HURJ Volume 23 - Fall 2016

Page 1

START YOUR MISSION Johns Hopkins alumna, Caroline Henriques, has worked for Textron, Inc. for four years and is now a Mechanical Engineer with Textron Systems Unmanned Systems in Hunt Valley, MD. She currently supports the Universal Ground Control Station, a command and control system for unmanned aerial vehicles. The UGCS provides warfighters with reliable intelligence during mission critical moments. From defense to commercial applications, Textron Systems employees design, engineer, and build products that are ready for any mission. Start your mission today at

Caroline Henriques, Mechanical Engineer

Textron Systems Unmanned Systems is a business of Textron Systems. Š 2016 AAI Corporation. All rights reserved.


Fall 2016:

International Conflict Resolution

table of contents fall 2016 focus: Building the Case for an IDF Intervention in Southern Syria: On National Security or Humanitarian Grounds? Jia Yao Kuek

Instrumental Democracy: An Examination of Jazz in Egypt Alejo Perez-Stable Husni

spotlight: The Consequences of Disproportional Publishing of Positive versus Negative Experimental Results Alex Serafini

The Relationship Between Sensory System and the Transmitted Content Kirsten Burke

table of contents articles: Craniostentosis: The Bioarchaeology of Catoctin Furnace Adrienne Kramer

Perceptual (A)symmetry: Visual Search and Change Detection for Basic Image Types J.J. Valenti

How Maternal Substance Abuse Affects the Physical Health of Children Milena Berhane

Altera: Heart Rate Variability Analysis Software for Small Animal Models Paul Kim

Corticosteroid Treatment as a Secondary Measure for Kawasaki’s Disease Sun Jay Yoo

Ethnic differences in parental feeding behaviors in a socio-economically diverse UK sample of preschoolers and their parents CeCe Gu

hurj editorial board

hurj fall issue 2016 contributors

Presidents James Shamul Hirsh Shekhar Editors-in-Chief Harry Burke Jaya Jasty Humanities & Social Sciences Emily Karcher Eric Guo Gulnar Tuli Veronica Reardon Lauren Pomerantz Science & Engineering Jane Miglo Aneek Patel Gurgan Raghuraman Cindy Li Aditya Mohan Layout Sarah Sukardi Cindy Li Copy Kyoung-A Cho Rebecca Lin Vicky Wang William Xie Fangjia Wen Wendy Tsai Sashinya DeSilva Jamie Chen Jhillika Patel Design Specialist Aneek Patel Aditya Mohan Webmaster Benjamin Bao

Adrienne Kramer Alejo Perez-Stable Husni Alex Serafini J.J. Valenti Jia Yao Kuek

Kirsten Burke Milena Berhane Paul Kim Sun Jay Yoo CeCe Gu

about hurj: The Hopkins Undergraduate Research Journal provides undergraduates with valuable access to research undertaken by their peers. The journal features four sections: a main Focus, Spotlight, Science & Engineering, and Humanities & Social Sciences. Students are highly encouraged to submit their original work.


The views expressed in this publication are those of the authors and do not constitute the opinion of the Hopkins Undergraduate Research Journal.

contact us: Hopkins Undergraduate Research Journal Mattin Center, Suite 210 3400 N Charles St Baltimore, MD 21218

a letter from the presidents


The Hopkins Undergraduate Research Journal was the first interdisciplinary undergraduate research journal on the Homewood campus. Within the pages of this issue, you will encounter a sampling of the incredible research engaging Hopkins students in the sciences, humanities, social sciences, and engineering. For each biannual issue of HURJ, the editorial staff selects one topic of interest to highlight in the magazine’s Focus section. This fall, we received submissions on “International Conflict Resolution,� which is a pressing message to spread around the world; we can come together even stronger than when we fight with each other. As you will find upon perusing the articles inside, the link between music, democracy, and cooperation in international relations is actually much more salient than at first glimpse. The authors have worked tirelessly with our diligent body of editors-in-chief, content editors and copy editors to assemble this elegant body of work for your enjoyment. Our Spotlight, Humanities & Social Sciences, and Science & Engineering sections showcase research endeavors that fall outside our chosen theme. This semester, we bring you articles on a fascinating range of subjects, including the sensory processing of artwork, genetics of ancient beings, substance abuse patterns, and computational heart rate analysis. The articles in this fall 2016 issue of HURJ reveal the diverse research interests of Johns Hopkins University students. We are proud of this publication as a testament to the boundless academic and intellectual talents of our peers. We would like to give special thanks to the HURJ staff and contributors for their hard work. Enjoy and Happy Holidays!

James Shamul Co-President

Hirsh Shekhar Co-President


hurj fall 2016: issue 22

Building the Case for an IDF Intervention in Southern Syria: On National Security or Humanitarian Grounds? Jia Yao Kuek, Class of 2019 International Studies

Introduction In this paper, I delve into the legality of possible Israeli military intervention in southern Syria. Concluding that the case for a legally justifiable, standalone Israeli Defense Forces (IDF) military operation in Syria is tenuous, I will instead offer up alternative scenarios for a limited use of force, and a joint Israeli civil-military intervention in southern Syria. Ultimately, after weighing the advantages and disadvantages of an Israeli military intervention in southern Syria, I conclude that by offering a synthesized approach that minimizes the risks to Israel, an IDF presence in southern Syria is a potentially advantageous course of action. First, I must qualify the use of several sources throughout my essay. Most of the papers cited offer arguments with specific reference to the United States (US); hence, the relevance of these points to the Israeli context should be viewed with a pinch of salt.

Legality of an Israeli Military Intervention

Considering the proximity of threats such as Hezbollah to the state of Israel, as well as Hezbollah’s well-publicized ongoing involvement in the Syrian civil war, a tentative case can be made for Israeli military intervention in southern Syria as an act of self-defense on the basis of ‘R2P’ alone. The legality of military intervention must also be examined with a contemporary lens, through the present-day framework of international relations. In the wake of the Cold War, the United States emerged as the preeminent power in the world and established itself as a ‘global policeman’ in enforcing international law. In keeping with this, Ikenberry described America as a liberal leviathan that runs a liberal hegemonic order in the post-Cold War world. However, its authority is not all-powerful, nor are its actions always consistent. A more recent piece by Dudney and Ikenberry describes the emergence of a nascent liberal internationalist system , similar to a Kantian triangle that balances international organizations, democracy, and economic interdependence. This relates to any proposed Israeli action in Syria – I believe that with the glare of international scrutiny, there is a need for Israel to keep its actions open to interpretation. Winning the information war is as crucial as any action taken by the IDF on the ground in Syria. As such, any proposal to set-up a buffer zone hinges on the success of efforts by the IDF and Israeli government to portray it as a justifiable course of action in front of the international community and the Israeli electorate, whether for national security or humanitarian reasons.

Winning the information war is as crucial as any action taken by the IDF on the ground in Syria.

Article 51 of the UN Charter that I will discuss below has traditionally been interpreted as excluding measures of self-defense against private actors, unless a degree of state involvement can be inferred. This leaves the question as to who decides whether such actions of self-defense are justified. A key tenet of international law is that it has no sovereign power with which to enforce its principles. International law is enshrined in treaties (which are neither necessarily binding nor enforceable), customs, and international tribunes such as the International Court of Justice or the Permanent Court of Arbitration. Ultimately, discussing the legality of any state-sponsored military action is not an open-and-shut case. International law depends on a variety of measures, such as sanctions, to support its implementation. These measures are seldom consistent, or universally applied. As seen from Russia’s unilateral annexation of Crimea, great powers such as Russia and the US have much greater bandwidth to exercise their military options within the grey area of international


law. In Russia’s case, it used the convenient excuse of ‘R2P’ or “Responsibility to Protect” (its ethnically Russian citizens in Crimea) as a legal justification for using armed forces in Crimea.

There are two avenues to provide legal justification for military intervention in Syria. Under Chapter VII of the UN Charter, use of force (authorized by the UN Security Council) is permitted when “necessary to maintain or

spotlight hurj fall 2016: issue 22

restore international peace and security”. Nonetheless, Russia and China’s continued opposition precludes the UNSC from adopting this course of action. The second scenario is provided under Article 51 of the UN Charter, which mentions the right of self-defense “if an armed attack occurs”. Israel’s history of military conflict with Syria provides a much stronger basis for the Israeli government to make the case that it faces a military threat on its northern border, requiring preemptive self-defense actions. Article 51 states that a state exercising its right of self-defense may do so until the Security Council can intervene, such that there is a period of time in which a state may act in self-defense without Security Council approval. Such unilateral actions on the part of the victim state operate in a legal grey area, yet provide for a preemptive use of force against anticipated attacks, under certain conditions. Hence, for Israel to make a convincing legal justification for a military intervention in Syria, it has to demonstrate the urgency of the threat facing it to the extent that it necessitates an immediate military response. Former Israeli Defense Minister Moshe Ya’alon declared in a June 2015 interview that Israel had three red lines proceeding any military intervention in Syria: Transfer of advanced weaponry to any terrorist group; transfer of chemical materials to any terrorist group; and any violation of Israeli sovereignty. As of April 2016, Israel


had conducted at least fifteen military strikes to enforce these lines. Despite Israel’s isolated military strikes in Syria, it continues to maintain its policy of non-intervention, as it deems its military responses as being proportionate to the opposing threat. One reason that Israel might avoid military entanglement in Syria is the necessity of maintaining its delicate ties with Russia. The risks of military intervention seem to outweigh its potential benefits, considering the potential damage that this would cause to Israeli-Russian ties. These ties are important but sensitive, as Russia remains to be a large source of Jewish migrants to Israel and still wields a degree of influence in the region. Crucially, Russia’s military involvement has been instrumental in engineering the Assad regime’s recent military successes, and reflects a turning point in the Syrian conflict in favour of the Damascus authorities. Russia’s continued military assistance to Hezbollah and the creation of a new Russian-Iranian-Syrian-Hezbollah axis poses a threat to Israel’s security interests. However, this works both ways, as Russia’s post-war influence in Damascus may be valuable to Israel by serving as a tether restricting the actions of Hezbollah or the Assad regime. Lastly, any discussion on the possibility of an IDF intervention in Syria also has to consider recent precedents of Israeli military operations outside its borders. The prospect of Israeli intervention in southern Syria differs


focus greatly from Israel’s existing threat scenarios, such as Israel’s ongoing conflict with Hamas. For the latter, and relating to previous clashes such as the 2014 Operation Protective Edge, there is little doubt as to Israel’s legal right to use military forces in the Gaza Strip, with the only questions being the scope (actions of self-defense must be proportionate to the threat) and objectives (parties are only allowed to attack military targets) of the operation. In the Syrian conflict, however, the international legal basis for military intervention is far from clear-cut. Despite the growing conflict in Syria, Israel has successfully isolated itself from the immediate effects of the crisis.

Advantages of Military Intervention A recent policy paper by The Institute for National Security Studies in Israel has also argued against military action, outlining several other reasons precluding an Israeli military intervention in southern Syria. However, I offer an alternative interpretation of these factors. Many see Israeli intervention as involving more risks than benefits, but it is possible that many factors perceived as risks may be advantages instead. Specifically, the paper highlights Israel’s preference for short-term tactical measures over long-term strategic measures as a factor that works against the prospects for military intervention. This is a much more risk-averse course of action, given the uncertainty over the Syrian conflict’s possible endgame. Indeed, based on the tumultuous nature of Middle East politics in general and the dynamic relationships between regional actors and external powers, I agree with the original premise. However, as aforementioned, Assad’s recent military successes point towards the all but certain survival of the Syrian regime, regardless of the outcome of the civil war. In the face of such developments, my opinion on what constitutes the best short-term tactical measure to adopt differs from that of the author’s. Commenting on Israeli Prime Minister Netanyahu’s June visit to Moscow, labeled the talks a failure since no understanding between Russia and Israel had been reached. Russia rejected Israeli requirements that it restrict supplies of weapons to Hezbollah weapons, in direct opposition to Israel’s national security interests. As such, I believe that an Israeli presence of any kind in southern Syria would greatly enhance its bargaining position vis-à-vis Russia. This, in a sense, can be viewed as a tactical move on Israel’s part. Another reason mentioned by the report speaks of Israel’s limited means of influencing the events in Syria, even as the threat of Israeli intervention already provides a credible means of deterrence. Building on this premise, I believe that establishing a safe zone in southern Syria would enhance Israel’s ability to respond to any provocation, while lending further credibility to its threats to intervene in an even greater armed capacity. One possible approach to tackle non-state actors (such


hurj fall 2016: issue 22 as Hezbollah) centers on the use of force and leverage in the areas of control, containment, marginalization and isolation. By establishing a ground presence in southern Syria, the IDF can more effectively conduct operations to restrict Hezbollah’s freedom to maneuver and communicate in the three southern-western governorates of Syria: Quneitra, Daraa, and As Suwayda. A parallel policy of marginalization and isolation can also be carried out amongst the local population, thereby stemming the spread of Hezbollah and Salafi jihadist political or ideological influence, and preventing them from building local bases of power. I have already discussed how Israel’s limited intervention in southern Syria gives it extra leverage in negotiations with Russia, thereby qualifying it as an act of tactical expediency. This previous paragraph, however, also demonstrates how the strategic implications of Israel’s inaction are potentially dangerous in the long run. The fundamental nature of threats posed by non-state actors, such as Hezbollah, arises not from their potential to wield advanced weaponry, but rather the entrenched yet haphazard nature of these armed elements within their area of operations. In Lebanon, Hezbollah serves as a social and political organization as much as it is a terrorist organization by building schools, hospitals, and maintaining efficient public services. Ultimately, Hezbollah’s involvement with terror is complicated by the legitimacy it enjoys amongst the local Shi’ite plurality, and even many non-Shi’ite Lebanese. In southern Syria, Hezbollah maintains a growing presence, particularly in Shi’ite-dominated areas such as the Syrian government-controlled city of Busra. Ultimately, I believe that Israel has a narrow window of opportunity to tackle the Hezbollah presence in southern Syria before it entrenches itself amongst the Shi’ite population of southern Syria and matures into a more formidable threat.

Cooperation with Jordan However, any success of the above suggestions hinges on a supportive local Syrian population. Doubts also exist regarding the reliability of any local Syrian partners for agreements, considering the existing mistrust and anathema with which Israel is regarded amongst the local Muslim population. To overcome this obstacle, as well as to establish the international legitimacy of such a safe zone, I propose enlisting Jordanian support in the plan’s implementation. In examining the reasons why Jordan would agree, we have to look at the main threats that the Jordanian government perceives will arise from the Syrian conflict. First, the paramount concern remains to be the threat of Islamic terrorism, with the Islamic State’s aspirations of a unified and centralized Arab caliphate in direct opposition to the rule of Arab monarchies such as the Hashemite Kingdom of Jordan. The Islamic State’s earlier incarnations, such as the Jordanian militant jihadist group Jama’at al-Tawhid wal-Jihad, espoused the goal of overthrowing the Jordanian

spotlight hurj fall 2016: issue 22


monarchy, as was publicly admitted by Tawhid operative Shadi Abdallah during the 2002 trial. Following the immolation of Lt Kasasbeh by Daesh militants in 2015, public opinion in Jordan turned decisively against Salafi jihadist elements, and this groundswell of public support can be tapped upon to overcome any reservation from cooperating with Israel in such an overt manner. Closely related to the first problem is the second issue of a growing refugee crisis on the Jordanian-Syrian border. In response to the Islamist threat and a suicide bombing that killed six Jordanian soldiers, Jordan declared its borders with Syria and Iraq as closed military zones in June. Nonetheless, closing the borders is a very immediate short-term solution, as it paves the way for a larger humanitarian crisis in southern Syria. A workable long-term solution to ensure refugee access to international aid, while minimizing the national security risks to Jordan, would similarly require the establishment of a buffer zone within southern Syria. In public, Jordan has consistently rejected any prospect of military intervention in Syria. In private, however, Amman has considered the creation of a buffer zone centered on Daraa, Syria, to stop the flow of Syrian refugees into Jordan. While both statements may seem contradictory, they mirror Israel’s similar considerations: The need to stay within a grey area by not siding decisively with any of the conflicting Syrian parties, while still taking precautionary measures to secure its territory. Nonetheless, the growing refugee pressures on its border with Syria necessitate a strategic rethink, with a new approach to prepare for a buffer zone in Syria. The logistics surrounding such an initiative would no doubt require the involvement of the Jordanian Armed Forces.

This begs the question of which courses of action Israel can pursue while exposing itself to the minimum level of risk. As such, I propose the creation of a safe zone in southern Syria as being a joint civil-military initiative, allowing it to resolve Israel’s national security interests while being packaged as a humanitarian mission that can likewise also benefit the flood of refugees fleeing the nucleus of the conflict further north. Another silver lining is the potential entry of other Arab states (particularly Jordan) into the Syrian conflict as well. The three southernmost governorates of Syria are wedged in on three sides by Israel and Jordan; this presents an opportunity for cooperation between both states in setting up a humanitarian buffer zone, but with the involvement and facilitation of both the Israeli and Jordanian militaries. This would also have the added advantage of rebutting accusations of an Israeli ‘invasion’ of Syria and diminishing comparisons with the IDF’s past occupation of southern Lebanon.

These risks mainly revolve around a potential international backlash against any Israeli ‘adventurism’ overseas

How to Interpret Israel’s Military Intervention To synthesize the previous sections, it is far from clear that a legal justification exists for Israel’s military intervention in southern Syria, with a potentially negative backlash from the international community if Israel does move troops into Syria. Despite the above points strengthening the argument for Israeli intervention in Syria, they do little to reduce the considerable risks that come with any physical military presence in southern Syria. These risks mainly revolve around a potential international backlash against any Israeli ‘adventurism’ overseas, or even domestic Israeli opposition to such a venture, given the painful memory of Israel’s intervention in Lebanon’s internal struggles in the 1980’s.

