Page 1

Multiple Sclerosis and Related Disorders 27 (2019) 164–170

Contents lists available at ScienceDirect

Multiple Sclerosis and Related Disorders journal homepage:

Diabetes and anxiety adversely affect cognition in multiple sclerosis a,b,⁎





Ruth Ann Marrie , Ronak Patel , Chase R Figley , Jennifer Kornelsen , James M. Bolton , Lesley Graffc, Erin L. Mazerolleh, James J. Marriotta, Charles N. Bernsteina, John D. Fiski, for the Comorbidity and Cognition in Multiple Sclerosis (CCOMS) Study Group



Department of Internal Medicine, Max Rady College of Medicine, Rady Faculty of Health Sciences, University of Manitoba, Winnipeg, Canada Department of Community Health Sciences, Max Rady College of Medicine, Rady Faculty of Health Sciences, University of Manitoba, Winnipeg, Canada c Department of Clinical Health Psychology, Max Rady College of Medicine, Rady Faculty of Health Sciences, University of Manitoba, Winnipeg, Canada d Department of Radiology, Max Rady College of Medicine, Rady Faculty of Health Sciences, University of Manitoba, Winnipeg, Canada e Division of Diagnostic Imaging, Winnipeg Health Sciences Centre, Winnipeg, Canada f Neuroscience Research Program, Kleysen Institute for Advanced Medicine, Winnipeg Health Sciences Centre, Winnipeg, Canada g Department of Psychiatry, Max Rady College of Medicine, Rady Faculty of Health Sciences, University of Manitoba, Winnipeg, Canada h Department of Radiology, Faculty of Medicine, University of Calgary, Calgary, Canada i Departments of Psychiatry, Psychology & Neuroscience, and Medicine, Dalhousie University, Halifax, Canada b



Key words: Multiple sclerosis Cognition Comorbidity Anxiety Diabetes

Objective: To determine whether comorbid diabetes and hypertension are associated with cognition in multiple sclerosis (MS) after accounting for psychiatric comorbidities. Methods: Participants completed a structured psychiatric interview, the Hospital Anxiety and Depression Scale (HADS), a comorbidity questionnaire, and cognitive testing including the Symbol Digit Modalities Test (SDMT), California Verbal Learning Test (CVLT-II), Brief Visuospatial Memory Test-Revised (BVMT-R), and verbal fluency. Test scores were converted to age-, sex- and education-adjusted z-scores. We evaluated associations between diabetes and hypertension and the four cognitive z-scores using a multivariate linear model, adjusting for comorbid depression and anxiety disorders, psychotropic medications, disease-modifying therapies, smoking status and body mass index. Results: Of 111 participants, most were women (82.9%) with relapsing remitting MS (83.5%), of mean (SD) age 49.6 (12.7) years. Comorbidity was common; 22.7% participants had hypertension, 10.8% had diabetes, 9.9% had current major depression, and 9.9% had current anxiety disorders. Mean (SD) z-scores were: SDMT −0.66 (1.15), CVLT-II −0.43 (1.32), BVMT-R −0.49 (1.07) and fluency −0.59 (0.86). Diabetes (p = 0.02) and anxiety disorder (p = 0.02) were associated with cognitive function overall. Diabetes was associated with lower BVMT-R (β = −1.18, p = 0.0015) and fluency (β = −0.63, p = 0.037) z-scores. Anxiety was associated with lower SDMT (β = -1.07, p = 0.0074) z-scores. Elevated anxiety symptoms (HADS-A ≥ 11) were associated with lower z-scores on the SDMT and CVLT-II. Conclusion: Comorbidities, including diabetes and anxiety, are associated with cognitive dysfunction in MS. Their presence may contribute to the heterogeneous pattern of impairments seen across individuals and they may represent targets for improved management of cognitive symptoms.

1. Introduction The prevalence of cognitive impairment in MS ranges from 40–70%, and the patterns of cognitive impairment are heterogeneous across individuals (Rocca et al., 2015). Cognitive impairment is associated with difficulty maintaining employment, and impaired social roles (Goverover et al., 2015). Effective pharmacologic therapies for cognitive impairment in MS are lacking (He et al., 2013). Although successful

trials of cognitive rehabilitation have emerged in recent years, questions remain regarding their accessibility, effectiveness and functional impacts (Chiaravalloti et al., 2013). A need exists for greater understanding of potentially modifiable factors that may affect the risk of cognitive impairment and explain the heterogeneity of cognitive impairment in MS. Comorbid conditions may represent such modifiable factors. Depression is associated with cognitive impairment in MS (Arnett et al.,

Corresponding author at: Health Sciences Center, GF-543, 820 Sherbrook Street, Winnipeg, MB R3A 1R9, Canada. E-mail address: (R.A. Marrie). Received 1 June 2018; Received in revised form 18 October 2018; Accepted 22 October 2018 2211-0348/ © 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (

Multiple Sclerosis and Related Disorders 27 (2019) 164–170

R.A. Marrie et al.

2001), but less attention has been paid to the effects of other comorbidities. Notably, up to half of the MS population is affected by vascular comorbidity, including diabetes, hypertension, hyperlipidemia and ischemic heart disease (Marrie et al., 2010); hypertension is the third most prevalent comorbidity in MS (Marrie et al., 2015b). The incidence of vascular comorbidities is also rising in the MS population (Marrie et al., 2016), making it important to understand their impacts on outcomes. Hypertension and diabetes are associated with accelerated ambulatory and visual disability progression (Marrie et al., 2010,2011), and increased mortality in MS (Marrie et al., 2015a). However, the effects of vascular comorbidities on cognition in MS are unknown. In the general population, vascular conditions including diabetes and hypertension adversely affect cognition (van den Berg et al., 2009); although no consistent findings have been reported for hyperlipidemia. We aimed to determine whether vascular comorbidity in the form of diabetes and hypertension were associated with impaired cognitive function in MS. We hypothesized that cognitive function would be lower in persons with MS who had comorbid vascular conditions, as compared to individuals without such comorbidities and that the impacts of vascular comorbidities would be seen even after accounting for the presence of comorbid psychiatric conditions.

