Merged document 5

Page 1

Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

23

Table 3: Substance Use Pattern Comparisons Substance Use Measure

Percent drinking days

Male Gay/Bisexual (n = 16)

Female

Heterosexual (n = 48)

Statistical Comparisons

Lesbian/ Bisexual (n = 46) M = 32.50 SD = 29.82 n = 40 M = 3.98 SD = 2.68 n = 40 M = 10.36 SD = 15.76 n = 43 M = 9.98 SD = 7.59 n = 46 37% (n = 16 of 43)

Heterosexual (n = 108)

Statistical Comparisons

M = 32.37 SD = 30.84 n = 97 M = 3.74 SD = 2.18 n = 87 M = 9.08 SD = 10.75 n = 102 M = 9.56 SD = 7.75 n = 108 46% (n = 47 of 102)

t(135) = -0.023

M = 50.04 SD = 37.73 n = 15 M = 3.69 SD = 1.61 n = 14 M = 14.45 SD = 13.39 n = 15 M = 11.19 SD = 6.25 n = 16 60% (n = 9 of 15)

M = 43.26 SD = 32.38 n = 43 M = 4.70 SD = 2.52 n = 39 M = 16.10 SD = 16.43 n = 45 M = 12.58 SD = 7.90 n = 48 64% (n = 29 of 45)

t(56) = -0.668

AUDIT Score 8+

69% (n = 11 of 16)

Z = -0.158

52% (n = 24 of 46)

56% (n = 61 of 108)

Z = -0.492

Lifetime drug use

86% (n = 12 of 14)

Z = 1.418

90% (n = 37 of 41)

78% (n = 79 of 101)

Z = 1.679+

Past Year Drug Use

86% (n = 12 of 14)

Z = 2.427*

59% (n = 22 of 37)

53% (n = 51 of 97)

Z = 0.715

Past Year Marijuana Only

17% (n = 2 of 12)

Z = -1.289

50% (n = 11 of 22)

39% (n = 20 of 51)

Z = 0.855

Past Month Drug Use

58% (n = 7 of 12)

71% (n = 34 of 48) 66% (n = 29 of 44) 49% (n = 21 of 43) 38% (n = 8 of 21) 33% (n = 14 of 42)

Z = 1.567

38% (n = 14 of 37)

33% (n = 31 of 93)

Z = 0.487

Mean drinks per drinking day Mean drinks per week AUDIT score Unsafe Drinking on AUDIT

t(51) = 1.390

t(58) = 0.352

t(62) = 0.642

Z = -0.309

t(63) = -0.506

t(143) = -0.568

t(152) = -0.312

Z =-0.984


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

Past Month Marijuana Only

29% (n = 2 of 7)

Prior SUD Treatment

23% (n = 3 of 13)

Prior Self-Help

25% (n = 3 of 12)

Currently Considering Treatment

8% (n = 1 of 13)

Other Suggested Treatment

8% (n = 1 of 13)

Trying To Cut Down

75% (n = 3 of 4)

Abstinence Goal

17% (n = 2 of 12)

36% (n = 5 of 14) 16% (n = 5 of 32) 13% (n = 4 of 32) 13% (n = 4 of 32) 16% (n = 5 of 32) 62% (n = 8 of 13) 14% (n = 4 of 29)

24

Z = -0.327

64% (n = 9 of 14)

65% (n = 20 of 31)

Z = -0.015

Z = 0.593

3% (n = 1 of 37)

16% (n = 15 of 93)

Z = -2.103*

Z = 1.01

8% (n = 3 of 37)

17% N = 16 of 92)

Z = -1.346

Z = -0.465

0

9% (n = 8 of 94)

Z = -1.831+

Z = -0.710

11% (n = 4 of 36)

10% (n = 9 of 94)

Z = 0.261

Z = 0.493

30% (n = 3 of 10)

83% (n = 24 of 29)

Z = -3.117**

Z = 0.237

3% (n = 1 of 35)

11% (n = 10 of 89)

Z = -1.477

*The percentages reported for drug use are the percentage of the group including missing data. The numbers reported are the numbers of respondents that reported drug use out of the number with data for that variable.


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

25

Table 4: Social Network and Social Support Comparisons Social Network Variable

Currently in a Romantic Relationship Current Relationship Length Less than 6 months 6 – 12 months 1 – 5 years 5 – 10 years 10 – 25 years 25+ years

Cohabiting With Partner

Gay/Bisexual (n = 16) 69% (n = 11 of 16) 9% (n = 1 of 11) 0% (n = 0) 36% (n = 4 of 11) 36% (n = 4 of 11) 9% (n = 1 of 11) 9% (n = 1 of 11) 82% (n = 9 of 11)

Married

36% (n = 4 of 11)

Children

21% (n = 3 of 14)

Number of people reporting 3 or 0 fewer important people (max of 5)

Male Heterosexual (n = 48)

Female Statistical Comparisons

Lesbian/ Bisexual (n = 46) 63% (n = 29 of 46) (n = 7 of 29) (n = 1 of 29) (n = 8 of 29) (n = 7 of 29) (n = 4 of 29) (n = 2 of 29)