Humanitarian Intervention



The creation of such a buffer zone would facilitate the reconstruction of the administrative, economic, and social infrastructure of southern Syria. This centers on a 21st-century repackaging of humanitarian principles generally labeled as “R2P” or the “Responsibility to Protect” civilians. This cluster of ideas that seeks to delineate a state’s responsibilities in the international arena can be summed up as: 1) A state’s duty to protect civilians within its borders; 2) The international community’s responsibility to ensure that each state carries out this duty; and 3) A state’s right to use all means necessary to enforce this protection of non-combatants. R2P redefines the concept of sovereignty, such that a state’s autonomy within its own territory is neither complete nor unconditional. Nonetheless, the ambiguity surrounding what constitutes a failure to protect civilians on the part of the target state, as well as the range of options available to the international community in preventing these human rights violations, leaves any unilateral action potentially open to critique. In defining the IDF’s potential move into southern Syria, one must divorce this concept of ‘R2P’ from the traditional law that allows states to intervene in another country so as to protect its civilians or property, due to the negative historical connotations associated with this latter justification for action. Specific to Israel, this would spark parallels with the historical Arab-Israeli conflict and Israel’s control over areas such as the Sinai Peninsula, to ensure its national security. For international commentators to draw such parallels would hurt Israel’s overall goal of humanitarian assistance, while turning public opinion in


focus other Sunni Arab states against potential cooperation with Israel in southern Syria. Despite the security that the creation of a buffer zone would bring for many internally displaced refugees in southern Syria, I acknowledge that Israelis are under no moral obligation to assist the civilians there. However, regardless of the main humanitarian objectives of an IDF intervention in southern Syria, such actions as mentioned earlier would also be highly advantageous to Israel’s national interests. The domestic Israeli and international audiences have very different priorities. The Israeli government, when communicating the rationale for such a potential intervention, should adopt a nuanced and comprehensive argument that lists the benefits of a buffer zone, both for the Syrian people and for Israelis. The humanitarian basis driving the creation of a buffer zone is ultimately crucial, because it lends credibility to such a joint Israeli-Jordanian initiative.

hurj fall 2016: issue 22 tary force only as an enabler. The legitimacy of Israel’s possible military intervention in Southern Syria (in the eyes of the international community) hinges greatly on its success in the public relations sector. The recent shift in the tide of the Syrian conflict in favour of Assad does not preclude any Israel intervention in southern Syria, but rather, makes it more urgent for Israel to act to stem the creation of a Russian-backed, Iran-Hezbollah-Syria alliance along its border.


A Narrow Window of Opportunity In the end, perhaps precedent is the most important factor in determining public response to any Israeli plans to enter southern Syria. Irrespective of previous successful interventions (such as in the Kosovo crisis of 1999), or the more recent widely-perceived failures of Resolution 1973 in Libya, perhaps the most telling bell-weather of foreign intervention in Syria, whether humanitarian in nature or otherwise, would be the news of Turkey’s recent armored thrust against the town of Jarabulus in northern Turkey. Moving against both Kurdish Yekîneyên Parastina Gel (YPG) forces as well as the Islamic State, the Turkish military’s actions in support of Turkish-affiliated rebel groups set a massive precedent for other regional actors. Despite the clear national security interests driving Turkey’s actions, international criticism has been rather muted. US Vice President Joe Biden, on a visit to Turkey, even warned Syrian YPG units to withdraw east of the Euphrates or risk losing US support, while US warplanes supported the Turkish and Syrian rebel incursion. Overlooking the differences between Turkish and Israeli regional standing, these recent actions present the clearest window of opportunity for Israel and the IDF to move into southern Syria, on the basis of humanitarian intervention.

Conclusion To conclude, the legal justification for Israel’s military possible military intervention remains inconclusive, and crucially, hinges on the Israeli government making a convincing argument for the urgency of this security threat to Israel’s sovereignty, as well as the presence of any state involvement (whether Syrian or Iranian) in these non-state actors against Israel. In the absence of these arguments, the only convincing case for Israeli intervention has to be humanitarian in nature, with mili-


Jia Yao is a sophomore studying International Studies, Public Health, and East Asian Studies. He is an avid globetrotter who looks forward to serving in the Singapore Armed Forces upon graduation, and loves reading and riding in his spare time. Jia Yao writes on various issues that have touched him on travels, with a specific research focus on the geopolitics of Southeast Asia.

hurj fall 2016: issue 22


Instrumental Democracy: An Examination of Jazz in Egypt Alejo Perez-Stable Husni, Class of 2019 History It comes out of a communal experience. We take our respective instruments and collectively create a thing of beauty.”1 Jazz appears to be apolitical because it lacks lyrics yet Roach’s quote supports the idea that the ethos of Jazz—an improvised, communal music that does not require language to deliver its message—can still craft a powerful democratic statement. The exportation of Jazz to the Middle East began in the 1950s and 1960s when the United States government began sending Jazz “ambassadors” to play concerts overseas in places such as Cairo and Iraq. When Louis Armstrong visited Egypt, he brought the music of oppressed Blacks in the United States to the people of Egypt, thus the tours were a huge success. In terms of popularity, Jazz has never gained much traction in Egypt because older, orchestrated music known as al-Arabiyya al-Musiqa, western influenced pop music known as Shababi, and working class political music known as Sha’bi dominate the Arab world musical scene.4 Yet therein lies the politicization of Jazz—it is unclassifiable, it is malleable, and it does not seek to capture listeners. Novelist Ralph Ellison characterized Jazz as being par-

allel to the American political system; a soloist can play anything as long as the rhythm and harmony of the song are maintained.5 In this paper, I will explore how the ethos of Egyptian Jazz makes a democratic statement similar to how American Jazz had in previous decades, and that even though Egyptian Jazz may only culturally appeal the wealthier political progressive, it is an essential player in the movement against musical and political repression in Egypt. Modern day Egyptian musicians face censorship and lack of promotion unless they are producing music that is marketable in the eyes of mainstream music producers and the Ministry of Culture. One example of how Jazz has moved against musical repression is the founding of the annual International Cairo Jazz Festival in 2009, a festival that is funded independently with no corporate or government sponsorship. The festival seeks to bring Egyptians music that is not approved under military rule. However, the festival mainly attracts a largely foreign and upper middle class audience due to ticket prices. The festival exists on the margin of main-



hurj fall 2016: issue 22

stream musical production and has helped promote other unrecognized genres and styles.6 Although the festival is not very large, as only about five thousand people attend yearly, it represents a move away from the monotonous, controlled mainstream music scene and a move towards a free, collaborative, and improvised endeavor.7 In the context of Ralph Ellison and Max Roach’s ideas of Jazz as an unrestricted, independent, and democratic form of music, the Cairo Jazz Festival is an important example of the resistance to musical censorship in Egypt. Perhaps even more importantly, the Cairo Jazz Festival does not only seek to promote Jazz, but also to expose Egyptian people to an array of different cultures and forms of music that were not as available during military rule.8 Although its appeal is not massive, the Jazz scene in Egypt strikes a balance between a sophisticated musical endeavor and member of the music underground and the established, censored music that is aided by the mainstream labels and the Ministry of Culture. Jazz’s promotion as a democratic art form in Egypt is what separates it from its other independent counterparts. Another form of promotion that is key to the promotion of democracy through music is Mini Mobile Concerts, which are free concerts held by street musicians who play music ranging from Punk Rock to Jazz for a less affluent crowd.9 These musicians are funded independently and largely disagree with the way music has been promoted in Egypt. They see themselves trapped by the music industry and the Ministry of Culture.10 Both the International Cairo Jazz Festival and the Mini Mobile concerts serve as outlets of expression for musicians. In this sense, Jazz, a heterogeneous, polyglot genre, is the epitome of resistance to a homogenized music scene that does not support freedom of expression.

International Cairo Jazz Festival in Egypt. Amidst the rise of political oppression in Egypt, as Egyptian musician Randy Essam sought refuge in Sweden and sensitive articles were pulled from news sources, the ethos of Jazz as an improvisational and collaborative art form represented a strong resistance to this political censorship. Despite the chaos amid popular protests in 2011, the Cairo Jazz Festival was held that spring, and had, up until then, its most successful year. After the fall of Mubarak, this rise in attendance could be attributed to the newfound freedom and desire of the Egyptian people to absorb different styles of music and culture that drew them away from the political unrest.12 Although Jazz does not have a deep political history in Egypt, the music’s origin in the United States as the music of oppressed Blacks facilitates its politicization. Ultimately, as Amro Salah puts it, “Jazz is expressing freedom.”13 While the ethos of Jazz and the International Cairo Jazz Festival fosters democracy and expresses freedom, it is also important to contextualize the impact of the music in terms of Arabic music. Many would dismiss Jazz as music of protest due to its evolution from the clubs of New York City into a global music that targets a more privileged and increasingly white population.14 Although the production of Jazz may stand for the same ideals of freedom of expression as most Sha’bi music, Jazz is not truly Sha’bi music since it is not music of the streets of Egypt and does not stem from the Arabic tradition of poetry. Furthermore, Egyptian Jazz’s audience does not consist of poor and working class Egyptians. Yet if the American counterpart of modern day Sha’bi music in Egypt is Hip-Hop, then the concept of Jazz as music of protest must be seen in a different light. Jazz is rooted in the Black American struggle for civil rights and equality of the twentieth century and therefore making it more equitable with Sha’bi.

Before we can examine the impact of Jazz in Egypt, it is essential to see Jazz’s political impact in the United States. As a Black art form, Jazz started from the Swing Era, a fairly restricted music form with few differences in sound. Sonically, the Bebop era of the 1940s and 1950s was a drastic transformation of the Jazz scene. Harmony incorporated dissonance and musical extensions had never been explored. All tunes were played at a breakneck pace, and each section of the song was cluttered with hundreds of notes. Yet the Bebop Era gave way to a resolution of sorts in the Blue Note Era. After the Civil and Voting Rights Acts had passed, Jazz began to embrace many other styles of music, from Brazilian Bossa Nova to Spanish Flamenco.11 This simultaneous political and sonic resolution— although not perfect—created what has now been exported to other countries and the

While Jazz was not exported as music of protest in the same vein as Hip-Hop, both Jazz and Hip-Hop have held strong roots in the Black community of the United States as music that represented a struggle for equality, albeit in different points of time. By virtue of Jazz’s origins as music of democracy in the United States, the impact of Jazz in Egypt must be taken seriously. On one hand, Jazz may not be the prototypical Sha’bi music of Cairo’s poorest neighborhoods, but on the other hand, Jazz seeks to deliver a similar message, both musically and politically. As famous Jazz trumpeter Miles Davis once said, “Jazz is the big brother of Revolution, Revolution follows it around.”15

“Jazz is expressing freedom.”


hurj fall 2016: issue 22 References (1) BBC. “Egypt Profile.” BBC. February 17, 2015. March 1, 2015. <>. (2) Billet, Alexander. “Thanks, Stanley Jordan, for Pulling out of Israeli Jazz Fest.” The Electronic Infitada. January 11, 2013. February 24, 2015. < thanks-stanley-jordan-pulling-out-israeli-jazz-fest>. (3) Crouch, Stanley. “Putting the White Man in Charge.” JazzTimes. April 2003. March 1, 2015. <>. (4) Danielson, Virginia. “New Nightingales of the Nile: Popular Music in Egypt since the 1970s.” Popular Music 15, No. 3 (October 1996): 299-312. March 1, 2015. <>. (5) El-Saket, Ola. “Telling the Story of Egyptian Jazz.” Egypt Independent. March 3, 2012. February 24, 2015. < story-egyptian-jazz>. (6) Kaplan, Fred. “When Ambassadors Had Rhythm.” The New York Times. June 18, 2008. February 24, 2015. <http://www.nytimes. com/2008/06/29/arts/music/29kapl.html?_r=1&>. (7) Kirkpatrick, David D. “Egypt Seizes Newspapers to Censor an Article.” The New York Times. October 1, 2014. February 28, 2015. <http://www.nytimes. com/2014/10/02/world/middleeast/egypt-seizes-newspapers-tocensor-an-article.html?_r=0>. (8) Morgan, Marwa. “Egyptian Singer Moves to Sweden Seeking ‘Safe City Residency’.” Daily News Egypt. October 29, 2014. February 28, 2015. < safe-city-residency/>. (9) Muscara, Aprille. “Q&A: Amro Salah, Founder of Cairo Jazz Festival.” Scoop Empire. March 13, 2014. February 24, 2015. <http://>. (10) Pinkerton, Steve. “Ralph Ellison’s Righteous Riffs: Jazz, Democracy, and the Sacred.” African American Review 44, No. 1/2 (Spring/ Summer 2011): 185-206. March 1, 2015. <>. (11) Pomerantz, James. “Happy Birthday, Miles Davis.” The New Yorker. Last modified May 24, 2013. Accessed March 1, 2015. http://www. birthday-miles-davis. (12) Roshdy, Adham. “The Egyptian Underground Music Scene.” Daily News Egypt. Last modified December 10, 2012. Accessed February 24, 2015. the-egyptian-underground-music-scene/. (13) Sprengel, Darci. “The International Cairo Jazz Festival.” Ethnomusicology Review 18 (August 23, 2013). Accessed February 28, 2015. (14) “The International Cairo Jazz Festival and Mini Mobile Concerts: Two Musical Approaches to ‘Post’-Revolutionary Egypt.” Ethnomusicology Review 18 (October 8, 2013). Accessed February 24, 2015.

focus Author

Alejo is majoring in History and plans to declare a second major in Economics. His research interests include Intellectual History, Critical Theory, and Political theory. Although born and raised in San Francisco, he has adjusted to life on the eastern seaboard, but will always miss the West Coast. Alejo’s career aspirations are uncertain, but he wouldn’t mind becoming an academic.


The Consequences

Of Disproportional Publishing of Positive versus Negative Experimental Results

Alex Serafini Neuroscience Class of 2018

hurj fall 2016: issue 22 The Current Trend in Medical Publications (Positive versus Negative Results) Positive findings are the only good findings. After all, that is what scientific publishers want, as Dwan et al. found that studies reporting positive (or significant) results are both more likely to be published and fully reported.1 However, from an objective point of view, all findings should be considered equal. In other words, a lack of a significant relationship between two factors should be just as important as a significant relationship that occurs. The disproportional publication of positive over negative findings should not be taken lightly, as publication of negative findings prevents resource wastefulness by barring other scientists from performing the same experiments in search of significant findings, while also keeping an unbiased distribution of published scientific literature.2

Figure 1. “Admission Rates of Questionable Research Practices (QRP) in self and non-self reports.” This data suggests that there is a very high level of publication of falsified data and studies conducted with bias or carelessness, even on a very small scale (Fanelli, 2009)3

It is one things to be wasteful with resources by showing preference for positive over negative research results, but it is criminal to falsify data (through “Questionable Research Practices” – QRPs) in order to improve one’s chances of achieving a high-impact publication – with reason. Scientific research serves as the foundation for some of the largest industries in modern society, such as the pharmaceutical industry. The problem with QRPs, according to John et al., is that increasing prevalence of these practices produces pristine results that are nearly impossible to achieve through unbiased research methods. This, in turn, threatens the integrity of the research field as a whole, as it becomes a “race to the bottom” – who can falsify data most efficiently to get the major publication first.4 It is difficult to imagine pharmaceutical companies, whose values are directly related to improving patient health, partaking in this data falsification. But it happens – and more than one can imagine.

spotlight How Merck’s Questionable Research Practices Hurt/Killed Thousands For an example of this, it is fit to look at Vioxx, a nonsteroidal anti-inflammatory drug (NSAID) introduced to the market by Merck Pharmaceuticals in 1999. Vioxx was commonly prescribed for arthritis, acute pain, and painful menstrual cycles.5 In 2001, however, a study of Vioxx called VIGOR (Vioxx Gastrointestinal Outcomes Research – a study attempting to show that Vioxx caused fewer gastrointestinal complications than similar treatments; n = 8076 patients) found that the drug caused twice as many heart attacks, strokes and cardiovascular deaths than the control group. On the publication in the New England Journal of Medicine, the following statement was used to explain this phenomenon: “The difference in major cardiovascular events in the VIGOR trial may reflect the play of chance,” the reason being that “the number of cardiovascular events was small” (less than 70).6 Later on in the article, however, the authors highlighted that there were only 53 patients with gastrointestinal issues, even though 70 cardiovascular complications were not considered important. By all means, the presentation of this data falls within the realm of Questionable Research Practices, specifically biased findings. What is more important is the consequence of not withdrawing the drug after these findings: over 27,000 heart attacks or cardiac deaths.7

The Link Between Increasing Rates of Positive-result Publications and QRPs

Figure 2. Percentage of publications (per publication year) claiming positive findings versus those claiming negative findings (de Winter and Dodou, 2015).8


spotlight As aforementioned, more positive findings are published in journals today than negative findings, but what exactly have been the trends over the last two decades regarding this matter? Furthermore, if there has been an increasing trend in the publication of positive findings, it is important to consider whether or not this trend is linked to an increase in QRPs. A study performed by de Winter and Dodou analyzed an increase in both positive findings (0.041<p<0.049) and negative findings (0.051<p<0.059) since 1990 (figure 2) – without a doubt, the rate of publishing positive findings has been increasing at a much higher rate than that of negative findings.8 The difficulty with comparing publication trends regarding positive/negative results with QRP trends is that one must be able to find a study that sets consistent criteria over a long period of time. One measure of QRPs that will serve well in this analysis, however, is the retraction of publications for fraud over time (figure 3), of which there is also a heavy increase.9

The Privatization of Scientific Research

hurj fall 2016: issue 22 Vioxx scandals and put the lives of thousands of patients at risk. This does not mean that pharmaceutical industry research and products should be deemed unethical or untrustworthy as a whole – in fact, society owes a debt of gratitude to these companies for their impact on mortality in the U.S. According to the PhRMA 2015 Biopharmaceutical Research Industry Profile, the pharmaceutical industry was responsible for preventing 862,000 premature deaths in the United States due to HIV/AIDS alone. Furthermore, there has been a 22% decline in cancer death rates since the 1990’s, with 83% of survival gains being attributed to newly developed treatments. However, these treatments come at a cost, a large one at that, which might be where the problems previously discussed stem from – the average cost for the development of a single drug is approximately $2.6 billion, with 23.4% of all drug sales going back into research and development for new drugs.10 The incentives to recover the large expenses incurred in drug development and yield high profit margins to pay off shareholder dividends is burdensome for pharmaceutical companies – so much so that failure is not an option, and QRP becomes one of the few routes for survival.

Courses of Action to Change the Direction of Current Research Trends

Figure 3. Papers published and subsequently retracted since 1973 (focus on retractions for fraud, although error could be considered QRP in certain situations). It is also important to note that the sharp decline in fraudulent retractions around 2007 is most likely due to the fact that papers published recently have not yet been retracted (Steen et al., 2013).9

Fraudulence, especially in research, has always been an issue, and will continue to be one. The real concern at this point is that trends in “positive publishing” and fraudulent retractions (assuming that recent publications have not yet had the chance to be retracted) are continuing upwards. But why now? The answer might lie in the privatization of research. According to Dr. Johns Abramson, author of Overdo$ed America, approximately two-thirds of current clinical trials in the United States are being funded or performed by for-profit research companies.6 Abramson goes on to quote Dr. Drummond Rennie, the deputy editor of the Journal of the American Medical Association, regarding the privatization of university-run medical research: “They are seduced by industry funding, and frightened that if they don’t go along with these gag orders, the money will go to less rigorous institutions. It’s a race to the ethical bottom.”6 In other words, success in many areas of research is no longer determined by integrity and lack of bias, but rather who can get the results that support corporate interests the most. It is these interests that lead to


If changes are to be made that can return scientific research to an unbiased process, it is not going to be through pharmaceutical legislation reform, as that would simply be too large of a task, considering how interwoven pharmaceutical companies and regulatory agencies have become. For example, in an analysis of FDA employee conflicts of interest (CoIs), not only was there a relatively high number of CoIs reported (figure 4), but a large number of those CoIs were related to consulting income (or, in other words, associations with the private sector, such as pharmaceutical companies).11 Therefore, one must consider what has changed recently with regards to the funding and performance of scientific research. The first major change has been the skewing of funding towards the private sector, which leads to incentivized research. By allocating more government finances to grant funds, such as those provided by the NIH, researchers would be relieved of privatized pressure and could perform their research without bias, at least for the most part. There would still be the issue of publications preferring research that presents positive results. This is where a regulatory committee assembled by the government would come into the equation. This committee would be able to establish legislation that could 1) incentivize medical publishers to have a required, minimum number of negative-results publications per issue, 2) identify and resolve any outstanding conflicts of interest within regulatory agencies, and 3) prevent privatized pressure on research by requiring an in-depth analysis of data compilation and presentation in order to de-incentivize QRPs.