course and current disease-modifying therapy used (if any) from medical records. All other medications were recorded by patient interview. For this analysis, medications of interest were those that might affect cognition including psycholeptics (e.g. anxiolytics, antipsychotics), psychoanaleptics (e.g. antidepressants), anticholinergics and opioids. A certified neurologist (RAM or JMM) assessed neurologic disability using the Expanded Disability Status Scale (EDSS) (Kurtzke, 1983). 2.4. Vascular comorbidity We focused on hypertension and diabetes because of their associations with disability progression in MS (Marrie et al., 2010), and their effects on cognition in the general population (van den Berg et al., 2009). Participants reported these comorbidities using a validated questionnaire developed and tested in two MS populations (Horton et al., 2010). They reported if a physician had diagnosed the comorbidity, and if yes, the year of diagnosis and whether the condition was currently treated. We augmented the information provided by questionnaire with additional assessments. Blood pressure was measured once in the seated position using an automatic blood pressure machine. We collected a serum sample to measure a hemoglobin A1c (HbA1c). We classified participants as hypertensive (any of self-reported physician-diagnosed hypertension, use of hypertensive medications, measured BP > 140/90), or not hypertensive (Weinstein et al., 2015). We classified participants as having diabetes (any of self-reported physician-diagnosed diabetes, use of medications for diabetes, HbA1c > 6.5%(American Diabetes Association, 2011)) or not. We also classified participants as having hyperlipidemia (self-reported physician-diagnosed hyperlipidemia or use of lipid-lowering medications) or not. As few participants with any of these conditions reported being untreated, we did not pursue analyses stratifying these conditions by treatment status.

2. Methods 2.1. Participants The source population was participants in a longitudinal study of the effects of psychiatric comorbidity on immune-mediated inflammatory diseases (the ‘IMID’ study) that recruited 255 persons with definite MS (Polman et al., 2011) from the province-wide Winnipeg MS Clinic (Marrie et al., 2018). The present study enrolled a subgroup of 111 IMID study participants aged ≥18 years, with adequate knowledge of the English language who underwent a study visit between September 2016 and July 2017. The sample was restricted by availability of funding for extended testing and included cognitive testing likely to be affected by diabetes or hypertension. Exclusion criteria included comorbid brain tumors or neurodegenerative disorders. Most studies of cognition in MS exclude persons with comorbidities but we did not do so because comorbidities were the exposures of interest, and the burden of comorbidity is large in a representative MS population. All participants provided informed consent. The University of Manitoba Health Research Ethics Board approved the study. As described in detail elsewhere (Marrie et al., 2018), participants completed questionnaires, and underwent standardized clinical assessments conducted by trained personnel.

2.5. Psychiatric comorbidity Psychiatric comorbidity is common in MS, and has been reported to affect cognition (Arnett et al., 1999). In the general population, one study suggested that hypertension adversely affected cognition only in the presence of depression (Scuteri et al., 2011). Therefore, we considered it important to account for any possible effects of psychiatric comorbidity. At initial enrollment in the IMID study, the Structured Clinical Interview for DSM-IV (SCID) (First et al., 2002) was used to establish current diagnoses of depression (yes/no) and anxiety disorders (yes/no). At enrollment in the present study, SCID current episode modules for depression and anxiety disorders were used to identify new diagnoses and to confirm continuing diagnoses. For our analysis, we classified diagnoses of major depression present at enrollment as current depression, and diagnoses of generalized anxiety disorder and panic disorder present at enrollment as current anxiety. While generalized anxiety disorder is persistent and panic disorder is episodic, we combined these disorders because of their frequent comorbidity, and because both affect cognition in the general population (Millan et al., 2012). We did not include phobias in the absence of evidence that they affect cognitive function in the general population. The Hospital Anxiety and Depression Scale (HADS) uses 14 items to assess the severity of current symptoms of anxiety (HADS-A) and depression (HADS-D) in medically ill populations (Zigmond and Snaith, 1983), irrespective of whether criteria for a formal diagnosis of anxiety or depressive disorders are met. A cut-point of 8 has been proposed for the HADS in MS although recent studies have suggested that a threshold of at least 9–10 is preferable for the HADS-A to ensure adequate specificity (Watson et al., 2014; Patten et al., 2015; Marrie et al., 2017). Thus, for our primary analysis, we chose to categorize the HADS at the more specific cut-points of ≥11, which indicate severe, clinically meaningful symptoms of anxiety and depression (Zigmond and Snaith, 1983).

2.2. Sociodemographic information and health behaviours Participants reported sex, date of birth, ethnicity, highest level of education attained, annual household income, and marital status. Highest level of education completed was reported as elementary school, junior high school, high school diploma/General Education Diploma (GED), college, technical/trade, university bachelor's degree, university master's degree, university doctorate or other. Annual household income was reported as < $15,000, $15,000–29,999, $30,000–49,999, $50,000–100,000, > $100,000 or ‘I do not wish to answer’. We classified participants who reported ever smoking ≥100 cigarettes as smokers (Grant et al., 2004). Participants reported current smoking status as not at all, some days, or every day. We calculated body mass index (BMI, kg/m2) based on height and weight measured at the study visit. 2.3. Clinical characteristics We extracted age at symptom onset, age at MS diagnosis, clinical 165

Multiple Sclerosis and Related Disorders 27 (2019) 164–170

R.A. Marrie et al.

2.6. Cognition (Outcome measures)

Table 1 Cohort demographic and clinical characteristics.