Concluding Remarks The research community has reached an impasse – one

hurj fall 2016: issue 22


Figure 4. Percentage of meetings where at least one CoI was reported by an FDA employee (n=221) and percentage of CoIs by “Person-Meetings” (n=3718). (right) Distribution of CoIs with range of income per employee (Lurie et al., 2006).11

where it can decide to salvage the concept of unbiased research aimed at improving the human condition, or where it can decide that it is appropriate to use fraudulent data in order to boost corporate profits. Once again, this is not to say that corporations such as pharmaceutical companies are inherently bad – our community owes a lot to these companies, as they have alleviated a significant amount of human suffering over the last few decades. What this article does call for, however, is tighter regulation and an increase in government funding of research. It is morally unethical for such a large proportion of the research field to be funded by the private sector, as this research will most likely be inherently biased. Just one case of a Questionable Research Practice led to the suffering of over 27,000 individuals. A change needs to be made now, because the scientific community is heading in the wrong direction.

too). PeerJ 3:2015. <>. (9) Steen RG, Casadevall A, Fang FC. Why Has the Number of Scientific Retractions Increased? PLoS ONE 8(7): 2013. < plosone/article/asset?id=10.1371/journal.pone.0068397.PDF>. (10) “2015 profile: Biopharmaceutical Research Industry.” PhRMA. <http://>. (11) Lurie P, Almeida CM, Stine N, Stine AR, Wolfe SM. Financial Conflict of Interest Disclosure and Voting Patterns at Food and Drug Administration Drug Advisory Committee Meetings. JAMA 295(16):2006. <http://jama.>.


References (1) Dwan K. et al. Systematic Review of the Empirical Evidence of Study Publication Bias and Outcome Reporting Bias. PLoS ONE 3(8): 2008. < pone.0003081.pdf>. Web. (2) Granqvist E. “Why science needs to publish negative results.” Elsevier. 2 Mar. 2015. <>. Web. (3) Fanelli D. How Many Scientists Fabricate and Falsify Research? A Systematic Review and Meta-Analysis of Survey Data. PLoS ONE 4(5): 2009. < pone.0005738>. (4) John LK, Loewenstein G, Prelec D. Measuring the Prevalence of Questionable Research Practices with Incentives for Truth Telling. APS 20(10): 2012. <>. (5) “Vioxx (rofecoxib) Questions and Answers.” FDA. 30 Sep. 2004. <>. (6) Abramson J. Overdo$ed America: The Broken Promise of American Medicine. New York, New York: Harper Collins Publishers, 2008. Print. (7) Berenson A, Harris G, Meier B, Pollack A. “Despite Warnings, Drug Giant Took Long Path to Vioxx Recall.” The New York Times. 14 Nov. 2004. <>. (8) de Winter JC, Dodou D. A surge of p-values between 0.041 and 0.049 in recent decades (but negative results are increasing rapidly

Alex Serafini is a junior from Sunnyvale, CA who is researching the molecular mechanisms of chronic pain through the BS/MS Neuroscience Program. He will be attending Mount Sinai School of Medicine in 2018, with the goal of combining research, clinical practice, and business to improve the quality of industrialized medicine.


spotlight Introduction It is known that the visual and auditory sensory systems are not completely independent; that one can affect the other. At the physiological level, Robinson and Sloutsky (2013) have shown that vision and hearing share some of the same processing components. They suggest that “cross-modal competition can occur at different points in the course of processing, with auditory input attenuating visual processing and visual input dominating the response.” This suggests that at some point in the processing of sensory information, the content does not play a role in sensory processing. At the psychological level, on the other hand, Baumgartner (2005) has shown that music can enhance the emotional experience of pictures. This suggests that at the psychological level, the content of stimuli, in this case the music and photographs themselves, affect the affective experience. Baumgartner employed photographs from the International Affective System and well-known classical recordings. Extending upon Baumgartner study, Jeong et al. (2011) focused on the influence of music on the perception of emotional facial expressions. They found that subjects rated “happy” or “sad” facial expressions at higher frequency when simultaneously exposed to music of a concordant emotion (e.g., a happy facial expression and happy music) and lower when exposed to music of a discordant emotion.

hurj fall 2016: issue 22 were Caspar David Friedrich’s Easter Morning (Figure 3) and Jan van Goyen’s Landscape with Two Oaks (Figure 3) (shown below). The two “happy” musical excerpts were from Franz Liszt’s Les Preludes and Giuseppe Verdi’s Overture to Nabucco and the two “sad” excerpts were from Daft Punk’s Adagio for Tron from the movie Tron: Legacy and Hans Zimmer’s A Way of Life from the movie The Last Samurai. PsychoPy software was used to preform this psychological test. In the alone condition, subjects were individually shown each painting and asked to rate the work; likewise, they were provided each music excerpt and were asked to rate piece. In the combined condition, subjects were presented with both a picture and music excerpt, and asked to rate them. The ratings were based on a 15-point scale of emotional valence, ranging between -7 (saddest) to +7 (happiest). Subjects were played random noise, and engaged in visual distraction tasks, between each stimuli presentation in the various conditions. All stimuli were randomized in terms of their presentation. The paired t-test was used to compare the subjects’ answers across the alone and combined conditions.

What is not known is whether the combination of visual and auditory stimuli can have an additive effect on emotion. In other words: When stimulus that elicits a certain level of emotion is combined with another stimulus with its own level of emotion, is either stimuli affected by the other stimulus? If so, what is the relationship between their changes, and what does this mean for our understanding of perception? I hypothesize that both visual and auditory modalities can affect each other, but that the effect depends, in part, on the content of the stimuli. In other words, although the sensory modality plays a role in perceptual processing, the content also plays a role in perceptual processing. Therefore, when we examine multi-modality processing, we are observing the effects of the sensory physiology and the content of the sensory stimuli; since the modality and its content can never be separated, our understanding of perception must be one that combines both the sensory modality and the stimuli content. The sensory system and the content of the stimuli cannot be studied separately because there is never a content-independent perceptual processing.

Figure 1

Methods Four paintings, two “happy” and two “sad”, and four short musical excerpts, two “happy” and two “sad” were selected. The two “happy” paintings were Pieter de Hooch’s A Woman Drinking with Two Men (Figure 1) and Carl Spitzweg’s Summer Day (Figure 2), and the two “sad” paintings


Figure 2

The Relationship

Between the Sensory System and the Transmitted Content

Kirsten J. Burke Art History Class of 2017

Sad music was affected by sad paintings. Discordant stimuli: Happy paintings were affected by sad music. Happy music was not affected by sad painting. Sad paintings were affected by happy music. Sad music was affected by happy paintings. (Significant findings are in bold)

Discussion Figure 3

Figure 4


Previous studies to examine the interaction of visual and auditory processing have focused on the ways in which the two modalities can affect each other at the sensory level, without taking into account the content effects of visual and auditory stimuli. However, this study found that “happy” and “sad” do not have equal effects; that it is hard to affect “happy” and relatively easy to affect “sad”. These results suggest that the content of the stimuli, as well as the modality of the stimuli, affect emotional states. Furthermore, there does not appear to be such a thing as a “pure” stimulus in perceptual processing, as perception involves the interaction of sensory and content processes. The existence and character of the cross-model competition depends in part on the nature of the stimuli. An implication of this finding is that, although a great deal of the physiological and psychological literature is based on the assumption that one can separate the sensory system from the sensory content, and study each separately, I found that the sensory system and the content of the stimuli cannot be studied separately because there is no content-independent perceptual processing.

References (1) Baumgartner, T., Esslen, M., & Jäncke, L. (2006). From emotion perception to emotion experience: emotions evoked by pictures and classical music. International Journal of Psychophysiology, 60(1), 34-43. (2) Jeong, J. W., Diwadkar, V. A., Chugani, C. D., Sinsoongsud, P., Muzik, O., Behen, M. E., ... & Chugani, D. C. (2011). Congruence of happy and sad emotion in music and faces modifies cortical audiovisual activation. NeuroImage, 54(4), 2973-2982. (3) Juslin, P. N., & Laukka, P. (2004). Expression, perception, and induction of musical emotions: A review and a questionnaire study of everyday listening. Journal of New Music Research, 33(3), 217-238 (4) Robinson, C. W., & Sloutsky, V. M. (2013). When audition dominates vision: Evidence from cross-modal statistical learning. Experimental psychology, 60(2), 113.


Concordant stimuli: Happy paintings were not affected by happy music. Happy music was not affected by happy paintings. Sad paintings were not affected by sad music.


Kirsten Burke is a senior History of Art major from Charlottesville, Virginia. She has conducted research for the Woodrow Wilson Fellowship and Provost’s Undergraduate Research Award on topics ranging from Northern Renaissance art to psychology and perceptual processing. After graduation, she plans to pursue graduate school in art history.

hurj fall spring 2016: 2016: issue issue 22 22

humanities and social sciences

Craniostentosis: The Bioarchaeology of Catocin Furnace Adrienne Kramer, Class of 2017 Archaeology

Abstract Bioarchaeological research was conducted on the skeletal remains of African slaves who labored at the Catoctin Furnace, an industrial village continuously occupied from 1776 to 1903. The remains of 34 individuals were examined by a team at the National Museum of Natural History, Smithsonian Institution under the supervision of Douglas Owsley. A number of pathologies were identified. Among these, multiple Schmorl’s depressions and arthritis, associated with hard physical labor, and craniostenosis are considered significant. My analysis focused on craniostenosis, a syndrome that results in premature cranial suture closure. The results of this study indicate that the incidence of craniostenosis is high. Six out of 34 individuals were identified with the sagittal form. Comparison with environmental factors and other genetic diseases, such as rickets, is examined to account for the high frequency of the syndrome in this population. One hypothesis, which explains the large number of individuals with craniostenosis compared to the estimated probability of being born with the syndrome, is that craniostenosis is genetically linked. Further research is proposed.

Bioarchaeology of Slaves There are specific bioarchaeological markers observed in slave populations that can be examined in order to reveal systems of enslavement in the past. Bioarchaeological research is most effective “when the data derived from the bodies of individuals believed to have been slaves... are incorporated within the broader ethnohistoric and archaeological records” (Harrod and Martin 2015: 52). Because these individuals were forced into “systems of exploitation,” they were treated differently, and these differences can be observed in the skeletal remains (Harrod and Martin: 2015: 52-53).

Skeletal analyses can disclose the “accumulative effects of subjugation and hard labor” (Harrod and Martin 2015: 54). For example, cribra orbitalia is a type of skeletal display that reflects malnutrition (Boyd 1996: 192). Also, osteoarthritic patterns in joints allow researchers to reconstruct occupational history, which is especially important in slave populations to ascertain what types of physical labor individuals did (Boyd 1996: 193, 225). Analyses of skeletal trauma and malnutrition can be interpreted by bioarchaeological researchers to “document excessive, grueling, long hours of physical labor, poor diet and repeated bodily abuse” (Harrod and Martin 2015: 54). Additionally, stable carbon isotopes in bone and teeth can be assessed to record diet, health, and activity patterns (Boyd 1996: 190). For example, certain levels of isotopes can be suggestive of a diet high in specific foods. This allows researchers to infer relatedness, and can also indicate migration (Boyd 1996: 196, 205).

Previous Archaeological Research Doug Owsley, head of the division of Physical Anthropology at the Smithsonian National Museum of Natural History, began to expand Angel’s preliminary analyses in 2015. New technological advances, such as stable isotope data, mitochondrial DNA analysis, and craniometric data, have allowed Owsley and his team to refine Angel’s original research (Anderson 2015). For example, Kelley and Angel (1983: 3) reported that out of 27 individuals, ten had bowed femora, which they attributed to malnutrition during adolescence. However, Owsley and the Smithsonian team reported just two individuals with bowed femora, out of a total of 34 (Smithsonian Institution n.d.). Additionally, Kelley and Angel (1983: 4) identified only one child, aged three,


humanities and social sciences

hurj fall 2016: issue 22

with premature closure of the sagittal suture, but Owsley’s Smithsonian team reported two individuals between the ages of 0-5 with sagittal stenosis (Smithsonian Institution n.d).

Results 34 graves were excavated at the cemetery site; some were marked by head and foot fieldstones (Anderson 2013: 67; Burnston and Thomas 1981). Three graves did not contain any human remains (Smithsonian Institution n.d.). The individuals ranged from neonate to over 60 years in age (Figure 1). Overall, the causes of deaths were unknown (Smithsonian Institution n.d.). There are a wide variety of associated pathologies evident for each individual (Figure 2). Only eight of 34 were found with no identifiable pathology (Smithsonian Institution n.d.). Some of these pathologies include bowing femora associated with malnutrition, dental pathology, spinal degeneration and multiple Schmorl’s depressions associated with hard physical labor, and craniostenosis. In one grave, a mother and child were found buried together. Both displayed sagittal stenosis.

Figure 1

Figure 2


Figure 3

Discussion of Craniostenosis Craniostenosis is a premature fusion of cranial sutures, and can be easily detected in skulls found in archaeological sites (Johnson and Wilkie 2011). The early fusion of cranial sutures causes the skull to become misshapen as it grows in the direction of least resistance. Identification of individuals with the syndrome is fairly simple (Gordon 1959: 793). Sagittal stenosis is a type of isolated craniostenosis that shows a single, prematurely fused suture (Panigrahi 2011). Six of the 34 individuals in the Catoctin Furnace sample showed craniostenosis, and more specifically, sagittal stenosis (Smithsonian Institution n.d.) (Figures 2 and 3). Thus, the individuals in the sample had long narrow heads, and the area that should reveal the sagittal suture was completely smooth. The proportion of sagittal stenosis in the Catoctin Furnace sample is considered uncommonly high (Figures 2 and 3). For example, the Catoctin Furnace sample represents a 17.6% occurrence versus a 0.04% chance, or 1 in 2500 births, estimated today (Panigrahi 2011). As Angel observed in his preliminary study, the syndrome was detected in the skulls of children (2) and teens (3) (Figure 3). The Owsley team identified the syndrome in a larger age group, including a 30-35 year old. These data also account for the woman and child with the syndrome that were buried together. Owsley and his team did not believe that craniostenosis was a cause of death for these individuals, and there were no skeletal defects or other anomalies found on the skeletons that are normally associated with non-isolated forms of craniostenosis (Panigrahi 2011). Based on research on types of craniostenosis, the type of stenosis found in individuals at Catoctin Furnace is most likely isolated sagittal stenosis. This type of craniostenosis was further researched to analyze the prevalence of this syndrome in the Catoctin Furnace group.

hurj fall 2016: issue 22

Figure 4

Genetic Links In contemporary populations, craniostenosis is an uncommon syndrome, with only about a 0.02% estimated incidence for sagittal stenosis (Boyadijev 2007: 131). According to one study by McGillivray et al. (2005) on familial scaphocephaly, sagittal stenosis is the most common form of craniostenosis and occurs in 4058% of cases. This statistic also held true in two other studies conducted by Boyadijev (2007) and Panigrahi (2011). Although sagittal stenosis is the most commonly affected suture in craniostenosis cases, the general incidence of craniostenosis is quite low. There is some question as to whether genes play a role in transmission of the syndrome. The prevalence was about 440 times higher in the Catoctin Furnace population than in the estimated birth incidence. Panigrahi’s (2011) study indicated that isolated craniostenosis is a “clinically and genetically heterogeneous condition.” This point is crucial because a dominant heterogeneous trait would explain the high incidence of the syndrome in the Catoctin Furnace population. Additionally, it was found that non-syndromic craniostenosis is thought to have a “strong genetic component” and possibly a gene-gene or gene-environment interaction (Boyadijev 2007: 129). Muenke (2014) cited a study done in 1995, which claimed that mutations of the FGFR1 and FGFR2 genes cause genetically linked craniostenosis. Furthermore, a continuation of this research found a three point mutation of the FGFR genes linked to autosomal dominant craniostenosis (Muenke 2014). An autosomal dominant genetically linked syndrome would explain a high prevalence of the syndrome in a population because a majority of the individuals would be carriers for the trait. Boyadijev (2007: 132) asserted that although the sagittal suture is the most commonly affected suture in craniostenosis cases, isolated sagittal stenosis is rare and the genetic cause is unknown.

Environmental Conditions The possibility of a gene-environment interaction as a risk factor in craniostenosis should be considered because the individuals lived and worked under harsh

humanities and social sciences conditions. It is possible that a susceptible gene, such as the FGFR series, could interact with the environment in such a way to change the gene epigenetically. Thus, craniostenosis could be hereditary, or a result of environmental conditions, or both (Boyadijev 2007: 130). The environmental conditions of slaves at the site were very poor. It is important to note how much physically demanding labor went into simply fueling the furnace. “Burning crew” members lived in pits of charcoal for four days straight at a time keeping watch over the coaling process (Anderson 2013: 70). Burning the charcoal in the mountains often caused raging mountain fires, such as one in April 1826 that lasted three days (Anderson 2013: 69). However, these poor conditions impacted not only male workers but also their families. They lived so close to the furnace that the “flames and gases [flickering] over the furnace stack” allowed “children in Catoctin Furnace [to] live with an almost constant fireworks display” (Anderson 2013: 79). When the furnace was fueled, sparks and ash would settle over the houses, machines, roads, water, and anything else nearby (Anderson 2013:79). The consistent smoke and ashy air conditions, and the constant exposure to chemicals from the burning of the furnace are important factors to consider when analyzing the impact of environmental conditions. Increased parity and maternal smoking are also risk factors (Boyadijev 2007: 130). These factors correlate well with conditions at Catoctin and slave sites in general.

Association with Rickets It is important to note the possible relationship between rickets and craniostenosis. While many cases of rickets are due to malabsorption of vitamin D or malnutrition, there are some cases where “resistance to normal doses of vitamin D” is due to an enzyme deficiency (Coleman and Foote 1954: 561). The deformities that occur with familial rickets are similar to those of classical rickets, however, “in the familiar resistant variety craniostenosis usually occurs” (Coleman and Foote 1954: 561). One case demonstrated that in an otherwise healthy two-year-old whose mother had rickets, had both craniostenosis and rickets, although their diets were not deficient (Coleman and Foote 1954: 562).

Comparative Data A recent excavation of an African slave cemetery in New York contributes to the bioarchaeological research on American slave sites and slave data. The DNA of thirty- two individuals was extracted and showed a “high probability of maternal ancestors” living in various locations in Africa (Mack and Blakey 2004: 11). Similar to individuals found at Catoctin Furnace, the skeletons showed evidence of intense physical labor, poor dental health conditions, and malnutrition (Mack and Blakey 2004: 11-13). One child exhibited “premature sagittal suture closure,” which is thought to be a result of “congenital, nutri-


humanities and social science tional and/or mechanical causes” (Mack and Blakey 2004:13). This is the same type of craniostenosis exhibited in the Catoctin Furnace population, and the researchers concluded a gene-gene or gene-environment interaction as the cause. The conclusion made by Mack and Blakey was very similar to the one formed in this paper, although the prevalence of craniostenosis was much lower in the New York burial population compared to the Catoctin Furnace population. However, the incidence of craniostenosis at the New York burial site is still 78 times higher than the general estimation, giving reasons to believe that the incidence of craniostenosis in slave populations does not occur by chance alone. Additionally, there were remains found showing osteological markers of traumatic effects of physical labor and illness at birth, signaling “poor maternal health” (Mack and Blakey 2004: 13). As previously discussed, poor maternal health is thought to increase the risk factor for craniostenosis. The living conditions at Catoctin Furnace were discussed earlier in the paper. Living conditions at the New York site were also grim, as Mack and Blakey claimed that there was increased disease risk due to “environmental insults” (2004: 12). The remains of children were found with traumatic injuries that were “indicators of children’s exposure to hazardous work environments” (Mack and Blakey 2004: 13). Quite similar to Catoctin Furnace where the slave families lived in close proximity to the furnace and were exposed to its dangerous effects, the New York site slave families also lived near their working location and were constantly exposed to the effects of their work. The comparison between the New York burial and the Catoctin Furnace grave site demonstrates the consistency in poor living and working conditions of slaves, and its effects on their health, disease, mortality, and the prevalence of syndromes such as craniostenosis.

hurj fall 2016: issue 22 ulation. It is possible that the two are genetically linked, although no individuals were found with both bowed femora and craniostenosis.

Further Research Further research could include a more in depth analysis of the environmental and physical conditions at the furnace that may have had epigenetic effects on the population. Because research thus far has shown that it is possible that the high frequency of craniostenosis in this population may be associated with external risk factors like environment and maternal physical condition, studying what these conditions were like may help answer some more difficult questions about craniostenosis or causes of death. DNA analysis has the ability to tell what genes may have had mutations or epigenetic changes, thus assisting the search for the type of craniostenosis and its cause. DNA research may also be able to reveal kinship connections and ancestry, which is important to broader research about slaves.

DNA analysis has the ability to tell what genes may have had mutations or epigenetic changes.

Conclusions This paper examined possible causes for the unusually high prevalence of sagittal stenosis in the slave population from Catoctin Furnace. Based on skeletal analysis alone, the most common type of craniostenosis found in the population is isolated sagittal stenosis. This form of craniostenosis is associated with gene mutations on the FGFR gene series. Other studies show that frequency of craniostenosis can be affected by environmental conditions and external risk factors such as maternal health. The risk factors associated with craniostenosis are quite similar to those that would have been present at a furnace site. Additionally, the paper touched on the association between rickets and craniostenosis, as individuals with bowed femora were also found in this pop-


Given the comparisons between data from the New York site and Catoctin Furnace, it is critical to continue to seek out comparative data that indicates the frequency of craniostenosis in the African slave diaspora. Moreover, it would be worthwhile to investigate the incidences of craniostenosis among the African nations from which the slaves at Catoctin Furnace originated.

Acknowledgements I would like to thank Professor Lisa DeLeonardis, Johns Hopkins University for guiding me through the research process and for aiding me in writing this paper. I would also like to thank and acknowledge Dr. Elizabeth Anderson Comer for the opportunity to conduct an internship with the Catoctin Furnace Project. Dr. Doug Owsley, Kari Bruwelheide, and the rest of the team at the Smithsonian’s National Museum of Natural History are thanked for their guidance, allowing me to take part in their work, and access to their research.

References (1) Anderson, Elizabeth 2013. Catoctin Furnace: Portrait of an Iron-making Village, edited by Elizabeth A. Comer. History Press, Charleston and London. (2) Anderson, Erik 2015. Buried Secrets: Study of Skeletal Remains May Hold Key to Slave History. The Frederick News Post, April 12. (3) Boyadjiev, Simeon 2007. Genetic Analysis of Non-syndromic Craniosynostosis. Orthodontics and Craniofacial Research 10 (3): 129-37. (4) Boyd, Donna 1996. Skeletal Correlates of Human Behavior in the Americas. Journal of Archaeological Method and Theory 3 (3): 189251. (5) Burnston, Sharon Ann, and Ronald A. Thomas 1981. Archaeological Data Recovery at Catoctin Furnace Cemetery, Frederick County,

hurj fall 2016: issue 22 Maryland. Prepared for Orr and Son, Consulting Archaeologists. (6) Burnston, Sharon Ann, and Ronald A. Thomas 1981. Archaeological Data Recovery at Catoctin Furnace Cemetery, Frederick County, Maryland: Appendices. Prepared for Orr and Son, Consulting Archaeologists. (7) Coleman, E. N., and J. B. Foote 1954. Craniostenosis with Familial Vitamin-D-resistant Rickets. British Medical Journal 1 (4861): 561-62. (8) Fennell, Christopher C. 2011. Early African America: Archaeological Studies of Significance and Diversity. Journal of Archaeological Research 19 (1): 1-49. (9) Gordon, H. 1959. Craniostenosis. British Medical Journal 2 (5155): 792-95. (10) Harrod, Ryan P., and Debra L. Martin 2015. Bioarchaeological Case Studies of Slavery, Captivity, and Other Forms of Exploitation. In Archaeology of Slavery: Comparative Approaches to Captivity and Coercion, edited by Lydia W. Marshall, pp. 41-63. Southern Illinois University, Carbondale. (11) Heuman, Gad J., and James Walvin 2003. The Slavery Reader. Routledge, London. (12) Kelley, Jennifer O., and J. Lawrence Angel 1983. Workers of Catoctin Furnace. Maryland Archeology 19 (1): 2-17. (13) Johnson, David, and Andrew O. M. Wilkie 2011. Craniosynostosis. European Journal of Human Genetics 19 (4): 369-376. (14) Mack, Mark and Michael Blakey 2004. The New York African Burial Ground Project: Past Biases, Current Dilemmas and Future Research Opportunities. Historical Archaeology 38(1): 10-17. (15) McGillivray, G., R. Savarirayan, T. Cox, C. Stojkoski, R. McNeil, A. Bankier, J. Bateman, T. Roscioli, R. Gardener, and S. LamandĂŠ 2005. Familial Scaphocephaly Syndrome Caused by a Novel Mutation in the FGFR2 Tyrosine Kinase Domain. Journal of Medical Genetics 42 8: 656-662. (16) Muenke, Maximillian 2011. Genetics and Genomics of Craniosynostosis Syndromes. Monographs in Human Genetics Vol. 19, edited by M. Muenke, W. Kress, H. Collmann, and B. D. Solomon. Karger, Basel, New York. (17) Panigrahi, Inusha 2011. Craniosynostosis Genetics: The Mystery Unfolds. Indian Journal of Human Genetics 17 (2): 48-53. Smithsonian Institution n.d. Data files, National Museum of Natural History, Department of Anthropology, Washington, D.C.

humanities and social science the health trends of iron furnace slaves.


Adrienne Kramer is a senior at Hopkins majoring in Archaeology with a minor in Psychology. She is from Hagerstown, Maryland. Adrienne plans on applying to dental school next year and is interested in forensic odontology. She is also completing an honors thesis on


humanities and social sciences

hurj fall 2016: issue 22

Perceptual (A)symmetry: Visual Search and Change Detection for Basic Image Types J.J. Valenti, Class of 2018 Cognitive Science

Abstract The present study explores the similarity between pairs of images and the effect such similarity has on a change-detection task. Materials were borrowed from previous research on visual similarity and a flicker paradigm was used for the change-detection task. Subjects had to detect a change between two rapidly flickering arrays of 16 stimulus objects as quickly as possible. Additionally, subjects were asked to indicate the location of change. Accuracy in change detection was strongly correlated with reaction time, suggesting that the use of less similar pairs made the task easier. Further analysis was performed on possible asymmetry between judgments of similarity. Results indicate that many of the test stimuli pairs possess robust asymmetries in judgments of similarity.

Introduction Human vision is filled with paradoxes and often times it violates our basic intuitions about how “seeing things” works. How do we manage to represent complex scenes so vividly in the brain, and yet reliably fail the most banal change-blindness task (Rensink et al., 1997)? Perhaps the hardest intuition to shake is the notion that vision operates like a projector and a screen, such that one’s eyes capture the outside world and project the image to the brain; the brain somehow manages to “see” this image and make sense of it. Scientific psychology has long rejected this conception of vision, and replaced it with an empirically informed view, which holds that vision (along with the other modalities of perception) is a process of representation; certain brain mechanisms take information captured on the retina and represent this information though a means of neural computation. Representations become more refined as they continue down the visual stream (Rensink et al. 1997; Mohan & Arun, 2012). Two challenges encountered by the human visual system are representing small changes in the scene (change detection) and localizing objects in the visual field (visual search). Previous research indicates that visual search performance is strongly affected by the familiarity of stimuli (Wang et al., 1994), the presence of distractor objects (Shen & Reingold, 2001), and the angle at which stimuli are viewed (Von Grünau & Dubé, 1994). Research on change detection shows that we


become utterly oblivious to changes under certain circumstances, but we are acutely aware of other kinds of changes to the scene. Our capacity to detect changes is rather fickle; with the mere addition of a blank screen between two changing arrays, change detection becomes very difficult (Rensink et al., 1997). The literature on change detection suggests that our visual representations of the scene around us are rather fluid. In the present study, we focus on another aspect of visual cognition, perceptual similarity, and its effects on both change detection and visual search. Perceptual similarity is the judged similarity between two presented objects or images. Perceptual similarity can be estimated graphically by measuring the pixel-by-pixel differences between two images (Mohan & Arun, 2012) or it can be estimated experimentally by testing subjects’ own judgments (Wolfe, 2001). Categorization of visual stimuli—a task that required the ability to discern changes—also relies heavily on quick judgments of perceptual similarity (Mohan & Arun, 2012). Perceptual similarity, like other domains of vision, is plagued by unique paradoxes. Chief among them is a phenomenon known as perceptual asymmetry. Asymmetry arises when image A is judged to be very similar to image B, but B is not judged to be similar to A. Attempts to formalize the underpinnings of these judgments through a pixel-by-pixel analysis of the relevant images have proved fruitless (Wolfe, 2001). In the present study, we borrow image-pair stimuli from Wolfe (2001) and Wang et al. (1994) (See Figure 1) and incorporate those stimuli into a change detection task that utilizes a flicker paradigm. We predict that accuracy on this change detection task will be negatively correlated with reaction time, suggesting that less similar image-pairs will elicit easier detection of change. An additional goal of the present study is to reveal where true asymmetries exist by using the change detection data. We hypothesize that image-pairs demonstrating strong differentials in change detection reaction time possess inherent perceptual asymmetries.

Materials and Methods Participants We tested 10 undergraduates from The Johns Hopkins University. All participants had normal or corrected-to-normal acuity and were awarded class credit for certain psychology class requirements.

hurj fall 2016: issue 22

humanities and social sciences

Apparatus and Stimuli The experiment was run on an iMac computer using custom software written in Matlab Psychotoolbox-3C. Participants sat approximately 50 cm from the screen. The computer’s frame rate was irrelevant in the present experiment because the stimuli did not involve motion on the screen. Head restraints were not used on participants during the experiment.

Procedure Subjects entered the lab and participated in our 2-session experiment, with each session interrupted by a five-minute rest period. Each session consisted of 10 randomized blocks. Each block had 48 trials, with each image-pair repeated 4 times, for a grand total of 960 trials over the course of the entire experiment. Subjects completed a short 8 trial demonstration to familiarize themselves with the structure of the experiment before beginning the formal trials. Within the image-pairs, the image that served as the “distractor” and the image that served as the “target” were counterbalanced across trials. After each block, subjects were informed of their cumulative accuracy for change detection on the monitor. Image-pair stimuli were selected from previous research in the similarity literature (Wolfe, 2001; Wang et al., 1994) (See Figure 1).

Figure 1. Image-pair stimuli, borrowed from Wolfe (2001) and Wang et al. (1994). Most image-pairs resemble orthographic symbols in English. Some pairs are reflections of one-another, others add or remove a feature. Some pairs represent geometric shapes (8 and 12 for example). Finally, one image-pair represents a schematic elephant, oriented both normally and upside down.

Each trial (See Figure 2) utilized a flicker paradigm in which an array of 16 images was presented on the monitor for 200 ms, followed by a “blank” screen that had black fixation “+” located at the center of the monitor for 150 ms. Next, an array of 16 images was presented on the monitor for 200 ms, followed by another “blank” screen that included the black fixation “+” fixation for 150 ms. This cycle repeated itself 6 times on each trial.

Figure 2. Schematic diagram of a single trial. The first array is shown for 200 ms followed by a 150 ms blank. The array is shown again for 200 ms, either with a change to one of the images (“present” condition), or with no change (“absent” condition), followed by another 150 ms blank. This cycle repeats 6 times, completing the Flicker Paradigm.

Each trial corresponds to one of two conditions, which were counterbalanced across the experiment. In the “present” condition, a change between the two arrays of 16 images was present, meaning that the first array contained 15 of one image and 1 of the other, while the second array contained 16 of the first image. In the “absent” condition, there was no change between the two arrays of 16 images, meaning that both arrays contained 16 of the same image. The precise location of the changed stimulus on the array was randomized. Subjects were given a specified set of directions prior to completing 8 familiarization trials. Subjects were informed that their task during the study was to determine whether or not there was a change between arrays (change detection task) and to indicate the approximate location of the changing stimulus—left side or right side of the monitor (visual search). Subjects had to decide whether or not a change was present within 6 cycles of the flicker paradigm, which lasts approximately 4 seconds. Subjects were instructed to press the [C] key on the keyboard when a change was present and the [N] key when a change was absent. Whenever a change was present—meaning subjects had entered the [C] key— they were then instructed to press the LEFT arrow key if the change was present on the left two columns of the array, or the RIGHT arrow key if the change was present on the right two columns of the array.

Results Prior to the collection of any data, a criterion of accuracy was set, such that subjects with a cumulative accuracy on the change detection task lower than 80% were automatically excluded from any further statistical analyses. The preliminary data comport with both the change detection and visual search literature. Collapsing sub-


humanities and social sciences jects and image-pairs estimates an overall measure of accuracy on the change detection task. Mean accuracy in the “present” condition (M= .84, SD= .21) was significantly poorer than the mean accuracy in the “absent” condition (M= .97, SD= .048) (t(478) = -9.35, p < .001). This measure suggests that one can more accurately—or more easily—determine that there is no change present than one can determine that there is a change. A similar measure for reaction time (RT) is ascertained by collapsing across subjects and image-pairs. Mean RT, in ms, in the “present” condition (M= 1538.30, SD= 461.31) was significantly faster than the mean RT in the “absent” condition (M= 2154.80, SD= 775.24) (t(477) = -10.23, p < .001). This measure suggests that subjects do indeed wait longer to choose “no change” than when choosing “change,” a finding first described by Wolfe (2001). Next, we performed a correlation analysis using image-pairs as individual data-points, comparing change detection accuracy to RT. For each condition (“present” or “absent”) a separate correlation analysis was performed. For the “present” condition (See Figure 3), our analysis found a strong negative correlation between accuracy and RT (Pearson’s r = -.946). These data suggest that subjects were able to distinguish certain image-pairs from another more easily and rapidly. For the “absent” condition (See Figure 3) however, our analysis found a more moderate correlation (Pearson’s r = -.617). It should be noted that the “absent” condition required no direct comparison of images within the pair. We suspect this reason explains why accuracies for all image-pairs were quite high (> .90).

hurj fall 2016: issue 22 plotted on the Y-axis. The plot shows a strong negative correlation between accuracy and RT (Pearson’s r = -.946). Bottom, Correlation Plot for the “absent” condition. Accuracy in this condition was high among all image-pairs (>.90). In this condition, subjects only see one image from the pair. The plot shows a moderate correlation (Pearson’s r = -.617).

Next we performed a set of post-hoc t-tests, comparing pairwise similarity for the 12 image-pairs. Each image-pair has two separate images and for any given trial, one image may serve as the “distractor” and the other may serve as the “target.” These tests compare image-pairs on this dimension. For this analysis we exclusively compared data from the “present” condition; it is in this condition where subjects have to do the rapid computations responsible for discriminating similar image. For this reason, RT data serves as the best measure for asymmetries in visual search (See Mohan & Arun, 2012). Still, we performed tests on dimensions of RT (in ms), accuracy, and the number of frames seen before entering the first keystroke (See Table 1). Pairwise comparisons for RT (See Figure 4) reveal which image-pairs possess visual search asymmetries. Post-hoc t-tests demonstrated that image-pairs One, Two, Four, and Five possess very strong asymmetry effects (p < .001). Image-pairs Six, Seven, Eight, Eleven and Twelve possess moderately strong asymmetry effects (p < .05). Finally, image-pairs Three, Nine, and Ten show no effects of asymmetry (p > .05). Refer to Figure 1 for visuals of the image-pairs used.

Figure 4. Pairwise RT comparisons for the “present” condition. Black bars refer to trials where the first image in the pair served as the target image; grey bars refer to trials where the second image in the pair served as the target image. Large differences in RT suggest visual search asymmetries.