We selected validated neuropsychological assessments that examined the cognitive domains most often or most severely affected in MS, diabetes and hypertension (van den Berg et al., 2009; Genova et al., 2009). These included all those recommended for the Brief International Cognitive Assessment for Multiple Sclerosis (BICAMS)(Langdon et al., 2012) as well as assessments examining most cognitive domains addressed via the Minimal Assessment of Cognitive Function in MS (MACFIMS) (Benedict et al., 2002): information processing speed, verbal learning and memory, visual learning and memory, and verbal fluency/executive ability. We assessed information processing speed using the Symbol Digit Modalities Test (SDMT) (Smith 2002), verbal learning and memory using the California Verbal Learning Test (CVLT-II; Trial 1–5 total recall score) (Delis et al., 2000), visual learning and memory using the Brief Visuospatial Memory Test-Revised (BVMT-R; summed recall score for all three learning trials) (Benedict and Hopkins 2001), and language and executive abilities using tests of verbal fluency (letter and animal categories) (Strauus et al., 2006). Raw test scores were converted to age, sex and education-adjusted z-scores using Canadian regression-based norms (Berrigan et al., 2014; Walker et al., 2017). Z-scores of ≤−1.5 were classified as impaired. The Wechsler Test of Adult Reading (WTAR) (The Psychological Corporation, 2001) was included to provide an age-, sex-, education-, and ethnicity-adjusted Full Scale IQ estimate of premorbid intelligence, and was used to characterize the sample. 2.7. Statistical analysis We summarized categorical variables using frequency (percent) and continuous variables using mean (standard deviation [SD]) or median (interquartile range [IQR]) as appropriate. Univariate analyses used chi-square tests, student's t-tests, and Wilcoxon or Kruskal–Wallis tests as appropriate. Multivariate regression is useful when examining patterns of differences between groups for a set of variables. It is particularly useful when the set of variables are at least moderately correlated and it reduces type I errors related to multiple hypothesis tests. Therefore, we evaluated the association between the independent variables of interest (hypertension, diabetes, hyperlipidemia) and cognitive function using multivariate regression, in which all of the z-scores for the cognitive measures were dependent variables. If a statistically significant global association was identified between an independent variable of interest and cognition, we explored this further using linear models which included only one z-score as the dependent variable. Non-significant global associations were not examined further. Covariates included SCID diagnosis of anxiety (yes vs. no [reference]), SCID diagnosis of depression (yes vs. no [reference]), age at symptom onset (continuous), BMI (continuous), smoking (past, current, never [reference]), use of MS disease-modifying therapies (yes vs. no [reference]), and use of psychotropic therapies (yes vs. no [reference]). We did not include age, sex or education as covariates since these were accounted for in the cognitive test z-scores. We did not include ethnicity given the low number of individuals who identified as non-white. We did not include disability status as covariate to prevent overadjustment bias(Schisterman et al., 2009), because we considered it to lie on the causal pathway between the independent variables of interest and cognitive function(Marrie et al., 2010; Zhang et al., 2016; Tettey et al., 2017; McKay et al., 2018). We tested for interactions between diabetes and hypertension, and between depression and hypertension based on prior findings in the literature(Scuteri et al., 2011). We assessed model assumptions (multivariate normality, linearity of relationships between dependent variables, and homogeneity of variance and covariance) using standard methods. We report partial eta-squared (np2), the proportion of variance explained after accounting for other independent variables, as a measure of effect size, which we interpreted as small (0.01), medium (0.09) or large (0.25)(Cohen 1992).



N Age, yr mean (SD) Sex, n (%) Male Female Ethnicity, n (%) Caucasian Other Missing Education, n (%) Elementary school Middle school High School/ GED College Technical/Trade Bachelor degree Master's degree Doctoral degree Annual income, n (%) < $15,000 $15,000–29,999 $30,000–49,999 $50,000–100,000 > $100,000 I do not wish to answer MS Characteristics Age at MS onset, years, mean (SD) Age at MS diagnosis, years, mean (SD) Current course, n (%) Relapsing remitting Secondary progressive Primary progressive Possible/uncertain EDSS, median (p25-p75) Any disease-modifying therapy, n (%) Any psychotropic medication, n (%)

111 49.6 (12.7) 19 (17.1) 92 (82.9) 89 (80.9) 21 (19.1) 1 1 (0.9) 4 (3.6) 31 (27.9) 34 (30.6) 13 (11.7) 25 (22.5) 2 (1.8) 1 (0.9) 13 (11.8) 7 (6.4) 14 (12.7) 38 (34.5) 27 (24.5) 11 (9.9) 29.3 (10.5) 34.9 (10.1) 91 (83.5) 12 (11.0) 6 (5.5) 2 3.5 (2.5–4.0) 64 (57.7) 66 (59.5)

EDSS = Expanded Disability Status Scale.

We also conducted complementary analyses which (i) substituted the dichotomized HADS scores for current diagnoses of depression and anxiety disorder; (ii) classified individuals as hypertensive only if their measured BP was > 140/90. We applied a Benjamin–-Hochberg correction for multiple comparisons, with a false discovery rate of 0.05. Statistical analyses were performed using SAS V9.4 (SAS Institute Inc., Cary, NC). 3. Results Of the 111 participants, most were women with relapsing remitting MS and moderate or greater levels of disability (Table 1). Of the vascular comorbidities, the most common was hypertension (54%), followed by hyperlipidemia (21%). Of the 25 participants who reported physician-diagnosed hypertension, 24 (96.0%) were taking medications for hypertension. Among participants with hypertension according to our study definition, the median (IQR) systolic blood pressure was 138.5 (127.5–148.5) and diastolic blood pressure was 79 (70.5–84.0), whereas among participants without hypertension the median (IQR) systolic blood pressure was 114.0 (107.0–122.0) and diastolic blood pressure was 68.0 (62.0–72.0). Among participants with diabetes, the mean (SD) HbA1c was 7.45% (0.96) with a minimum value of 6.5%, whereas among participants without diabetes the mean (SD) HbAc1 was 5.44% (0.35). All but one participant who reported physician-diagnosed diabetes was taking medications for diabetes. Of those who reported physician-diagnosed hyperlipidemia, 16 (76.2%) were taking anti-lipid agents. Based on the SCID, a small proportion of participants met diagnostic criteria for current major depression (10%) or anxiety disorder (10%) (Table 2). 166

Multiple Sclerosis and Related Disorders 27 (2019) 164–170

R.A. Marrie et al.