Figure 3. Top, Correlation Plot for the “present” condition. Each point represents an image-pair and target/distractor status. Accuracy in the change detection task is plotted on the X-axis and RT is


In the present study, we investigated how the use of perceptually similar visual stimuli hinders performance in a change detection task. Our investigation also revealed which image-pairs may possess inherent perceptual asymmetries. Our primary analysis shows that in the “present” condition, subjects performed poorer than when in the “absent” condition. This finding is well document in the visual search literature, and suggests that subjects are often blind to (and do not attend to)

hurj fall 2016: issue 22 relatively stark changes in the visual field. We also found that in the same change detection task, subject RTs were substantially higher in the “absent” condition than in the “present” condition. Humans seem to be hesitant to affirm that their environment has not changed; they are always open to the possibility of a stark but unnoticed change. This finding has had some prior discussion (Wolfe, 2001). However, we believe that this finding has a straightforward explanation. Essentially, we believe that when subjects in the “present” condition encounter a change between the two arrays, they immediate press the approximate key on the keyboard; in these cases, subjects do not double check to confirm what they saw. In the “absent” condition however, subjects are much more hesitant to press the appropriate key without substantial confirming evidence. We propose a follow up experiment testing precisely this; we hypothesize that subjects will continue to scan the two arrays more robustly in the “absent” condition than in the “present” condition. Eye trackers will be integral to this proposed experiment. The results from our correlation analysis support our primary hypothesis that image-pairs that are intuitively judged as being less similar will be easier to discriminate from one another and therefore, lead to higher accuracy and lower RTs in a change detection task. Our correlation analysis revealed a strong negative correlation between accuracy and RTs, suggesting that not only do less similar image-pairs make the change detection task easier for subjects, but subjects actually pick-up on changes more quickly when image-pairs possess inherent asymmetries.

humanities and social sciences (4) Rensink, R. A., O’Regan, J. K., & Clark, J. J. (1997). To see or not to see: The need for attention to perceive changes in scenes. Psychological science, 8(5), 368-373. (5) Shen, J., & Reingold, E. M. (2001). Visual search asymmetry: The influence of stimulus familiarity and low-level features. Perception & Psychophysics, 63(3), 464-475. (6) Von Grünau, M., & Dubé, S. (1994). Visual search asymmetry for viewing direction. Perception & Psychophysics, 56(2), 211-220. (7) Wang, Q., Cavanagh, P., & Green, M. (1994). Familiarity and popout in visual search. Perception & psychophysics, 56(5), 495-500. (8) Wolfe, J. M. (2001). Asymmetries in visual search: An introduction. Perception & Psychophysics, 63(3), 381-389.

Author J.J. Valenti is a junior at Johns Hopkins studying Cognitive Science and Philosophy. He is originally from Chicago. His research interests lies at the intersection of human/primate social cognition and the evolution of mating and cooperation behaviors. Valenti plans on pursuing a PhD in the brain sciences.

More work in this domain surely needs to be conducted. Studying the nature and intricacies of perceptual judgements have tangible real world benefits. Orthographic processing disorders such as dyslexia are essentially deficits in the ability to compare and discriminate similarity in letters and figures. We hope this work will contribute someday in determining the exact causes of such orthographic processing disorders. Additionally we believe that the human mind’s ability to make snap judgments for perceptual similarity plays an important role in driving safely whilst avoiding pedestrians and attending to street signs, making medical diagnoses before and during surgical procedures, and adjudicating implicit racial bias in situations are racial categorization. The more we understand the nature of perceptual similarity, the more able we are to employ it when perceptual judgments are beneficial and combat it when perceptual judgments are harmful.

References (1) Brainard, D. H. (1997). The psychophysics toolbox. Spatial vision, 10, 433-436. (2) Lisi, M., & Cavanagh, P. (2015). Dissociation between the perceptual and saccadic localization of moving objects. Current Biology, 25(19), 2535-2540. (3) Mohan, K., & Arun, S. P. (2012). Similarity relations in visual search predict rapid visual categorization. Journal of vision, 12(11), 19.


humanities and social sciences

hurj fall 2016: issue 22

How Maternal Substance Abuse Affects the Physical Health of Children Milena Berhane, Class of 2019 Public Health Studies

Abstract An estimated 7.93 million women in the United States (6.5% of the female population) regularly use illicit drugs (Substance Abuse and Mental Health Services Administration, 2003). Maternal drug abuse affects not only the mother herself, but also affects her children. Substance abuse can affect not only the psychological upbringing of the child but also the physical health and well-being of the child. Although there have been multiple studies focusing on the psychological and mental health of the children, little has been studied about the physical health of children of substance using parents. The current study focuses on how children of substance using mothers report their physical health. Findings are reported from a sample collected in the Midwest comprised of mothers and their children which vary in age from 8-16 years old. When asked to report their physical health, the majority of the children did not believe that they physical health would decline severely or their health was worse than other children. This study could motivate more studies to explore how the physical health of the child is affected by their mother’s substance abuse.

Introduction An estimated 7.93 million women in the United States (6.5% of the female population) regularly use illicit drugs (Substance Abuse and Mental Health Services Administration, 2003). Maternal substance abuse is also the most common factor of children being referred to Child Protective Services due to suspected neglect and/or parental abuse (Child Welfare League of America, 1998; Department of Health and Human Services [DHHS], 1999). Substance abuse is associated with low self-esteem, peer pressure and depression (National Center on Addiction and Substance Abuse at Columbia University, 2003). Drug and alcohol use is considered substance abuse when it leads to a clinically significant impairment or distress. Substance abuse not only affects the psychological and physical health of the mothers but also the physical health and development of their children. Children of substance abusing women have more chances of developing psychological disorders and physical issues such as asthma. Maternal drug abuse poses multiple risks for the children’s health and well-being. In the current study, the children’s self-report of their physical health is described. Previous studies have looked into parental substance


use and how it causes maladjustment of the child (Suchman, 2000). It has also been shown that parental warmth and communication lead to better child behaviors and overall upbringing (Suchman, 2007). Other studies have shown how negative maternal parenting behaviors cause significant vulnerability in children (Luthar & Sexton, 2007). Among children of substance-abusing mothers, a lack of parental warmth and poor parenting behaviors have shown to be associated with symptoms of depression, stress, and other psychological disorders (Suchman, Pajulo, DeCoste, Mayes, 2006; Suchman, 2007; Suchman, 2000). Although it has been established that maternal substance abuse has been linked to insufficient parenting and potential psychological issues in the child, it is not yet understood how the child’s physical health is affected.

Participants Participants included N=183 substance using mothers with at least one mother in their care. These mothers were recruited from a community center in a large Midwestern city. The eligibility criteria for mothers were that they had to 1) be seeking outpatient treatment for their substance use disorder, 2) meet diagnostic criteria for an alcohol or drug use disorder as defined by DSM IV, and 3) have a child between the age of 8-16 years who either resided with the participating mothers at least 50% of the time in the past two years or 100% of the time in the past 6 months, ensuring that the child has been in the care of the mother long enough to have been exposed to the mothers substance abuse. This substance abuse involves alcohol or marijuana, if not both. If more than one child was eligible, the child reporting substance use or other problem behaviors was selected as the target child. The mothers ranged in age from 22 to 54 years old. They were primarily white, non-Hispanic (53.6%), or African-American (42.6%). Majority of the mothers (85%) had a high school degree or less and only 19.1% of the women reported that they were married. In regards to income, 60% of the families had an annual income of $15,000 or below and only about 18% of families had an annual income of $30,000 or greater. Mothers reported having between 1 to 11 children, with the target child’s ages ranging from 8 to 16. Out of the target children, 51.9% of them were male.

hurj fall 2016: issue 22

humanities and social sciences

Procedure The community treatment center was used to screen the mother for eligibility and interest in the research. After parental permission was obtained from the eligible mothers, research assistants contacted the target child and informed him or her on the research study. At baseline, both the mother and target child completed an assessment battery including individual and family measures, as semi-structured substance use assessment. Assessment procedures for all research participants were identical.

Measures The child’s physical health will be measured using the Short-Form-36, which is a multi-purpose health survey (Ware, et al., 1993). This form was derived from the Rand Corporation’s Medical Outcome Study (MOS) (Ware & Sherbourne, 1992). It can be used as a generic assessment of the child’s health status and an outcome measure in clinical practice. The survey includes 36 questions that yield 8 subscales assessing physical health (physical functioning, role-physical, bodily pain, and general health) and mental health (vitality, social functioning, role-emotional, and mental health). For the current study, the physical health portion of the survey will be used. The questions are to be answered true/false, with a numeral system 1-5. An answer of 1 indicates “definitely true”, 2 is “mostly true”, 3 is “don’t know”, 4 is “mostly false”, and 5 is “definitely false”. The question regarding general health is answered on scale of 1 to 5, where 1 indicates “excellent”, 2 is “very good”, 3 is “good”, 4 is “fair”, and 5 is “poor”. Reliability estimated ranged from .65 to 0.94 across scales (median = 0.85) and varied across patient subgroups (McHorney, et al., 1994). SF-36 scales achieved about 80-90% of their empirical validity in studies involving physical and mental health criteria (McHorney et al., 1993) and showed high validity and reliability in psychiatric assessments of mental health patients (Ware et al, 1993). It has been shown that the SF-36 is suitable for self-administration, computerized administration, or administration by a trained interviewer in person or by telephone (Ware et al., 1993).

ple”, majority of children responded false. This indicates that most children did not believe that they were more prone to sickness than any other person. For the statement “I am as healthy as anyone else”, the majority of children also responded false. This response shows that most of children thought that their health is not satisfactory compared to others. For the statement “I expect my health to get worse”, the majority of the children responded false, which indicated that they did not believe that their health was at risk of getting any worse. When asked if the children thought “my health is excellent”, the majority of the children reported false as well. This response shows that the children do not think that their health is as strong as it could be, which could be for a number of reasons. Refer to Tables 1-4 for percentages. When assessing the general health of the children, 182 out of the 183 samples were valid. Most of the children reported that their general health was “good”, reporting a score of 3. This response is also another indication of the children thinking that their health is not the best that it could be. Refer to Table 5 percentages.

The findings showed that children of substance using mothers rated their physical health positively.

Baseline Demographic Characteristics The ethnicities of both the mothers and children were measured. The ages of the children varied from 8-16 years of age.

Results The health information of the children was obtained using the SF-36 self-assessment form. For the statements “I seem to get sick a little easier than other peo-

In summary, the findings showed that children of substance using mothers rated their physical health positively. Although the majority of children did not say that their general health was poor or that they believed that their health was significantly worse than others, most children believed that their health was not excellent.

Discussion Maternal drug abuse can pose a number of risks for the children under the mother’s care. Much research has shown that maternal substance abuse affects the psychological health and well-being of the children (Suchman, 2000). Maternal substance abuse has been linked to poor parenting behaviors, which causes maladjustment of the child (Suchman, 2000). Among children of substance-abusing mothers, a lack of parental warmth and poor parenting behaviors have shown to be associated with symptoms of depression, stress, and other psychological disorders (Suchman, Pajulo, DeCoste, Mayes, 2006; Suchman, 2007; Suchman, 2000). How maternal substance abuse affects the physical health of the child has yet to be thoroughly investigated. Most of the findings did not support the hypothesis of this study. Although the majority of the children did not believe that their health was “excellent”, most believed that their health was “good”. An outlier in the results was that when asked whether the children were as healthy as anyone they knew, the majority replied false. This response shows that many children believe that


humanities and social sciences

hurj fall 2016: issue 22

their health is good enough for them personally, but do not believe that they are as healthy as others. Future research might determine what health disparities can come about from maternal substance abuse. With the utilization of medical records and a self-assessment that is focused on physical health, more information can be obtained to do an in-depth study on the physical health of children under the care of substance abusing mothers.

Limitations Limitations of this study include the sample of convenience recruited from one substance use treatment facility in the Midwest. This sample was only limited to mothers and their children from the Midwest, which is only one specific population. Accuracy of the physical health of the child is also limited because of the self-assessment system of measurement. Because the SF-36 form is not focused to assess physical health only, the results may not be an accurate report of the child’s overall physical health.

Implications This study could induce other studies to further explore the affects of maternal substance abuse on the child’s physical health. Future studies could utilize medical records and a self-assessment survey that focuses solely on physical health in order to get specific information on health issues such as asthma and others.



This research was supported by NIDA Grant R01 DA023062 to Dr. Natasha Slesnick. Thanks to Dr. Slesnick and Brittany Brakenhoff for the assistance and support throughout the duration of this project.

hurj fall 2016: issue 22 References (1) Achenbach, T.M., & Edelbrock, C. S. (1982). Manual for the Child Behavior Checklist and Child Behavior Profile. Burlington, VT; Child Psychiatry, University of Vermont (2) Black MM, Nair P, Kight C, Wachtel R, Roby P, Schuler M. Parenting and early development among children of drug-abusing women: Effects of home intervention. Pediatrics. 1994; 94:440–448. (3) Chaplan, T. M., & Aldao, A. (2013). Gender differences in emotion expression in children: A meta-analytic review. Psychological bulletin, 139(4), 735-765. (4) Child Welfare League of America. Alcohol and other drug survey of state child welfare agencies. Washington, DC: Author; 1998. (5) Department of Health and Human Services. Blending perspectives and building common ground: A report to congress on substance abuse and child protection. Washington, DC: U.S. Government Printing Office; 1999. (6) DiClemente, C.C. (1999). Motivation for change: Implications for substance abuse treatment. Psychological Science, 10(3), 209. (7) Enlow, M.B., Englund, M. M., & Egeland, B. (2016). Maternal childhood maltreatment history and child mental health: Mechanisms in intergenerational effects . Journal of Clinical Child & Adolescent Psychology: Online First: DOI: 10.1080/15374416.2016.114189 (8) Isenhart, C. E. (1994). Motivational subtypes in an inpatient sample of substance abusers. Addictive Behaviors, 19, 463–475. (9) Luthar, S. S., & Sexton, C. C. (2007). Maternal drug abuse versus maternal depression: Vulnerability and resilience among school-age and adolescent offspring. Development and Psychopathology, 19(1), 205–225. (10) McHorney, C. A., Ware, J. E. & Raczek, A. E. (1993). The MOS 36-Item Short-Form Health Survey (SF-36®): II. psychometric and clinical tests of validity in measuring physical and mental health constructs. Medical Care, 31(3), 247-63 (11) McHorney, C. A., Ware, J. E., Lu, J. F. R. & Sherbourne, C. D. (1994). The MOS 36-Item Short-Form Health Survey (SF-36®): III. tests of data quality, scaling assumptions and reliability across diverse patient groups. Medical Care, 32(4), 40-66. (11) Miller, W. R., & Marlatt, G. A. (1984). Manual for the Comprehensive Drinker Profile. Odessa, FL: Psychological Assessment Resources. (12) Miller, W.R. & Tonigan, J.S. (1996). Assessing drinkers’ motivation for change: The stages of Change Readiness and Treatment Eagerness Scale (SOCRATES). Psychology of Addictive Behaviors, 10, 8189. (13) National Center on Addiction and Substance Abuse at Columbia University, The Formative Years: Pathways to Substance Abuse Among Girls and Young Women Ages 8–22 (PDF), 2003 (14) Substance Abuse and Maternal and Child Health. (2003.). https:// (15) Suchman, N. E., & Luthar, S. S. (2000). Maternal addiction, child maladjustment and socio-demographic risks: implications for parenting behaviors. Addiction (Abingdon, England), 95(9), 1417–1428. (16) Suchman, N., Pajulo, M., DeCoste, C., & Mayes, L. (2006). Parenting Interventions for Drug-Dependent Mothers and Their Young Children: The Case for an Attachment-Based Approach. Family Relations, 55(2), 211–226. (17) Suchman, N., Rounsaville, B., DeCoste, C., & Luthar, S. (2007). Parental control, parental warmth, and psychosocial adjustment in a sample of substance-abusing mothers and their school-aged and adolescent children. (18) Tonigan, J., Miller, W.R., & Brown, J.M. (1997). The reliability of Form 90: An instrument for assessing alcohol treatment outcome. Journal of Studies on Alcohol, 58 (4), 358-364. (19) Ware J. J. & Sherbourne, C.D. (1992). The MOS 36-item shortform health survey (SF-36). I. Conceptual framework and item selection. Medical Care, 30, 473-83. (20) Ware, J. E., Snow, K. K., Kosinski, M., & Gandek, B. (1993). SF-36® Health Survey Manual and Interpretation Guide. Boston, MA: New England Medical Center, The Health Institute. (21) Westerberg, V.S., Tonigan, J.S., & Miller, W.R. (1998). Reliability of Form 90D: An instrument for quantifying drug use. Substance Use, 19, 179-189.

humanities and social sciences Author

Milena Berhane, Class of 2019. Major: Public Health Studies, Minor: Africana Studies. Hometown: Laurel, MD. Post-graduation aspirations/plans: Milena plans to attend medical school after graduating from Hopkins. She aspires to specialize in pediatric internal medicine, and hopes to serve in developing countries after medical school. Extracurriculars: Refugee Youth Project Vice President, HEES Public Relations Chair, Alternative Break Leader.


humanities and social sciences

hurj fall 2016: issue 22

Altera - Heart rate variability analysis software for small animal models Paul Kim Class of 2018, Biomedical Engineering

Abstract Altera is an advanced and easy-to-use program for heart rate variability (HRV) analysis. It is a MATLAB-based tool that supports several types of input data formats for electrocardiogram (ECG) data. Key features of the program are the R-peak detection, artifact correction via manual user input, and dynamic presets for different types of subjects. Compared to other programs that compute human HRV data, Altera can be effective when analyzing small animal data such as rats and mice. The program computes time-domain and frequency-domain HRV parameters that are commonly used today. Users can modify and adjust settings to optimize the analysis experience for several types of data. The program was developed for ease of use, with a simple graphical user interface built with MATLAB GUIDE, an interactive development environment. Introduction Heart rate variability (HRV) is the variation in beatto-beat intervals. Along with blood pressure (BP), HRV is increasingly acknowledged as an evaluative tool for autonomic nervous system (ANS) function and cardiovascular state. For example, a study has shown that a survivor of myocardial infarction lacks variability in HRV and has higher chance of fatal heart attacks [1]. Overall, this variance demonstrates the heart’s ability to adapt to different environments and states, as well as the balance between the different branches that compose the ANS. HRV research has been gradually gaining momentum due to increasing interest in the field alongside relevant technological advances. The significance of the HRV was discovered as early as the 1960s, in a study linking fetal health to fetal HRV. Yet another study from 1978 established a connection between high HRV and low mortality and myocardial infarction rates.3 In 1981, it was found that the sympathetic and parasympathetic nerves of the ANS affect the heart rate (HR) through pharmacological interventions [4]. The introduction of computer technologies during the 1980s gave way for new signal processing techniques for electrocardiogram (ECG) recordings and analysis algorithms, further expanding research in the field. Today, HRV parameters used in conjunction with the more established measure of HR for evaluative purposes. The following table addresses the general timeline of HRV research.


Table 1: Timeline of HRV research

There are several methods and algorithms that are used in HRV analysis. Common practice is to obtain HR and RR interval series by processing ECG signals. There are two main types of analysis: time domain analysis and frequency domain analysis. Time domain measures involve the analysis on the time-series of RR intervals. Some common parameters include the root-mean-squared successive differences (RMSSD), which calculates the square root of the mean of squared differences between successive RR intervals, and pNNx, a measure of the percentage of consecutive RR intervals that differ by more than an ‘x’ number of milliseconds. Both of these indices measure short-term variation in the NN interval because they are entirely based on comparisons between successive beats [5]. Frequency domain methods mostly consist of analyzing the power spectral density (PSD) of the RR intervals. The ANS is further subdivided into the sympathetic nervous system (SNS) and the parasympathetic nervous system (PNS), which are responsible for activating body processes for action and rest, respectively. Since the levels of PNS and SNS

hurj fall 2016: issue 22

humanities and social sciences

activity affect the frequency patterns of the heart, using PSD analysis techniques allows for the decomposition of the signal into different frequency components that correspond to each system [4]. In this paper, we present Altera HRV (ver. 1.0), an analysis tool that provides time and frequency domain analysis functions. In contrast to most HRV analysis tools currently available, Altera gives users the ability to save sets of manually processed data and employs algorithms optimized for smaller animal models (e.g. rats, mice).

Computational Methods Here, we elaborate on the algorithms and HRV parameters included in Altera. As of version 1.0, the program is capable of analyzing most time and frequency domain parameters. The user specifies a time window for parameter analysis, and the program subsequently runs the analysis algorithms for the given time period.

Preprocessing After the user inputs the sampling frequency of the given data set, Altera begins by automatically extracting the location of the R-peaks from the ECG recording through our custom QRS detection algorithm, developed based on the Pan-Tompkins algorithm. The data is normalized by subtracting the mean and then dividing it by the absolute maximum value. This is then run through a second order Butterworth filter (2-300 Hz) to reduce power line noise, baseline wander, and other noise components. The frequency response of the Butterworth filter is maximally flat in the passband and rolls off towards zero in the stopband [11]. This makes the phase response close to linear range in the frequency range of interest. In other words, the smoothness of the filtered signals comes at the price of decreased roll-off steepness (transition between passband and stopband). The transfer function coefficients of the filter, returned as row vectors of length n + 1 for low pass and high pass filters and 2n + 1 for bandpass and band stop filters.

where a and b are transfer function coefficients and n is the order of the filter.

Figure 1. Frequency response of an example bandpass filter [12]

In addition, to ensure that lower frequency components are passed under cutoff frequency of interest, we implement a low pass filter. Low pass filter does successful job in attenuating frequencies higher than desired low frequency. The initial filtered signals are convoluted with the low pass filter after every process in the algorithm. The input-output description of the filter operation on a vector in the Z-transform domain is a rational transfer function. A rational transfer function is of the form,

where a and b are the denominator and numerator coefficients, respectively. These convoluted samples are then squared to highlight the high value peaks. The squared samples are convoluted again with the low pass filter and go through a thresholding process. This is done by accepting the R-peak values that are as big as 40% of the maximum peak. Thresholding is necessary to delete possible outliers in the samples and discard any random peaks other than the R-peak from the QRS complex. Finally, a moving average filter is used to clear out the peaks that are close together.

where x is each sample and n represents the indices.


humanities and social sciences After the R-peaks are detected, they are interpolated to improve the time resolution of the detection. The code for this preprocessing is fully dynamic and can be easily modified by users with sparse knowledge of MATLAB. The final result of the preprocessing process is then stored in an array for analysis.

hurj fall 2016: issue 22 spectrum estimate is obtained by averaging the FFT spectra of these windowed segments. This method helps find and test weak periodic signals in otherwise random, unevenly sampled data.

Parameter Analysis: Time Domain After a processed RR series is extracted from the ECG reading, the program can perform time domain analysis techniques. The time domain methods are computationally simple and they are calculated directly from the RR series. The most obvious and easy to extract measures are the mean RR and HR interval values. Another measure is the standard deviation of normal-to-normal RR intervals (SDNN), which reflects the overall (short-term and long-term) variations in HRV. Similarly, the root mean square of successive differences (RMSSD) reflects short-term HRV. Another measure is NNx/pNNx which are the number of successive intervals differing by more than ‘x’ milliseconds and the percentage of successive intervals. The two main time domain graphs displayed in Altera are a histogram and a pNNx graph. The histogram displays the RR interval distribution, with counts for every 0.05 second interval window. The pNNx graph plots pNNx against the NN interval value (x).

Figure 2. Altera Sample Parameter Analysis Figures (Time and Frequency Domains)

Parameter Analysis: Frequency Domain Altera also analyzes data in frequency domain. It uses MATLAB built-in Lomb-Scargle method to calculate the power spectrum of the given samples. Using Lomb-Scargle periodogram, the RR series is divided into overlapping segments, each of which is windowed to decrease the leakage effect, and the


The spectrum estimates are then divided into very low frequency (VLF), low frequency (LF), and high frequency (HF) bands. Our preset for rat models use 0–0.195 Hz (VLF), 0.195–0.6 Hz (LF) and 0.6–2.5 (HF). HRV measures extracted from these frequency bands include peak frequencies (i.e. the frequency values corresponding to maximum power within VLF, LF, and HF bands), absolute and relative powers (for VLF, LF, and HF), normalized powers (for LF and HF), LF/HF power ratio, and the total spectral power. The band powers are computed by simply integrating the estimated power spectrum through MATLAB’s built-in integration

hurj fall 2016: issue 22

humanities and social sciences

function over different frequency band limits.

pear on the graph section so that users can clearly see which time window they are working on. Finally, for the results window, users are provided with all the statistical values calculated from the algorithm in their desired time window. Users can switch between time domain method and frequency domain window to record analyzed values or powers.

Table 2. Frequency classification and primary mechanism

Figure 3: Basic UI of Altera

Table 3: Common parameters for HRV analysis

Software Description Input Data Format As of version 1.0, raw ECG readings only in ‘.mat’ format are accepted. The user is prompted to indicate the sampling frequency of the reading at the start of the program. The program then automatically detects the peaks and generates the RR interval series in accordance to input data.

Save and Load Options

Altera provides session saving and loading functions that facilitate continuing post-processing work.

Due to computer memory usage issues stemmed from MATLAB and the length of some ECG recordings, some long data sets may be impossible to analyze beyond peak extraction and RR interval series generation. VLF, LF, and HF Altera is used by way of a graphic user interface (GUI). Ver. 1.0 was built with MATLAB’s GUI development tool and application compiler. The GUI is divided into three different sections: the RR interval series section, analysis options section, and the results section. For graph section, the program generates an ECG graph versus time with automatically detected R-peaks. Users can manually add R-peaks if the algorithm missed the peak due to noise or artifacts. There’s an additional graph with RR interval, which allows user to detect any errors in data due to noise or artifacts. For time input section, depending on the length of the data, user can choose the desired time window to analyze. When this is done, a rectangular shade will ap-

Altera provides session saving and loading functions that facilitate continuing post-processing work (i.e. adding undetected R-peaks) at later points in time. Saving a session will store all data excluding analyzed HRV parameters. This is primarily useful for storing arrays of peaks when working with arrhythmia-heavy datasets.

Sample Run In order to test whether the algorithms implemented in Altera were accurate and working as intended, we compared analysis results for three synthetic data sets (increasing frequency, decreasing frequency, and alternating patterned frequency) as well as one real, experimental data set. The synthetic data sets were created with different combination of sine waves with fundamental frequency of 6 Hz. This was to account for the bpm of rodent models whose bpm is around 360. The algorithm for generating the synthetic data sets was based on ECG simulator for MATLAB [10]. The first set was made so to reflect increase in heart rate every 5 minutes. The second set has the opposite pattern, with a decreasing heart rate every 5 minutes. The final set has a constantly changing heart rate in a pattern that remains constant throughout the dataset. Our test results are shown below with the comparison to Kubios, a leading HRV analysis software.


humanities and social sciences

hurj fall 2016: issue 22 Table 5: Parameter Analysis Results on Synthetic Data. Set 1 corresponds to the increasing heart rate set, while set 2 corresponds to the constantly changing heart rate set.

a) ECG with detected peaks in Altera

b) ECG with detected peaks in Kubios Figure 4. Comparison of QRS Detection Algorithms As noted in Figure. 3, Kubios’ peak detection algorithm both failed to detect and falsely detected multiple peaks in the data set, thus requiring extensive manual correction for proper analysis results.

Discussion In this paper, we introduced a new HRV analysis tool geared towards research use, with algorithms optimized for small animal model analysis. Through customized preprocessing algorithms, parameter extraction scripts for time and frequency domain analysis, and increased accessibility, Altera presents a complete solution for research labs conducting cardiac research on non-human subjects. Through the introduction and continual development of Altera, we hope to encourage and facilitate HRV research.


Table 4. Expected Parameter Analysis Results


[1]. Lombardi F, Sandrone G, Pernpruner S, Sala R, Garimoldi M, Cerutti S, Baselli G, Pagani M, Malliani A. (1987) Heart rate variability as an index of sympathovagal interaction after acute myocardial infarction. Am. J. Cardiol. 60(16):1239–1245. [2] Hon EH. (1963) Fetal heart rate monitoring. Biomed. Sci. Instrum. 1:287–291. [3] Wolf MM, Varigos GA, Hunt D, Sloman JG. (1978) Sinus arrhythmia in acute myocardial infarction. Med. J. Aust. 2(2):52–53. [4] Akselrod S, Gordon D, Ubel FA, Shannon DC, Berger AC, Cohen RJ. (1981) Power spectrum analysis of heart rate fluctuation: A quantitative probe of beat-to-beat cardiovascular control. Science. 213(4504):220–222 [5] Bilchick, Kenneth C., and Ronald D. Berger. “Heart Rate Variability.” J Cardiovasc Electrophysiol Journal of Cardiovascular Electro-

hurj fall 2016: issue 22

humanities and social sciences

physiology 17.6 (2006): 691-94. [6] Task force of the European society of cardiology and the North American society of pacing and electrophysiology,Heart rate variability – standards of measurement,physiological interpretation, and clinical use, Circulation 93(5) (1996) 1043–1065. [7] Matlab Help Files. “butter.” Mathworks, Cambridge MA. 2002. [8] Matlab Help Files. “filter.” Mathworks, Cambridge MA. 2002. [9] Lomb, N. R. (1976). “Least-squares frequency analysis of unequally spaced data”. Astrophysics and Space Science 39 (2): 447–462. doi:10.1007/BF00648343. [10] Raviprakash, Karthik, Matlab File Exchange, “ECG simulation using MATLAB.” Mathworks, Cambridge MA. 2006. [11] Giovanni Bianchi and Roberto Sorrentino (2007). Electronic filter simulation & design. McGraw-Hill Professional. pp. 17–20. ISBN 9780-07-149467-0. [12] Tessiro. (n.d.). In Wikipedia. Retrieved Novermber 11, 2016


Paul Kim is an avid undergraduate researcher at Johns Hopkins University Neuroengineering & Biomedical Instrumentation Lab. He is from Long Island, New York and currently a junior biomedical engineering student. As an engineer, Paul wants to surprise the world with his innovative biomedical devices in the future.


science and engineering

hurj fall 2016: issue 22

Corticosteroid Treatment as a Secondary Measure for Kawasaki’s Disease Sun Jay Yoo Department of Biomedical Engineering, Johns Hopkins University

Introduction Overview Kawasaki’s Disease (KD) was first investigated by pediatrician Tomisaku Kawasaki in 1967 as the mucocutaneous lymph node syndrome. At the time, it was not apparent that the heart was involved, but after years of autopsy studies of KD patients, it was established that the primary cause of death was through coronary artery aneurysms and the ensuing heart complications.1 Since this primary report, KD has been diagnosed in children worldwide and became the leading cause of acquired pediatric heart disease in Asia.3

Figure 2: A computer-rendered image depicting the heart and its coronary arteries. From left to right, the white arrows depict aneurysms in the RCA, LCA, and LAD.4

Figure 3: An artist’s rendition of coronary artery vasculitides and aneurysms as the result of KD.5 Figure 1: A description of the main symptoms of Kawasaki’s disease accompanied by artists’ renditions of a heart and a child affected by the disease.2

KD is an autoimmune disease categorized by small to medium vessel vasculitis.1, 3 Characterized by hypersensitive and overabundant white blood cells, KD poses a huge health risk when the target blood vessels are the coronary arteries. Through a quick cytokine storm, these vessels experience an inflammatory cascade. The development of “scar tissue” in the inner walls of the coronary arteries leads to their weakening and a potential blockage of blood that sustains the heart’s functionality.3 In the future, these coronary lesions and abnormalities can cause myocardial infarctions, pericardial effusions, and damage to the valvular muscles.1 KD primarily affects the right coronary artery (RCA), the left coronary artery (LCA), and the left anterior descending (LAD) artery.3


In Singapore, the KK Women’s and Children’s Hospital estimated that 51.4 out of 100,000 patients under five years old develop this disease, and about 30% of these children develop coronary aneurysms.6 Similarly, Kato et al., in a long-term study on KD, concluded that 25% of patients developed these coronary aneurysms, and only 55% showed slight improvement in the future with medication.7 Later in the study, it was discovered that 2% of KD patients experienced a heart attack, and death occurred in 1% of all patients.7 Through these estimations, it is evident that KD could pose a dangerous cardiac risk for pediatric patients, and there is a need for an effective treatment method. Although Son and Newburger explored research that suggested that infectious agents (such as Staphylococcus aureus and Streptococcus pyogenes) and genetic susceptibility cause KD, they concluded that the etiology of KD is still entirely unknown as these theo-

hurj 2014:issue issue 18 hurjspring fall 2016: 22

science and engineering science and engineering

ries have largely been disproven.1This lack of knowledge bacterial antigens.3 Often coupled with aspirin, IVIG is in the pathophysiology of this disease has limited medi- most effective when administered to KD patients in cal scientists’ ability in uncovering a targeted treatment early stages.3 The medical community has had mixed for KD. Thus, decades of research have instead been reviews on this standard treatment, and in recent times, focused around using empirical and clinical evidence research has shown that IVIG may not be an optimal and the patients’ overall clinical presentations to publish treatment procedure for KD patients. For instance, Uetreatment guidelines and protocols. Such was the work hara found that almost 20% of Japanese KD patients done by the Japanese Ministry of Health and Welfare in who received the standard IVIG treatment between 1984, and later, the American Heart Association.8 These 2003 and 2004 were nonresponders.11 Not only were guidelines have established a set clinical criterion to di- these patients unresponsive to the treatment, but they agnose KD and have proposed a standard treatment of were found to be at an even higher risk of developing IVIG to combat coronary abnormalities. The Japanese coronary aneurysms or even giant coronary aneurysms. Ministry classified coronary arteries as “abnormal” if the It is apparent that pediatricians need to be well precoronary artery inner-lumen diameter was greater than pared for KD patients who are unresponsive to primary 3.0 mm in patients younger than five or 4.0 mm in pa- IVIG. tients older than five. However, de Zorzi, in her study, found that these classifications were underestimating The Case for Corticosteroids the abnormalities, as they did not take into account the Corticosteroids, or intravenous methylprednisolone patient’s general size or body surface area (BSA).9 Thus, (IVMP), is the treatment of choice in similar forms of de Zorzi’s study proposed an alternate formula based vasculitis. As a steroidal anti-inflammatory, corticoupon BSA and changed the standard classification for steroids has hormonal capabilities that give it a wide determining KD health risks due to abnormal coronary range of uses. Much like IVIG, not much is known about arteries. Similarly, many other studthe pharmacokinetics and celluies have been done to further refine lar mechanics of corticosteroids and change the standard guidelines in KD patients. Corticosteroids is of KD diagnosis, classification, and hypothesized to manipulate signaltreatment.10, 3 However, more work ing pathways to decrease the tranneeds to be done to ensure to op- IVIG is a blood product scription of interleukin RNA and is timal efficiency and efficacy of KD the activation of consisting of natural an- theorized to block treatment, especially treatment for lymphocytes.3 IVIG nonresponders. tibodies that is adminstered through the vein. It was the initial treatment used in The Standard Treatment Japan to treat KD, before Furusho, Aspirin, or acetylsalicylic acid (ASA), in 1984, revealed that IVIG could is a nonsteroidal anti-inflammatobe more efficient.12 Furusho’s study ry medication that has been used provided overwhelming evidence to treat KD. Although it is hypothto show that high-dose IVIG radiesized to fight KD by stopping the cally eliminated fevers and corobinding of thrombocytes, it has been largely shown to nary abnormalities with more efficacy than corticostehave little effect on the frequency of coronary abnor- roids. Furthermore, an earlier study by Kato in 1979 even malities.1, 3 Not only is it ineffective for this purpose, it is suggested that corticosteroids may exert a detrimental also highly associated with the development of Reye’s effect as a primary treatment for KD.13 Through a rosyndrome, a potentially fatal brain disorder caused by bust randomized experiment, Kato’s study tested the the many side-effects of aspirin as long-term treatment. effects of aspirin, steroids, warfarin, and antibiotics as Thus, aspirin is usually a cautionary measure for patients KD treatment. Kato found that while aspirin did little to running high fevers. During the acute phases, aspirin is lower the prevalence of KD, steroids actually increased often administered with standard IVIG in an attempt to the likelihood of coronary abnormalities. Even though pose an additive anti-inflammatory effect.7 Kato’s study was difficult to reject due to its strong experimental design, subsequent studies have repeatedly IVIG is a blood product consisting of natural antibod- shown the opposite- that corticosteroids have no ill efies that is administered through the vein. Although it is fects and may actually be more effective than IVIG. primarily used to treat immunodeficiency such as HIV, it has been shown to reduce the occurrence of coronary For example, in 2004, Miura studied the role of cortiabnormalities in Kawasaki’s disease, a disease of hyper- costeroids as a secondary measure after the initial failimmunity.11 The cellular mechanisms of this treatment ure of primary IVIG.14 With a prospective observational are largely theoretical, but many have hypothesized study, the study found that the use of secondary corpossible processes such as white blood cell deactiva- ticosteroids coupled with the IVIG caused a statisticaltion, negative feedback of antibodies, the inhibition of ly significant decrease in the prevalence of coronary cytokine production, and the further protection against abnormalities, compared to IVIG alone. The study also


science and engineering found that the administration of the steroids quickly induced defervescence and suppressed fever recurrence. The potential side-effects of the steroids were not significantly seen, and all coronary artery lesions regressed after a year with steroid treatment. Similarly, Zhu continues to acknowledge that while IVIG/aspirin may be an acceptable primary treatment, corticosteroids also has its merits.15 The study by Zhu concluded that there is a significant decrease in coronary artery size after an initial treatment failure of IVIG when given corticosteroids. The study also concluded that corticosteroids decreases the duration of fever and the time for the C-reactive protein (CRP) level to return to normal. There is controversy regarding the secondary treatment for KD, with options such as re-administering IVIG or utilizing other treatments such as corticosteroids.