Table 2 Frequency of comorbidity and health behaviours (n = 111). Comorbidity


Psychiatric comorbidity Current anxiety disorder (SCID), n (%) Current generalized anxiety disorder, n (%) Current panic disorder, n (%) Current major depressive episode (SCID), n (%) HADS-D ≥ 11, n (%) HADS-A ≥ 11, n (%) HADS-D ≥ 8, n (%) HADS-A ≥ 9, n (%) Vascular comorbidity Hypertension (self-reported physician diagnosis), n (%) Hypertension (self-reported physician diagnosis, measured blood pressure and medication use), n (%) Hypertension (measured blood pressure), n (%) Hyperlipidemia (self-reported physician diagnosis), n (%) Hyperlipidemia (self-reported physician diagnosis and medications), n (%) Diabetes (self-reported physician diagnosis), n (%) Diabetes (self-reported physician diagnosis, medications and measured HbA1c), n (%) Health behaviours Current smoker, n (%) BMI (kg/m2), mean (SD) BMI ≤ 25 BMI > 25 and < 30, n (%) BMI > 30, n (%)

Table 3 Cognitive function. Cognitive measure

Z-scorea Mean (SD)

Impairedb N (%)

Symbol Digit Modalities Test California Verbal Learning Test-II Brief Visuospatial Memory Test-Revised Verbal Fluency

−0.66 −0.43 −0.49 −0.59

29 21 18 17

a b

(1.15) (1.32) (1.07) (0.86)

11 (9.9) 8 (7.2) 5 (4.5) 11 (9.9) 11 (9.9) 16 (14.4) 29 (26.1) 28 (25.2) 25 (22.7) 60 (54.1) 29 (26.1) 21 (18.9) 23 (20.7) 9 (8.1) 12 (10.8) 21 (18.9) 28.9 (6.3) 31 (27.9) 37 (33.3) 43 (38.7)

Table 4 Association of independent variables with cognitive function overall (MANCOVA).

(26.1) (18.9) (16.2) (15.3)

Age, sex and education-adjusted z-scores; mpaired = z-score ≤−1.5

Based on the WTAR, estimated premorbid intelligence fell in the average range (mean [SD] 106.1 [8.4]). Mean z-scores were lowest for the SDMT, followed by fluency (Table 3). When we dichotomized these adjusted z-scores as impaired or unimpaired, the frequency of impairment was highest for the SMDT, followed by the CVLT-II, BMVT-R, and verbal fluency tests. The cognitive measures correlated moderately with each other (Table e1), supporting the use of a multivariate model. Diabetes (F[4,97 ]= 3.06, p = 0.02), a current diagnosis of anxiety (F[4,97] = 3.02, p = 0.02) and BMI (F[4,97] = 2.71, p = 0.035) were associated with cognitive function overall (Table 4). We did not identify any interactions between diabetes and hypertension, or between depression and hypertension. When we examined the associations of diabetes, current anxiety disorder, and BMI with specific cognitive domains in models which remained adjusted for all other factors, diabetes was associated with lower performance on the BVMT-R and tests of fluency; anxiety was associated with lower performance on the SDMT; and higher BMI was associated with better performance on the BVMT-R and tests of fluency (Table 5). When we substituted severe symptoms of depression and anxiety for current diagnoses of depression and anxiety, diabetes (F[4,97] = 2.66, p = 0.037), anxiety symptoms [HADS-A score ≥11] (F[4,97] = 4.80, p = 0.0014), and higher BMI (F[4,97) = 3.54, p = 0.0097) were associated with cognitive function overall (Table e-2). When we examined the associations of these factors with specific cognitive domains in models which remained adjusted for all other factors, a similar pattern of results emerged: diabetes was associated with lower performance on the BVMT-R and tests of fluency; anxiety was associated with lower performance on the SDMT and CVLT-II; and higher BMI was associated


Pillai's trace F-value


Diabetes Hypertension Hyperlipidemia Major depression (SCID) Anxiety (SCID) BMI Current smoking Age at symptom onset Any disease-modifying therapy Any psychotropic therapy

3.06 0.96 0.68 1.37 3.02 2.71 0.91 0.43 1.27 0.30

0.02 0.44 0.61 0.25 0.02 0.035 0.46 0.79 0.29 0.88

bold = statistically significant after Benjamini–Hochberg correction.

with better performance on the BVMT-R and tests of fluency (Table e3). When we classified individuals as hypertensive only if their measured BP was elevated, the pattern of findings did not change. 4. Discussion We evaluated the association between several common comorbidities and cognitive function in persons with established MS. Diabetes, the presence of anxiety disorder, and elevated anxiety symptoms were all associated with lower cognitive function, after accounting for age, sex, education, smoking status, and BMI. However, the effects of diabetes and anxiety were seen on differing cognitive domains. We did not find that hypertension was associated with cognitive function nor did we find an association with depression. Our findings suggest that at least some of the widely recognized heterogeneity in cognitive outcomes in MS may be due to comorbid conditions and symptoms, and that these associations are complex. Notably, diabetes was associated with reduced cognitive function on tests of visual learning and memory, as well as verbal fluency, but hypertension and hyperlipidemia were not. Diabetes did not affect verbal memory, but visual learning and memory are impaired more often than verbal learning and memory in MS(Chiaravalloti and DeLuca 2008), and verbal and visual memory differ in their underlying neuroanatomy(Dalton et al., 2015). The number of individuals with 167

Multiple Sclerosis and Related Disorders 27 (2019) 164–170

R.A. Marrie et al.