Method Through a retrospective observational study, a clinical statistics analysis was conducted to investigate the effects of corticosteroid treatment compared to secondary IVIG on the inner diameters of the right coronary artery (RCA), the left coronary artery (LCA), and the left anterior descending (LAD) artery for Singaporean Kawasaki disease patients unresponsive to primary IVIG. This method with quantitative variables was chosen in regards to the scientific discipline of pediatric immunology and the systematic procedures of medical research study. This method was modeled after de Zorzi’s 1998 KD research, which suggests novel techniques to retrospectively analyze KD data through statistics.9

Obtaining Data As IVIG was naturally most used as the initial/primary treatment for KD, this study aimed to compare the effects of two alternating secondary treatments for Singaporean patients that have already been categorized as IVIG non-responders: continued secondary IVIG versus corticosteroids. As the judgments on the time and the method of delivery of these treatments were entirely up to the medical discretion of the attending pediatrician, this was a factor that was not standardized across all subjects. To obtain a sample of IVIG non-responders, the Electronic Medical Records Xchange (EMRX) computer software was used. The EMRX is a healthcare initiative implemented by Singapore’s Ministry of Health in 2004 and allows for the sharing of medical records across all public hospitals in Singapore.16 With the availability of Hospital Inpatient Discharge Summaries (HIDS) for patients that visited public hospitals, all records of patients diagnosed with Kawasaki’s disease between 2004 and 2015 were extracted using their Medical Record Number. Out of that selection, KD patients who were IVIG non-responders and received either secondary IVIG or corticosteroids were collected. These patient casenote records were then isolated and reviewed. It is noteworthy to mention that the subjects were sampled only from a population of patients that visited public hos-


hurj fall 2016: issue 22 pitals in the Singapore, and there was undercoverage from populations that visited private hospitals or did not visit any hospitals at all. It should also be noted that selective randomization was not necessary, as it was an assumption that the progression and physiology of KD and its treatment effects were independent of patient race, culture, or social status, as indicated by a risk factor assessment study by Han.17 To measure the progression and health of patients with Kawasaki’s, the study determined that the inner diameter of the right coronary artery (RCA), the left coronary artery (LCA), and the left anterior descending (LAD) were the most important factors in determining the effectiveness of any KD treatment and their overall health, as concluded in the study conducted by de Zorzi.9 For each patient, each of these diameters were already recorded via echocardiography and measurements were made between the inner edges of the artery lumen. These measurements were made in millimeters and were generally classified as either aneurysm or dilation by the discretion of the respective pediatrician. Furthermore, in order to track the progression of the disease and treatment across time, the measurements of these arteries were conducted several times per patient and each date of measurement was recorded along with the patient’s weight and height at the time. In addition to this data, other information such as gender, the age at KD onset, type of KD, and the existence of several clinical symptoms (conjunctivitis, body rash, swollen lymph nodes, strawberry tongue, etc.) were recorded. All the data was entered into IBM’s Statistical Package for the Social Sciences (SPSS).18 This computer software facilitated the compilation of mass data and the use of statistical analysis and testing. RStudio, for the R statistical programming language, was used to verify the results achieved by SPSS.19 All patients in the sample were non-identifiable and the confidentiality of patient data was placed under strict audit controls. Thus, an institutional review board (IRB) inspection was not necessary.

Standardization of Data As this study is concerned with the comparison of coronary abnormalities over time, a method of standardization was implemented for unbiased analysis. First, for each patient, the BSA was calculated for each time the coronary arteries were measured via echocardiography. In physiology and medicine, the BSA is an accurate indicator of metabolic mass and is calculated using the patient’s weight (in kilograms) and height (in centimeters). Haycock, in 1978, calculated a BSA formula for pediatric patients using a randomized selection of infants and children.20 As KD is most prevalent in children under the age of five, the following formula by Haycock was used: BSA = 0.024265 * Weight0.5378〗 * Height0.3964〗

hurj fall 2016: issue 22 Then, in order to standardize every coronary artery measurement to a hypothetical “average”, a z-score was calculated for each artery. This standardization was based off of the “normal” dimension distributions from the universal Boston standard that was calculated by McCrindle in his 2007 study.10 This would convert each artery measurement to an indication of how normal or abnormal the measurement was for each patient of his or her BSA. The following formulas from McCrindle’s study were used for the mean and standard deviation for each artery measurement (in millimeters):

science and engineering Findings and Results Out of the 78 IVIG non-responders, 33 were treated with secondary IVIG and 45 were treated with secondary corticosteroids. In order to first compare and verify that there were no differences in the general clinical characteristics between the two treatment groups, multiple two-sample Student’s t-tests were conducted alongside further verification with nonparametric tests (the Mann-Whitney U-test or the permutation test).

RCA Mean =0.26117*(BSA0.39992)-0.02756 RCA Standard Deviation =0.02407+(0.01597*BSA) LCA Mean =0.31747*(BSA0.36008)-0.02887 LCA Standard Deviation=0.03040+(0.01514*BSA) LAD Mean =0.26108*(BSA0.37893)-0.02852 LAD Standard Deviation =0.01465+(0.01996*BSA) Then, using these corresponding artery mean and standard deviation formulas, the z-score for each measurement was calculated with the following: Z-score = (observed artery measurement - mean) / standard deviation Table 1: Summary of the results of the hypothesis tests for the four selected clinical characteristics.

If this z-score was above 2.0 for a certain artery measurement, this artery for that t=time was categorized as “abnormal”. Thus, after calculating the z-scores of every measurement across time for each patient, the first coronary abnormality occurrence and the last abnormality occurrence was recorded for each patient. This allows for the calculation of a new figure: the time to coronary abnormality resolution (due to treatment). This value (in years) was calculated by subtracting the date of the last abnormality with the date of the first abnormality. Essentially, this number, for each patient, shows how long the coronary abnormality (whether they are dilations or aneurysms) lasted. A short duration of abnormality would mean that the treatment is effective, and a long duration would mean that the treatment is ineffective.

Statistical Hypothesis Tests and Analysis This study relied on several different hypothesis tests to reach a conclusion about the compared effectiveness of the two treatments. Every test that was run in the study compares certain values and aspects that are recorded for the secondary corticosteroid sample versus the secondary IVIG sample. Before the main hypothesis test comparing the effectiveness of the two treatments was run, general hypothesis tests were first conducted to compare the demographics of the two treatment samples. After these checks were verified, the specific hypothesis test to compare the coronary effectiveness of the two treatments was conducted.

The results of the preliminary hypothesis tests showed that there was no significant difference between the two treatment groups in regards to gender, age of onset, type of KD, and clinical symptoms (excluding artery abnormality). Now that these “control” variables were verified to have no difference, the test to compare effectiveness of IVIG and corticosteroids was conducted with further validity. In order to compare the treatment effectiveness, a Kaplan-Meier survival analysis hypothesis test was conducted between the two samples. While the Kaplan-Meier test is often conducted in medical research to compare the patient survival probability from treatments over periods of time, the test is used in this study to compare the probabilities of coronary abnormality resolution- the time it takes for a patient to become healthy. After standardizing the coronary artery measurement data with the BSA and calculating the time until abnormality resolution for each patient, the Kaplan-Meier test was conducted.


science and engineering

hurj fall 2016: issue 22 gument that it is fully up to the pediatrician in choosing a secondary treatment. A pediatrician would have to take into account the patient’s medical history, clinical status, and prognosis before selecting one of the two treatments. For example, a pediatrician might choose not to administer the corticosteroids, a synthetic hormone, to an infant if he or she believes that there could be major negative side effects that the infant may be susceptible to. Administering the natural blood product IVIG may be a safer option in some cases as no artificial compounds are introduced to the patient. However, if a patient is already unresponsive to IVIG, they may continue to that prognosis. Both treatments have their individual benefits and drawbacks, but in the long term, both treatments have the potential to be equally effective in eliminating coronary abnormalities.

Figure 4: A Kaplan-Meier analysis graph depicting the time to coronary abnormality resolution for secondary IVIG vs. corticosteroids. With the green line depicting the curve for corticosteroid patients and the blue line (the line that extends to the right) depicting that for secondary IVIG patients, it was seen that those who received corticosteroids have a larger probability of early resolution (within 40 months), whereas IVIG patients can take up to 110 months. However, the results of the Kaplan-Meier test revealed that p = 0.11 with the null hypothesis that there was no difference in resolution time. With a significance level at 0.05, this would mean that the time to achieve coronary abnormality resolution was not statistically significant between the two treatment groups. The graph may reveal that corticosteroids was more effective, but this may be due to random chance and sampling alone.

After conducting preliminary statistical hypothesis tests to verify that there were no other clinical characteristic differences between the two treatment groups, it was found that there was no statistically significant difference between the time until coronary abnormality resolution for patients receiving secondary IVIG versus those receiving corticosteroids. The Kaplan-Meier test failed to reject the null hypothesis that there is no significant difference, and according to our data, IVIG and corticosteroids are equally effective measures in eliminating coronary abnormalities after the failure of primary IVIG.

Discussion So far, it has been established that there is a high probability that a KD patient will be unresponsive to primary IVIG.1, 3, 11 A secondary treatment measure needs to be administered in order to control the potentially fatal coronary abnormalities; however, there exists a controversy between which treatment should be given.12-14 According to the data from this study, it was revealed that there is no significant difference in the effectiveness of corticosteroids versus IVIG in resolving coronary abnormalities as secondary measures. This supports the ar-


There have been laboratory tests hypothesizing the reasoning for the effectiveness of both IVIG and corticosteroids in KD patients. In 2002, Tsujimoto conducted an in vitro laboratory experiment to study the effects of IVIG on KD neutrophil levels.21 Neutrophils are specifically important to KD as they are known to be the first-responders of inflammatory cells to attack a site of inflammation. In KD, these neutrophils might travel to the coronary arteries and cause inflammation as if there was some sort of pathogenic infection. Tsujimoto found that the average number of peripheral neutrophils in the post-IVIG stage were significantly less than that of the pre-IVIG stage. Furthermore, the average proportion of apoptotic neutrophils was significantly higher after IVIG was administered. This suggests that IVIG is effective in combating coronary inflammation (abnormalities) in KD by inducing the accelerated apoptosis of inflammatory neutrophils. This would decrease immunological sensitivity and prevent further coronary abnormality. As IVIG is also an immune product that actually aids in immune response, it is hypothesized that the increased introduction of these immunoglobulins contributes to a negative feedback system that decreases neutrophil counts through their apoptosis. This study by Tsujimoto provides a novel explanation as to why IVIG alone could be an effective measure from a pathophysiological standpoint. Studies have also been conducted to explain why corticosteroids, administered after primary IVIG, might also be an effective treatment measure to eliminate coronary abnormalities and inflammation. In 2003, Okada used a multicenter prospective and randomized study to study the effect of corticosteroids on immune signaling if administered after and alongside IVGG (a general class of IV antibody treatment that include IVIG).22 These cytokines (immune signaling molecules), more specifically categorized as interleukins, are expressed by white blood cells in order to regulate and expedite immune responses. Okada first found that pre-IVIG serum levels of several specific interleukins were significantly higher in KD patients than in healthy controls.

hurj fall 2016: issue 22 Okada then found that IVGG alone failed to significantly decrease interleukin levels. Only when patients were administered with corticosteroids did their interleukin levels decrease significantly (both within 24 hours and after). As it was already previously hypothesized that corticosteroids may have the effect of blocking immune signaling pathways, Okada’s study brings forth a theory as to how corticosteroids effectively resolves coronary abnormalities. In addition to the explanations brought forth by these two research studies, Han’s 2000 study showed that the scope of generalization for most studies that investigate IVIG non-responder populations can be extended to IVIG responder populations as well. Han concluded that there were no significant differences in clinical demographics, symptoms, and overall outcomes between primary IVIG non-responders and IVIG responders. Thus, the result that there is no significant difference between the effectiveness of IVIG and corticosteroids can be extended to any KD patient population with similar characteristics. As these studies increase the validity of our findings, it can strongly be asserted that IVIG and corticosteroids are both equally effective measures to treat KD- whether the patient is an IVIG non-responder or not. The results of this study are valuable because they provide conclusions to clear up the controversy behind a safe and effective secondary treatment measure for KD. As the results rely on reliable and original data, this study provides further insight into a standard treatment protocol based on a population that has never been studied before in this regard. Doctors can now be more certain of what to do when the first administration of IVIG fails, and patients can be assured that either of the two treatments is equally as effective as the other in resolution.

science and engineering 1998; 154(2): 254-258. Doi: 10.1016/S0022-3476(98)70229-X. (6) McCrindle, Brian et al. Coronary Artery Involvement in Children With Kawasaki Disease. Circulation. 2007; 116: 174-179. doi: 10.1161/CIRCULATIONAHA.107.690875 (7) Uehara R, et al. Analysis of Potential Risk Factors Associated With Nonresponse to Initial Intravenous Immunoglobulin Treatment Among Kawasaki Disease Patients in Japan. Pediatric Infectious Disease Journal. 2008; 27(2): 155-160. doi:10.1097/INF.0b013e31815922b5. (8) Furusho K, Kamiya T, Nakano H, et al. High-dose intravenous gammaglobulin for Kawasaki disease. Lancet. 1984; 2(8411): 1055-1058. (9) Kato H, Koike S, Yokoyama T. Kawasaki disease: effect of treatment on coronary artery involvement. Pediatrics. 1979; 63(2): 175-179. (10) Miura M, Tamame T, Naganuma T, Chinen S, Matsuoka M, Ohki H. Steroid pulse therapy for Kawasaki disease unresponsive to additional immunoglobulin therapy. Paediatrics and Child Health. 2011; 16(8): 479-484. (11) Zhu B, Lv H, Sun L, et al. A meta-analysis on the effect of corticosteroid therapy in Kawasaki disease. European Journal of Pediatrics. 2012; 171(3): 571-578. doi:10.1007/s00431-011-1585-4. (12) Han R, Silverman E, Newman A, McCrindle B. Management and Outcome of Persistent or Recurrent Fever After Initial Intravenous Gamma Globulin Therapy in Acute Kawasaki Disease. Archives of Pediatrics and Adolescent Medicine. 2000; 154(7): 694-699. doi:10.1001/ archpedi.154.7.694. (13) Haycock G, Schwartz G, Wisotsky D. Geometric method for measuring body surface area: A height-weight formula validated in infants, children, and adults. The Journal of Pediatrics. 1978; 93(1): 62-66. Tsujimoto H, et al. Intravenous Immunoglobulin Therapy Induces Neutrophil Apoptosis in Kawasaki Disease. Clinical Immunology. 2002; 103(2): 161–168. doi:10.1006/clim.2002.5209. (14) Okada Y, et al. Effect of corticosteroids in addition to intravenous gamma globulin therapy on serum cytokine levels in the acute phase of Kawasaki disease in children. The Journal of Pediatrics. 2003; 143(3): 363–367. doi:10.1067/S0022-3476(03)00387-1.


Future Directions As only one sample was studied, further research must be done to validate the results. Observational research on other populations of KD IVIG non-responders must be continued and a larger randomized and prospective controlled trial must be conducted to confirm these preliminary observations. However, the finding of this study that both treatments are equally effective is a promising step towards tackling the knowledge gap in secondary treatment research for KD.

References (1) Son MB, Newburger J. Kawasaki Disease. American Academy of Pediatrics. 2013; 34(4): 151-162. doi: 10.1542/pir.34-4-151. (2) Newburger JW, Takahashi M, Gerber MA, et al. Diagnosis, Treatment, and Long-Term Management of Kawasaki Disease. Circulation. 2004; 110: 2747-2771. doi: 10.1161/01.CIR.0000145143.19711.78. (3) Kato H, et al. Long-term Consequences of Kawasaki Disease. Circulation. 1996; 94: 1379-1385. doi: 10.1161/01.CIR.94.6.1379. (4) Research Committee on Kawasaki’s disease. Report of subcommittee on standardization of diagnostic criteria and reporting of coronary artery lesions in Kawasaki disease. Tokyo, Japan: Ministry of Health and Welfare; 1984. (5) de Zorzi, Andrea et al. Coronary artery dimensions may be misclassified as normal in Kawasaki disease. The Journal of Pediatrics.

Sun Jay Yoo, a member of the Freshman Class of 2020, is majoring in Biomedical Engineering and Computer Science. Although he is South Korean, he has lived in Singapore all his life. After graduation, he plans on focusing in computational biology/medicine and entering either graduate school or industry.


science and engineering

hurj fall 2016: issue 22

Ethnic differences in parental feeding behaviors in a socio-economically diverse UK sample of preschoolers and their parents CeCe Gu Department of Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine

Introduction Obesity and being overweight are prevalent among UK children (1), and children of the two most numerous ethnic minority populations in the UK – Black Afro-Caribbean and South Asian (2, 3) – may be disproportionately affected. For example, the 1999 Health Survey for England data revealed that 13% of Afro-Caribbean girls, 9% of Pakistani boys, and 8% of Pakistani girls and Indian boys were obese, compared to 6% of boys and 6% of girls in the larger population (4). Furthermore, when compared to White British children of same weight status, South Asian and Black Afro-Caribbean children have higher insulin resistance, diastolic blood pressure, and HbA1c. This thus puts them at a higher risk for consequences of obesity such as cardiovascular diseases and diabetes (5), making it increasingly important to understand the origins of child obesity in these populations. Parents engage in various feeding practices to influence their child’s eating behaviors (6, 7). Restriction and Pressure to eat (measured using the Child Feeding Questionnaire [CFQ] (8)) reflect an authoritarian parental feeding style (high in both demandingness and responsiveness) (9). Restriction is defined as the parents’ limiting of the child’s access to foods, which has been correlated with avid appetite (10), higher child weight status and increased intake (11-14). Pressure to eat has been related to lower appetite (10, 12), lower fruit and vegetable intake (15), and lower child BMI (12-14). In contrast, Monitoring (measured by the CFQ (8)) and Prompting to eat (measured by the Parental Feeding Style Questionnaire [PFSQ] (16)) reflect a more authoritative style (low in demandingness and high in responsiveness) (9, 10). Monitoring, defined as a milder form of restriction, correlates with healthier eating behavior in children and healthier BMI (14, 17). The Prompting to eat was conceived as representing a gentler form of promoting intake than Pressure to eat (10). The authroritative feeding style has been associated with greater fruit and vegetable consumption over the authoritarian feeding style (18). Non-nutritive feeding practices include Emotional feeding, i.e. using food to influence the child’s emotions, and Instrumental feeding, i.e. using food to


reward certain behaviors. Both practices have been shown to be associated with the obesogenic eating behaviors in children such as increased snacking (19), and emotional and disinhibited eating (20). However, findings have been inconsistent, with some studies showing no association between child weight/adiposity and Restriction, Pressure to eat (10, 21), Monitoring (10, 22) or non-nutritive feeding practices (10, 23), pointing towards the importance of considering potential effect modifiers such as ethnicity and culture. Parents of ethnic minorities may differ in the degree to which they engage in controlling and non-nutritive feeding behaviors. Specifically, studies have suggested that parents of African-American and Hispanic backgrounds in the US (24-26) and parents of Black Afro-Caribbean background in the UK (27) display more controlling feeding behaviors than White parents in the respective regions. One study also demonstrated that South Asian parents in the US engaged in more non-nutritive feeding behaviors and Pressure to eat (28) than White American parents. Differences in levels of controlling and non-nutritive feeding practices and in the home food environment have also been observed among different minority groups. For example, Hispanic parents in the US reported higher control and more child-responsive feeding strategies than African-American parents (29). Cultural differences in weight and food-related attitudes may drive ethnic differences in parent feeding behaviors. For example, evidence suggests that low-income African-American parents may have a higher perceived child overweight threshold than low-income White and Hispanic-American parents (30), while other evidence suggests that parents may be more likely to engage in controlling feeding behaviors when they are more concerned about their child’s weight (31) and with higher child weight (32). Parents of ethnic minorities may hold different cultural attitudes towards food, such as associating greater food consumption and availability with higher social status (33). Black Afro-Caribbean and South Asian parents may also use food to preserve their cultural identities as immigrants (28, 34) , potentially leading them to pressure their children to eat ethnic foods. As minorities, they may also experience more stress and social

hurj 2014:issue issue 18 hurjspring fall 2016: 22 obstacles, potentially leading to a lack of patience with disruptive behaviors in their children (35), and resulting in the use of more controlling and low-responsive, non-nutritive feeding behaviors. Notably, more acculturated Latino parents have been shown to engage in less controlling general parenting practices which in turn are associated with unhealthy eating habits (36), highlighting the strong influence of culture. Minority families are also more likely to have lower incomes, making them vulnerable to consuming cheaper, more energy-dense foods (33). As the UK population becomes more diversified (3), it becomes increasingly important to explore ethnic differences in parental feeding behaviors and how they might contribute to disparities in pediatric obesity. Based on previous findings, we hypothesized that 1) South Asian and Black Afro-Caribbean parents would score higher than White British parents on authoritarian feeding behaviors (Restriction, Pressure to eat); and 2) White British parents would score higher than South Asian and Black Afro-Caribbean parents on authoritative feeding behaviors (Monitoring, Prompting to eat). As exploratory aims, we also tested if there were sex differences in ethnic effects, and whether ethnic differences were mediated by child BMI or perceptions and concerns relating to the child’s weight.

Subjects/Methods Participants and Procedures 271 White British parents, 59 South Asian parents, and 42 Black Afro-Caribbean parents from 12 London primary schools participated. The schools were chosen to represent a diverse socioeconomic range based on student eligibility for Free School Meals, a government-benefit for low-income families. Letters were sent to the parents with study information and the option to exclude their child prior to data collection. Questionnaires were then distributed to the parents on days of data collection and reminders were sent to those who did not respond within two weeks. Children were measured and weighed at school. Measures

Demographic and Anthropometric Information Demographic information including ethnicity, marital status, education level, employment status, and annual family income was self-reported by parents. Following categorizations used in previous studies of similar ethnic groups (27, 37), parents of Indian, Pakistani, and Bangladeshi origin were grouped into a single South Asian category, while Black African and Black Caribbean parents were combined into a Black Afro-Caribbean group. Children were weighed and measured by trained researchers using standard methods. BMI z-scores were based on 1990 UK growth reference curves and World Obesity Clinical Care weight categories were determined (38, 39).

science and engineering science and engineering Parental Feeding Behavior The Child Feeding Questionnaire [CFQ] (8)] was used to measure Restriction (8 items), Monitoring (3 items), and Pressure to Eat (4 items), as well as Perceived Child Weight (3 items), and Concern about Child Overweight (5 items). A Concern about Child Underweight (5 items) subscale was adapted from the Concern about Child Overweight subscale by changing the expressions ‘eating too much’ to ‘not eating enough’, ‘having to diet to maintain desirable weight’ to ‘having to eat high energy foods to maintain a reasonable weight’, and ‘overweight’ to ‘underweight’. Parental Prompting to Eat (8 items), Emotional Feeding (5 items) and Instrumental Feeding (4 items) were taken from the Parental Feeding Style Questionnaire [PFSQ] (16)].

Statistical Analysis To investigate ethnic differences (White British vs. Black Afro-Caribbean vs. South Asian) in sample characteristics, we ran chi-square tests for categorical variables and univariate ANOVAs for continuous variables, using specific chi-squared tests and Tukey’s Range Tests for post-hoc exploration of differences. To examine ethnic differences in parental feeding practices, we ran univariate ANOVAs for each parent feeding variable with Tukey’s Range Test for post-hoc differences. To investigate whether ethnic differences in parent feeding were mediated by perceptions and concerns relating to the child’s weight, we ran ANCOVAs for ethnic differences in each parent feeding behavior while controlling for each perception and concern variable separately. To explore the possibility that ethnic effects differed by child sex, we repeated these analyses for boys and girls separately, using Tukey’s Range Test to explore differences. We also repeated all of the above analyses controlling for a) child’s BMI z-score, b) parent education and c) income to exclude the influence of other potential confounders.

Results Table 1 shows descriptive statistics for each ethnic group. Most of the parents had a school degree or lower level of education, and earned less than £40,000 annually (p<0.01). Children’s ages ranged from three to six years. Black Afro-Caribbean children had higher BMI z-score values when compared to the other two ethnic groups (p<0.01). Chi-square tests showed that Black Afro-Caribbean mothers live without partner more frequently when compared to the other ethnic groups (p<0.01), and White British mothers perceived themselves as more overweight compared to Black Afro-Caribbean mothers (p<0.05). Black Afro-Caribbean parents reported more concern about their child being both over- and underweight (both p<0.001). White British parents reported greater use of Monitoring than Black Afro-Caribbean parents (p<0.01). Among negative parental feeding practices, White British parents reported lowest Pressure to Eat, while Black Afro-Caribbean parents showed higher scores (p<0.001). Regarding non-nutritive feeding practices, White British parents reported less use of both Emotional and Instrumental Feeding practices


science and engineering when compared to South Asian (p<0.05) and Black Afro-Caribbean (p<0.01), respectively. All ethnic groups had similar scores for Restriction and Prompting to Eat. These results remained significant when controlling for a) parent education, b) child’s BMI z-score, c) parental concern about child overweight, and d) parental concern about child underweight. Mean scores and standard errors for the four parental feeding practices showing differential ethnic effects for each sex are illustrated in Figure 2. For boys, but not girls, differences in Monitoring were apparent (p=0.002), such that Black Afro-Caribbean parents reported lower use of this practice when compared to White British and South Asian parents (Tukey p=0.001 and p=0.048, respectively). A slightly different pattern of significant differences for each sex (boys: p=0.002, girls: p=<0.001) was also observed for Pressure to Eat, such that both Black Afro-Caribbean and South Asian parents of boys scored higher than White British parents (Tukey p=0.011 and p=0.043, respectively), but only Black Afro-Caribbean parents of girls scored higher than those of White British parents of girls (Tukey p=0.001). For Emotional Feeding, the pattern of ethnic differences also differed for boys and girls, such that South Asian parents of girls reported greater use of Emotional Feeding when compared to White British parents of girls (main effect p<0.001; Tukey p<0.001). However, this was not observed in boys, which showed a significant sex difference by ethnic group interaction for Emotional Feeding (main effect p=0.001). For Instrumental Feeding, ethnic differences were only apparent in girls (p=0.005), such that South Asian parents reported greater use of Instrumental Feeding when compared to White British parents (Tukey p=0.010). All results remained significant when controlling for Perceived Child Weight, Concern about Child Overweight, Concern about Child Underweight (8), child’s BMI z-score, parent education and income.

Discussion The purpose of our study was to investigate whether parental feeding behaviors differed by ethnicity and, if so, to explore differences by child sex, and possible mediators of observed ethnic differences. As expected, we observed significant differences in feeding practices between the three ethnic groups. Specifically, South Asian and Black Afro-Caribbean parents scored higher than White families on Pressure to Eat, Emotional Feeding and Instrumental Feeding, and Black Afro-Caribbean parents scored lower than White families on Monitoring. Our findings are consistent with previous studies done on minority parents in both the US (24-26, 29) and the UK (27). Contrary to our predictions, the higher levels of authoritarian and non-nutritive feeding behaviors exhibited by minority parents could not be explained by parental concern about child weight (22, 31). They were also independent of parental education, income, and child’s


hurj fall 2016: issue 22 BMI z-score. This suggests that cultural factors could be more important determinants of ethnic variation in feeding behavior than these potential confounders. An interesting incidental finding was that ethnic minority parents were more concerned than White parents about both child over- and underweight. This relationship was driven primarily by the Afro-Caribbean group, and the association persisted even when controlling for perceived and actual child weight. This suggests that weight-related concerns may not necessarily be translated into differences in feeding practices among minorities. In support of this, a previous study in the US found that concern about child overweight led to increased restriction and slower increase in total fat mass in White children, but initial concern was not correlated with feeding practices in African-American parents and had no longitudinal effect on child fat mass (25). The ethnic differences in feeding that we observed could have resulted from a multitude of social and cultural pressures. Previous research has shown obesity risk among UK Bangladeshi children to be correlated with low SES, while within the Black Afro-Caribbean population a multitude of sociodemographic and cultural factors are implicated (40). For example, qualitative evidence suggests that Afro-Caribbean parents perceive themselves as making efforts to both hold onto their family traditions and live healthily in a community context that they see as stressful and fast paced (34) – a combination of factors that could lead to less responsive parenting practices. A qualitative study of Asian-Indian mothers in the US found that feeding behaviors, such as pressure to eat and use of food as a reward, were influenced by explicit efforts to encourage the consumption of Indian dishes and pass on cultural values of respecting and not wasting food (28). Interestingly, our results also showed that some of the trends we observed differed by sex, with a significant sex difference by ethnic group interaction emerging for Emotional Feeding, such that White British parents reported lower scores than South Asian parents for girls only. Child BMI may not be the ideal way to measure ethnic differences in body fatness, since previous research suggests that children from different ethnic groups have disproportionate fat mass at similar BMIs. This is due to differences in height and body composition, which may contribute to minority children’s increased risk of insulin resistance and cardiovascular diseases (5). As in much parent feeding research, practices were self-reported rather than directly observed, and we therefore cannot exclude the possibility that parents of different ethnic origins could have different interpretations of questionnaire items given different cultural ideals for parenting (29, 34). Larger studies should test whether the relationship between feeding practices and child BMI z-scores differs between ethnic groups (25), and whether this relationship is further impacted by child sex (29) – analyses for which this study was

hurj fall 2016: issue 22 underpowered. Nevertheless, our results provide compelling evidence that parental feeding differs by ethnicity, and suggest that cultural variation should therefore be considered in efforts to develop effective child obesity interventions.

References (1) Ng Fat L. Health Survey for England, 2014. Health & Social Care Information Centre; 2015. (2) Statistics OfN. Ethnicity and National Identity in England and Wales 2011. 2012. (3) Jivraj S. How has ethnic diversity grow 1991-2001-2011? 2012. (4) Saxena S, Ambler G, Cole TJ, Majeed A. Ethnic group differences in overweight and obese children and young people in England: cross sectional survey. Arch Dis Child. 2004;89(1):30-6. (5) Whincup P, Nightingale C, Donin A, Rapala A, Joysurry D, Prescott M, et al. Ethnic differences in cardiovascular risk in childhood: comparison of UK south Asian, African-Caribbean and European children. J Epi Comm Health. 2011;65(Suppl 1):A312-A. (6) Vollmer RL, Mobley AR. Parenting styles, feeding styles, and their influence on child obesogenic behaviors and body weight. A review. Appetite. 2013;71:232-41. (7) Clark HR, Goyder E, Bissell P, Blank L, Peters J. How do parents’ child-feeding behaviours influence child weight? Implications for childhood obesity policy. J Public Health (Oxf). 2007;29(2):132-41. (8) Birch LL, Fisher JO, Grimm-Thomas K, Markey CN, Sawyer R, Johnson SL. Confirmatory factor analysis of the Child Feeding Questionnaire: a measure of parental attitudes, beliefs and practices about child feeding and obesity proneness. Appetite. 2001;36(3):201-10. (9) Hughes SO, Power TG, Orlet Fisher J, Mueller S, Nicklas TA. Revisiting a neglected construct: parenting styles in a child-feeding context. Appetite. 2005;44(1):83-92. (10) Carnell S, Benson L, Driggin E, Kolbe L. Parent feeding behavior and child appetite: associations depend on feeding style. Int J Eat Disord. 2014;47(7):705 (11) Faith MS, Scanlon KS, Birch LL, Francis LA, Sherry B. Parent-child feeding strategies and their relationships to child eating and weight status. Obes Res. 2004;12(11):1711-22. (12) Ventura AK, Birch LL. Does parenting affect children’s eating and weight status? Int J Behav Nutr Phys Act. 2008;5(1):1-12. (13) Carnell S, Kim Y, Pryor K. Fat brains, greedy genes, and parent power: a biobehavioural risk model of child and adult obesity. Int Rev Psychiatry. 2012;24(3):189-99. (14) Jansen PW, Tharner A, van der Ende J, Wake M, Raat H, Hofman A, et al. Feeding practices and child weight: is the association bidirectional in preschool children? Am J Clin Nutr. 2014;100(5):1329-36. (15) Galloway AT, Fiorito L, Lee Y, Birch LL. Parental pressure, dietary patterns, and weight status among girls who are “picky eaters”. J Am Diet Assoc. 2005;105(4):541-8. (16) Wardle J, Sanderson S, Guthrie CA, Rapoport L, Plomin R. Parental feeding style and the inter-generational transmission of obesity risk. Obes Res. 2002;10(6):453-62. (17) Brann LS, Skinner JD. More controlling child-feeding practices are found among parents of boys with an average body mass index compared with parents of boys with a high body mass index. J Am Diet Assoc. 2005;105(9):1411-6. (18) Patrick H, Nicklas TA, Hughes SO, Morales M. The benefits of authoritative feeding style: caregiver feeding styles and children’s food consumption patterns. Appetite. 2005;44(2):243-9. (19) Sleddens EFC, Kremers SPJ, De Vries NK, Thijs C. Relationship between parental feeding styles and eating behaviours of Dutch children aged 6–7. Appetite. 2010;54(1):30-6. (20) Rodgers RF, Paxton SJ, Massey R, Campbell KJ, Wertheim EH, Skouteris H, et al. Maternal feeding practices predict weight gain and obesogenic eating behaviors in young children: a prospective study. Int J Behav Nutr Phys Act. 2013;10:24. (21) Campbell K, Andrianopoulos N, Hesketh K, Ball K, Crawford D, Brennan L, et al. Parental use of restrictive feeding practices and child BMI z-score. A 3-year prospective cohort study. Appetite. 2010;55(1):84-8. (22) Webber L, Hill C, Cooke L, Carnell S, Wardle J. Associations between child weight and maternal feeding styles are mediated by maternal perceptions and concerns. Eur J Clin Nutr. 2010;64(3):259-65.

science and engineering (23) Carnell S, Wardle J. Associations between Multiple Measures of Parental Feeding and Children’s Adiposity in United Kingdom Preschoolers. Obesity. 2007;15(1):137-44. (24) Cardel M, Willig AL, Dulin-Keita A, Casazza K, Beasley TM, Fernández JR. Parental feeding practices and socioeconomic status are associated with child adiposity in a multi-ethnic sample of children. Appetite. 2012;58(1):347-53. (25) Spruijt-Metz D, Li C, Cohen E, Birch L, Goran M. Longitudinal influence of mother’s child-feeding practices on adiposity in children. J Pediatr. 2006;148(3):314-20. (26) Spruijt-Metz D, Lindquist CH, Birch LL, Fisher JO, Goran MI. Relation between mothers’ child-feeding practices and children’s adiposity. Am J Clin Nutr. 2002;75(3):581-6. (27) Blissett J, Bennett C. Cultural differences in parental feeding practices and children/’s eating behaviours and their relationships with child BMI: a comparison of Black Afro-Caribbean, White British and White German samples. Eur J Clin Nutr. 2013;67(2):180-4. (28) Momin SR, Chung KR, Olson BH. A qualitative study to understand positive and negative child feeding behaviors of immigrant Asian Indian mothers in the US. Matern Child Health J. 2014;18(7):1699-710. (29) Hughes SO, Anderson CB, Power TG, Micheli N, Jaramillo S, Nicklas TA. Measuring feeding in low-income African–American and Hispanic parents. Appetite. 2006;46(2):215-23. (30) Sherry B, McDivitt J, Birch LL, Cook FH, Sanders S, Prish JL, et al. Attitudes, practices, and concerns about child feeding and

child weight status among socioeconomically diverse white, Hispanic, and African-American mothers. J Am Diet Assoc. 2004;104(2):215-21. (31) Gray WN, Janicke DM, Wistedt KM, Dumont-Driscoll MC. Factors associated with parental use of restrictive feeding practices to control their children’s food intake. Appetite. 2010;55(2):332-7. (32) Polfuss M, Frenn M. Parenting behaviors of African American and Caucasian families: parent and child perceptions, associations with child weight, and ability to identify abnormal weight status. J Pediatr Nurs. 2012;27(3):195-205. (33) Kumanyika SK. Environmental influences on childhood obesity: ethnic and cultural influences in context. Physiol Behav. 2008;94(1):61-70. (34) Bramble J, Cornelius LJ, Simpson G. Eating as a cultural expression of caring among Afro-Caribbean and African American women: understanding the cultural dimensions of obesity. J Health Care Poor Underserved. 2009;20(2 Suppl):53-68. (35) Hurley KM, Black MM, Merry BC, Caulfield LE. Maternal mental health and infant dietary patterns in a statewide sample of Maryland WIC participants. Matern Child Nutr. 2015;11(2):229-39. (36) Arredondo EM, Elder JP, Ayala GX, Campbell N, Baquero B, Duerksen S. Is parenting style related to children’s healthy eating and physical activity in Latino families? Health Education Research. 2006;21(6):862-71. (37) Pomerleau J, McKeigue PM, Chaturvedi N. Factors associated with obesity in South Asian, Afro-Caribbean and European women. Int J Obes Relat Metab Disord. 1999;23(1):25-33. (38) Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ. 2000;320(7244):1240-3. (39) Cole TJ, Lobstein T. Extended international (IOTF) body mass index cut-offs for thinness, overweight and obesity. Pediatr Obes. 2012;7(4):284-94. (40) Zilanawala A, Davis-Kean P, Nazroo J, Sacker A, Simonton S, Kelly Y. Race/ethnic disparities in early childhood BMI, obesity and overweight in the United Kingdom and United States. Int J Obes (Lond). 2015;39(3):520-9.


science and engineering

hurj fall 2016: issue 22


CeCe Cihang Gu is a senior majoring in history. She is originally from Bayside, New York and plans on attending medical school after graduation. She loves food and is interested in research about obesity, eating behaviors and nutrition. During her free time, she loves to cook for her friends and runs a food blog.

Figure Legends

Figure 1. Mean (+ SE) parent feeding practice scores by ethnic group. *p<0.05, **p=0.001; ***p<0.001.


Figure 2. Mean (+ SE) parent feeding practice scores by ethnic group for each sex (practices showing sex differences only). *p<0.05, **p=0.001; ***p<0.001; a: interaction between child sex and ethnic group.

hurj fall 2016: issue 22

science and engineering