Table 5 Associations of comorbidity with cognitive functiona. Variable

SDMTb β (SE)






BVMT-Rd β (SE)



Fluencye β (SE)



Diabetes Anxiety SCID BMI

−0.55 (0.40) −1.07 (0.39) 0.027 (0.02)

0.017 0.07 0.016

0.18 0.0074 0.19

−0.75 (0.46) −0.70 (0.44) −0.0096 (0.023)

0.034 0.02 0.002

0.10 0.12 0.68

−1.18 (0.36) −0.38 (0.35) 0.042 (0.018)

0.05 0.002 0.053

0.0015 0.28 0.023

−0.63 (0.30) 0.18 (0.29) 0.038 (0.015)

0.04 0.003 0.06

0.037 0.53 0.013

a Adjusted for hypertension, hyperlipidemia, SCID major depression, age at symptom onset, use of any DMT, use of any psychotropic medication, current smoking status. b Total variance explained = 0.12; c Total variance explained = 0.15; d Total variance explained = 0.17; e Total variance explained = 0.12; np2 = the proportion of variance explained after accounting for other independent variables; bold = statistically significant after Benjamini–Hochberg correction.

diabetes was small thus these results should be viewed cautiously. We were unable to identify other studies that have evaluated this association in the MS population. A systematic review of cross-sectional and longitudinal studies that assessed the impact of vascular risk factors on cognition in individuals without dementia found that diabetes and hypertension were consistently associated with reduced cognitive function although not necessarily reaching the threshold for “impairment”(van den Berg et al., 2009). In our cohort, most were treated for their vascular conditions and this may have ameliorated some of the adverse effects of these conditions; however, we did not observe an effect of hypertension even among individuals with elevated blood pressures at the time of their study visit. Several factors may contribute to effects of vascular comorbidity on cognition in MS. Vascular comorbidity is associated with increased peripheral inflammation, and the presence of elevated inflammatory markers which, in turn, is associated with brain atrophy(Jefferson et al., 2007). The chronic hyperglycemia and hyperinsulinemia of diabetes may also induce molecular changes in vasopressin-secreting neurons which impair long-term potentiation in the hippocampus(Klein and Waxman, 2003). Diabetes may also alter blood vessel function, thereby increasing inflammatory responses in the brain(Launer, 2005). By contrast, while hypertension is associated with reduced cerebral blood flow and metabolism, as well as endothelial injury(Waldstein, 2003), some observational studies in older persons suggest that successful blood pressure control mitigates the effects of hypertension on cognition (Obisesan et al., 2008), and that individuals with treated hypertension may still experience less decline in cognition even when blood pressure is not fully controlled (Rouch et al., 2015). More recent studies suggest specific anti-hypertensives may even have benefits on cognition, independent of their effects on hypertension (Stuhec et al., 2017). We found that while two-thirds of our cohort was overweight or obese, higher BMI was associated with better cognitive performance, even though the magnitude of the effects were small. This finding was surprising given reported associations between obesity and greater disability in MS (Tettey et al., 2017). However, our findings are consistent with some recent studies of general population samples which have suggested that elevated BMI in mid to late life is associated with a lower risk of cognitive impairment (Suemoto et al., 2015; Kim et al., 2016). For example, in the Korean Longitudinal Study of Aging, adults aged 45 years and older with a baseline BMI ≥25 were 27% less likely to develop severe cognitive impairment than those of normal weight. (Kim et al., 2016) Those findings were particularly evident among women and individuals with lower cognitive function at baseline. In that study, being underweight also conferred an increased risk of cognitive impairment, a finding similar to that of a recent American study. (Xiang and An, 2015) Although the etiology of these associations is uncertain, this issue appears to warrant further longitudinal investigation in the MS population, in whom lower BMI may not necessarily be associated with better overall health.

We found that a current diagnosis of an anxiety disorder (generalized anxiety disorder or panic disorder), or elevated symptoms of anxiety (HADS-A score ≥11) even in the absence of a diagnosed anxiety disorder, were associated with reduced information processing speed. Elevated anxiety symptoms were additionally associated with lower verbal learning and memory. Previous studies in the general population have suggested that generalized anxiety disorder and panic disorder are often associated with changes in attention and memory.(Millan et al., 2012) Unlike our study however, prior studies of cognition in MS have focused only on symptoms of anxiety rather than anxiety disorders. In a study of 190 persons with relapsing remitting MS, state anxiety was associated with impaired attention and processing speed as measured by the SDMT and Paced Auditory Serial Addition Test (PASAT). (Goretti et al., 2014) In a retrospective chart review of 151 persons with MS who had completed the HADS and the MACFIMS,29 HADS-A scores were associated with lower scores on the PASAT, and immediate and delayed BVMT-R recall.(Morrow et al., 2015) Another study of 322 persons with MS, reported that anxiety as measured by the Depression Anxiety Stress Scales was associated with impaired memory and verbal fluency (Ribbons et al., Epub ahead of print). However, none of these studies accounted for the potential presence of comorbid physical health conditions as was done in our study. Future studies examining the role of psychiatric comorbidity in cognition in MS should assess anxiety and depression. While potential associations of depression with cognitive function in MS have been explored for some time, early studies typically failed to detect them(Moller et al., 1994). Later studies have suggested that severe depression in MS is associated with impaired working memory and executive function as well as reduced information processing speed (Arnett et al., 2001; Diamond et al., 2008; Niino et al., 2014; Golan et al., 2017), a pattern of impaired cognitive functioning similar to that seen in depressed individuals without MS. These effects appear to be greatest when symptoms of depression are more severe, and are more readily seen for cognitively demanding tasks (Diamond et al., 2008; Golan et al., 2017). However, at least one study suggested that the variance accounted for by depression was 6% or less (Golan et al., 2017), and most studies did not account for the presence of comorbid anxiety. While our findings are incongruent with some other studies of the impact of depression on cognition in MS, only 10% of our participants had major depression or severe depressive symptoms and this may have limited our ability to detect an effect. However, another very recent study also reported that, after accounting for stress and anxiety symptoms, depressive symptoms were no longer associated with cognitive function in an MS sample (Ribbons et al., Epub ahead of print). Study limitations should be considered. Among them, our study was cross-sectional and we cannot infer causality from the associations observed; anxiety for example, could be a response to perceived cognitive changes. The heterogeneity of the study population may have reduced our ability to detect effects, but the study sample was recruited to better reflect a typical MS clinic population than most prior studies of 168

Multiple Sclerosis and Related Disorders 27 (2019) 164–170

R.A. Marrie et al.

cognition in MS. We did not evaluate all possible cognitive domains but we purposefully assessed those most likely to be affected, based on the existing literature regarding MS, psychiatric disorders and vascular disorder populations. The number of participants with diabetes and severe anxiety symptoms was small; so replication of our findings in larger samples is needed. Nonetheless, the consistency of the effects of diabetes in our sample with those reported in the general population, and consistency with the effects of diabetes and psychiatric comorbidities on disability progression in MS are reassuring. The number of participants affected by severe depressive symptoms was relatively small, reducing our power to detect related effects, and we did not use biological measures of hyperlipidemia which may have more accurately classified participants. We were also unable to account for the duration of these conditions or use of treatments of the vascular conditions in our analyses. All of these represent possible avenues for further study. We found that anxiety and diabetes are associated with lower cognitive function in MS, and potentially affect differing domains of abilities. Cognitive outcomes in MS are heterogeneous, and comorbid conditions and symptoms may influence not only the presence or absence of cognitive impairment, but also the patterns of cognitive dysfunction observed. Our findings suggest that studies of cognition in MS should consider and account for comorbidities which are common in MS. Moreover, our findings suggest that improved treatment of anxiety and diabetes may improve cognitive symptoms.

Acknowledgements This study was funded by the Waugh Family Foundation MS Society of Canada Operating Grant(EGID 2639), CIHR (THC-135234), Crohn's and Colitis Canada, a Manitoba Research Chair (to RAM) and the Waugh Family Chair in Multiple Sclerosis (to RAM). Dr. Bernstein is supported in part by the Bingham Chair in Gastroenterology. Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.msard.2018.10.018. References American Diabetes Association, 2011. Standards of medical care in diabetes–2011. Diabetes Care. 34 (Suppl 1), S11–S61. Arnett, P.A., Higginson, C.I., Randolph, J.J., 2001. Depression in multiple sclerosis: relationship to planning ability. J. Int. Neuropsychol. Soc. 7 (6), 665–674. Arnett, P.A., Higginson, C.I., Voss, W.D., Bender, W.I., Wurst, J.M., Tippin, J.M., 1999. Depression in multiple sclerosis: relationship to working memory capacity. Neuropsychology 13 (4), 546–556. Benedict, R.H.B., Fischer, J.S., Archibald, C.J., Arnett, P.A., Beatty, W.W., Bobholz, J., et al., 2002. Minimal neuropsychological assessment of MS patients: a consensus approach. Clin. Neuropsychol. 16, 381–397. Benedict, R.H.B., Hopkins, B.J., 2001. Verbal learning test-revised/brief visuospatial memory test-revised professional manual supplement. Psychol. Assess. Resour. Berrigan, L.I., Fisk, J.D., Walker, L.A., Wojtowicz, M., Rees, L.M., Freedman, M.S., et al., 2014. Reliability of regression-based normative data for the oral symbol digit modalities test: an evaluation of demographic influences, construct validity, and impairment classification rates in multiple sclerosis samples. Clin. Neuropsychol. 28 (2), 281–299. Chiaravalloti, N.D., DeLuca, J., 2008. Cognitive impairment in multiple sclerosis. Lancet Neurol. 7 (12), 1139–1151. Chiaravalloti, N.D., Moore, N.B., Nikelshpur, O.M., DeLuca, J., 2013. An RCT to treat learning impairment in multiple sclerosis: the MEMREHAB trial. Neurology 81 (24), 2066–2072. Cohen, J., 1992. A power primer. Psychol. Bull. 112 (1), 155–159. Dalton, M.A., Hornberger, M., Piguet, O., 2015. Material specific lateralization of medial temporal lobe function: an fMRI investigation. Hum. Brain Mapp. 37 (3), 933–941. Delis, D.C., Kramer, J.H., Kaplan, E., Ober, B.A., 2000. California Verbal Learning Test Second Edition Adult Version Manual. The Psychological Corporation. Diamond, B.J., Johnson, S.K., Kaufman, M., Graves, L., 2008. Relationships between information processing, depression, fatigue and cognition in multiple sclerosis. Arch. Clin. Neuropsychol. 23 (2), 189–199. First, M., Gibbon, M., Spitzer, R., Williams, J., 2002. User's Guide for the Structured Clinical Interview for DSM-IV-TR Axis I Disorders - Research Version - (SCID-I For DSM-IV-TR, November 2002 Revision). Biometrics Research Department, New York State Psychiatric Institute, New York, New York. Genova, H.M., Sumowski, J.F., Chiaravalloti, N., Voelbel, G.T., Deluca, J., 2009. Cognition in multiple sclerosis: a review of neuropsychological and fMRI research. Front Biosci. (Landmark Ed) 14, 1730–1744. Golan, D., Doniger, G.M., Wissemann, K., Zarif, M., Bumstead, B., Buhse, M., et al., 2017. The impact of subjective cognitive fatigue and depression on cognitive function in patients with multiple sclerosis. Mult. Scler. J Epub ahead of print(0): 1352458517695470. Goretti, B., Viterbo, R.G., Portaccio, E., Niccolai, C., Hakiki, B., Piscolla, E., et al., 2014. Anxiety state affects information processing speed in patients with multiple sclerosis. Neurol. Sci. 35 (4), 559–563. Goverover, Y., Strober, L., Chiaravalloti, N., DeLuca, J., 2015. Factors that moderate activity limitation and participation restriction in people with multiple sclerosis. Am. J. Occup. Ther. 69 (2) 6902260020p1-9. Grant, B.F., Hasin, D.S., Chou, S., Stinson, F.S., Dawson, D.A., 2004. Nicotine dependence and psychiatric disorders in the United States: results from the national epidemiologic survey on alcohol and related conditions. Arch. Gen. Psychiatry 61 (11), 1107–1115. He, D., Zhang, Y., Dong, S., Wang, D., Gao, X., Zhou, H., 2013. Pharmacological treatment for memory disorder in multiple sclerosis. Cochrane Database Syst. Rev. 12 CD008876. Horton, M., Rudick, R.A., Hara-Cleaver, C., Marrie, R.A., 2010. Validation of a self-report comorbidity questionnaire for multiple sclerosis. Neuroepidemiology 35 (2), 83–90. Jefferson, A.L., Massaro, J.M., Wolf, P.A., Seshadri, S., Au, R., Vasan, R.S., et al., 2007. Inflammatory biomarkers are associated with total brain volume: the Framingham Heart Study. Neurology 68 (13), 1032–1038. Kim, S., Kim, Y., Park, S.M., 2016. Body mass index and decline of cognitive function. PLoS One 11 (2), e0148908. Klein, J.P., Waxman, S.G., 2003. The brain in diabetes: molecular changes in neurons and their implications for end-organ damage. Lancet Neurol. 2, 548–554. Kurtzke, J.F., 1983. Rating neurologic impairment in multiple sclerosis: an Expanded Disability Status Scale (EDSS). Neurology 33, 1444–1452. Langdon, D.W., Amato, M.P., Boringa, J., Brochet, B., Foley, F., Fredrikson, S., et al., 2012. Recommendations for a brief international cognitive assessment for multiple

Author contributions Ruth Ann Marrie, Chase Figley, Jennifer Kornelsen, James Bolton, Lesley Graff, Erin Mazerolle, Charles Bernstein, James Marriott and John Fisk designed the study and obtained funding for the study. Ruth Ann Marrie, John Fisk and Ronak Patel developed the analytical plan. Ruth Ann Marrie drafted the manuscript. All authors assisted in interpretation of the data and revised the manuscript. Disclosures Ruth Ann Marrie receives research funding from: CIHR, Research Manitoba, Multiple Sclerosis Society of Canada, Multiple Sclerosis Scientific Foundation, Crohn's and Colitis Canada, National Multiple Sclerosis Society, CMSC. Ronak Patel receives research funding from the Workers Compensation Board of Manitoba. Chase Figley receives research funding from the Brain Canada Foundation, MS Society of Canada, Natural Sciences and Engineering Research Council of Canada, and Health Sciences Centre Foundation. Jennifer Kornelsen receives research funding from the MS Society of Canada, University of Manitoba and Health Sciences Centre Foundation. James Bolton receives research funding from CIHR, Brain and Behavior Research Foundation and the MS Society of Canada. Lesley Graff receives research funding from CIHR, the MS Society of Canada and the Health Sciences Centre Foundation. Erin Mazerolle received fellowship funding from NSERC and Alberta Innovates-Health Solutions. James Marriott has conducted trials for Biogen Idec and Roche, and receives research funding from the MS Society of Canada. Charles Bernstein has consulted to Abbvie Canada, Ferring Canada, Janssen Canada, Pfizer Canada, Shire Canada, Takeda Canada, and Napo Pharmaceuticals and has consulted to Mylan Pharmaceuticals. He has received unrestricted educational grants from Abbvie Canada, Janssen Canada, Shire Canada, and Takeda Canada. He has been on speaker's bureau of Abbvie Canada, Ferring Canada and Shire Canada. John Fisk receives research grant support from the Canadian Institutes of Health Research, the National Multiple Sclerosis Society, the Multiple Sclerosis Society of Canada, the Nova Scotia Health Authority Research Fund and the Dalhousie Medical Research Fund. 169

Multiple Sclerosis and Related Disorders 27 (2019) 164–170

R.A. Marrie et al. sclerosis (BICAMS). Mult. Scler. 18 (6), 891–898. Launer, L.J., 2005. Diabetes and brain aging: epidemiologic evidence. Curr. Diab. Rep. 5, 59–63. Marrie, R.A., Cutter, G., Tyry, T., 2011. Substantial adverse association of visual and vascular comorbidities on visual disability in multiple sclerosis. Mult. Scler. 17 (12), 1464–1471. Marrie, R.A., Elliott, L., Marriott, J., Cossoy, M., Blanchard, J., Leung, S., et al., 2015a. Effect of comorbidity on mortality in multiple sclerosis. Neurology 85 (3), 240–247. Marrie, R.A., Fisk, J., Tremlett, H., Wolfson, C., Warren, S., Blanchard, J., et al., 2016. Differing trends in the incidence of vascular comorbidity in MS and the general population. Neurol.: Clin. Pract. 6 (2), 120–128. Marrie, R.A., Graff, L.A., Walker, J.R., Fisk, J.D., Patten, S.B., Hitchon, C.A., et al., 2018. A prospective study of the effects of psychiatric comorbidity in immune-mediated inflammatory disease: rationale, protocol and participation. JMIR Res. Protoc. 7 (1), e15. Marrie, R.A., Reider, N., Cohen, J., Stuve, O., Trojano, M., Sorensen, P.S., et al., 2015b. A systematic review of the incidence and prevalence of comorbidity in multiple sclerosis: overview. Mult. Scler. J. 21 (3), 263–281. Marrie, R.A., Rudick, R., Horwitz, R., Cutter, G., Tyry, T., Campagnolo, D., et al., 2010. Vascular comorbidity is associated with more rapid disability progression in multiple sclerosis. Neurology 74 (13), 1041–1047. Marrie, R.A., Zhang, L., Lix, L.M., Graff, L.A., Walker, J.R., Fisk, J.D., et al., 2017. The validity and reliability of screening measures for depression and anxiety disorders in multiple sclerosis. Mult. Scler. Relat. Disord. 20, 9–15. McKay, K.A., Tremlett, H., Fisk, J.D., Zhang, T., Patten, S.B., Kastrukoff, L., et al., 2018. Psychiatric comorbidity is associated with disability progression in multiple sclerosis. Neurology. Millan, M.J., Agid, Y., Brüne, M., Bullmore, E.T., Carter, C.S., Clayton, N.S., et al., 2012. Cognitive dysfunction in psychiatric disorders: characteristics, causes and the quest for improved therapy. Nat. Rev. Drug Discov. 11 (2), 141–168. Moller, A., Wiedemann, G., Rohde, U., Backmund, H., Sonntag, A., 1994. Correlates of cognitive impairment and depressive mood disorder in multiple sclerosis. Acta Psychiatr. Scand. 89 (2), 117–121. Morrow, S.A., Rosehart, H., Pantazopoulos, K., 2015. Anxiety and depressive symptoms are associated with worse performance on objective cognitive tests in MS. J. Neuropsychiatry Clin. Neurosci appineuropsych15070167. Niino, M., Mifune, N., Kohriyama, T., Mori, M., Ohashi, T., Kawachi, I., et al., 2014. Apathy/depression, but not subjective fatigue, is related with cognitive dysfunction in patients with multiple sclerosis. BMC Neurol. 14 (1), 3. Obisesan, T.O., Obisesan, O.A., Martins, S., Alamgir, L., Bond, V., Maxwell, C., et al., 2008. High blood pressure, hypertension, and high pulse pressure are associated with poorer cognitive function in persons aged 60 and older: the Third National Health and Nutrition Examination Survey. J. Am. Geriatr. Soc. 56 (3), 501–509. Patten, S.B., Burton, J.M., Fiest, K.M., Wiebe, S., Bulloch, A.G., Koch, M., et al., 2015. Validity of four screening scales for major depression in MS. Mult. Scler. J. 21 (8), 1064–1071. Polman, C.H., Reingold, S.C., Banwell, B., Clanet, M., Cohen, J.A., Filippi, M., et al., 2011. Diagnostic criteria for multiple sclerosis: 2010 Revisions to the McDonald criteria. Ann. Neurol. 69 (2), 292–302. Ribbons, K., Lea, R., Schofield, P.W., Lechner-Scott, J., 2017. Anxiety levels are

independently associated with cognitive performance in an Australian multiple sclerosis patient cohort (Epub ahead of print). J. Neuropsychiatry Clin. Neurosci. 29 (2), 128–134. Rocca, M.A., Amato, M.P., De Stefano, N., Enzinger, C., Geurts, J.J., Penner, I.-K., et al., 2015. Clinical and imaging assessment of cognitive dysfunction in multiple sclerosis. Lancet Neurol. 14 (3), 302–317. Rouch, L., Cestac, P., Hanon, O., Cool, C., Helmer, C., Bouhanick, B., et al., 2015. Antihypertensive drugs, prevention of cognitive decline and dementia: a systematic review of observational studies, randomized controlled trials and meta-analyses, with discussion of potential mechanisms. CNS Drugs 29 (2), 113–130. Schisterman, E.F., Cole, S.R., Platt, R.W., 2009. Overadjustment bias and unnecessary adjustment in epidemiologic studies. Epidemiology 20 (4), 488–495. Scuteri, A., Spazzafumo, L., Cipriani, L., Gianni, W., Corsonello, A., Cravello, L., et al., 2011. Depression, hypertension, and comorbidity: disentangling their specific effect on disability and cognitive impairment in older subjects. Arch. Gerontol. Geriatr. 52 (3), 253–257. Smith, A., 2002. Symbol Digit Modalities Test. Western Psychological Services. Strauus, E., Sherman, E., Spreen, O., 2006. A Compendium of Neuropsychological Tests: Administration, Norms, and Commentary. Oxford University Press, New York. Stuhec, M., Keuschler, J., Serra-Mestres, J., Isetta, M., 2017. Effects of different antihypertensive medication groups on cognitive function in older patients: a systematic review. Eur. Psychiatry 46 (Supplement C), 1–15. Suemoto, C.K., Gilsanz, P., Mayeda, E.R., Glymour, M.M., 2015. Body mass index and cognitive function: the potential for reverse causation. Int. J. Obes. 39 (9), 1383–1389. Tettey, P., Simpson, S., Taylor, B., Ponsonby, A.-L., Lucas, R.M., Dwyer, T., et al., 2017. An adverse lipid profile and increased levels of adiposity significantly predict clinical course after a first demyelinating event. J. Neurol. Neurosurg. Psychiatry 88 (5), 395–401. The Psychological Corporation (2001). The Wechsler Test of Adult Reading (WTAR). van den Berg, E., Kloppenborg, R.P., Kessels, R.P., Kappelle, L.J., Biessels, G.J., 2009. Type 2 diabetes mellitus, hypertension, dyslipidemia and obesity: a systematic comparison of their impact on cognition. Biochim. Biophys. Acta 1792 (5), 470–481. Waldstein, S.R., 2003. The relation of hypertension to cognitive function. Curr. Dir. Psychol. Sci. 12 (1), 9–12. Walker, L.A.S., Marino, D., Berard, J.A., Feinstein, A., Morrow, S.A., Cousineau, D., 2017. Canadian normative data for minimal assessment of cognitive function in multiple sclerosis. Can. J. Neurol. Sci. / J. Canadien des Sciences Neurologiques 1–9. Watson, T.M., Ford, E., Worthington, E., Lincoln, N.B., 2014. Validation of mood measures for people with multiple sclerosis. Int. J. MS Care 16 (2), 105–109. Weinstein, G., Maillard, P., Himali, J.J., Beiser, A.S., Au, R., Wolf, P.A., et al., 2015. Glucose indices are associated with cognitive and structural brain measures in young adults. Neurology 84 (23), 2329–2337. Xiang, X., An, R., 2015. Body weight status and onset of cognitive impairment among U.S. middle-aged and older adults.". Arch. Gerontol. Geriatr. 60 (3), 394–400. Zhang, T., Tremlett, H., Leung, S., Zhu, F., Kingwell, E., Fisk, J.D., et al., 2016. Examining the effects of comorbidities on disease-modifying therapy use in multiple sclerosis. Neurology 86 (14), 1287–1295. Zigmond, A.S., Snaith, R.P., 1983. The hospital anxiety and depression scale. Acta Psychiatr. Scand. 67, 361–370.


Profile for Manitoba Neuroscience Network

Marrie et al. (2019)  

Diabetes and anxiety adversely affect cognition in multiple sclerosis Ruth Ann Marriea, Ronak Patel, Chase R Figleyd, Jennifer Kornelsend, J...

Marrie et al. (2019)  

Diabetes and anxiety adversely affect cognition in multiple sclerosis Ruth Ann Marriea, Ronak Patel, Chase R Figleyd, Jennifer Kornelsend, J...