Heterosexual (n = 108)

Statistical Comparisons

75% (n = 81 of 108) (n = 11 of 81) (n = 11 of 81) (n = 36 of 81) (n = 7 of 81) (n = 10 of 81) (n = 6 of 81)

Z = -1.503

65% (n = 31 of 48) (n = 5 of 31) (n = 7 of 31) (n = 12 of 31) (n = 2 of 31) (n = 3 of 31) (n = 2 of 31)

Z = 0.304

42% (n = 13 of 31) 33% (n = 10 of 30) 30% (n = 14 of 47) 0

Z = 2.275*

72% (n = 21 of 29)

58% (n = 47 of 81)

Z = 1.39

Z = 0.181

25% (n = 7 of 28)

35% (n = 28 of 81)

Z = -0.935

Z = -0.612

33% (n = 13 of 39)

24% (n = 25 of 103)

Z = 1.089

-----

5% (n = 2 of 40)

3% (n = 3 of 99)

Z = 0.565


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

26

Mean Percentage of Network Rated as Generally Supportive

M = 81.82 SD = 20.89 n = 11

M = 79.33 SD = 27.03 n = 30

t(39) = -0.275

M = 83.13 SD = 23.34 n = 32

M = 87.05 SD = 18.95 n = 88

t(118) = 0.940

More than Half of Network Rated Generally Supportive

91% (n = 10 of 11)

90% (n = 27 of 30)

Z = 0.087

91% (n = 29 of 32)

93% (n = 82 of 88)

Z = -0.470

Mean Percentage of Network Rated as Supportive of SUD Treatment More than Half of Network Rated as Supportive of SUD Treatment Mean Percentage of Network Rated as Moderate/Heavy Drinkers More than Half of Network Rated as Moderate/Heavy Drinkers Mean Percentage of Network Rated as Abstainers from Alcohol Percentage Reporting No Abstainers in Network

M = 47.27 SD = 41.25 n = 11

M = 43.33 SD = 36.42 n = 30

t(39) = -0.296

M = 53.13 SD = 40.52 n = 32

M = 58.86 SD = 39.23 n = 88

t(118) = 0.703

46% (n = 5 of 11)

43% (n = 13 of 30)

Z = 0.121

56% (n = 18 of 32)

59% (n = 52 of 88)

Z = -0.279

M = 41.82 SD = 35.16 n = 11

M = 38.67 SD = 31.04 n = 30

t(39) = -0.278

M = 39.39 SD = 25.24 n = 33

M = 35.45 SD = 26.73 n = 88

t(119) = -0.733

46% (n = 5 of 11)

33% (n = 10 of 30)

Z = 0.714

36% (n = 12 of 33)

26% (n = 23 of 88)

Z = 1.105

M = 18.18 SD = 20.89 n = 11

M = 25.33 SD = 28.25 n = 30

t(39) = 0.794

M = 13.94 SD = 16.19 n = 33

M = 25.23 SD = 25.82 n = 88

t(91) = 2.865**

46% (n = 5 of 11)

37% (n = 11 of 30)

Z = 0.511

49% (n = 16 of 33)

36% (n = 32 of 88)

Z = 1.214


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS Table 5: Perceived Social Support Comparisons Percentage of Network Rated as Generally Supportive M = 84.38 Entire Sample SD = 21.71 n = 160 M = 80.00 SD = 25.30 n = 41 M = 85.88 SD = 20.23 n = 119 M = 85.08 SD = 21.43 n =118 M = 82.38 SD = 22.61 n =42

Males Females Heterosexual LGB

27

Percentage of Network Rated as Supportive of SUD Treatment

Paired Sample t-test

M = 54.00 SD = 39.35 n = 160 M = 44.39 SD = 37.29 n = 41 M = 57.31 SD = 39.65 n = 119 M = 54.92 SD = 38.98 n =118 M = 51.43 SD = 40.76 n =42

t(159) = 9.373*** t(41) = 5.647*** t(118) = 7.562*** t(117) = 8.094*** t(41) = 4.679***


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

28

Table 6: Pearson Correlations Between Social Network Variables and Personal Substance Use

LGB Respondents Mean Percentage of Network Rated as Supportive of SUD Treatment r (43)= .368*

Mean Percentage of Network Rated as Moderate / Heavy Drinkers

Mean Percentage of Network Rated as Abstainers

Mean Percentage of Network Rated as Generally Supportive

r (44)= .081

r (44)= .070

r (120)= -.137

Heterosexual Respondents Mean Mean Percentage of Percentage of Network Rated Network Rated as Supportive as Moderate / of SUD Heavy Drinkers Treatment r (120)= .088 r (120)= .244**

r (42)= .114

r (43)= .039

r (43)= -.014

r (117)= -.067

r (117)= .013

r (117)= .292**

r (120)= .252** r (117)= -.210*

r (41)= -.138

r (41)= .313*

r (42)= .224

r (42)= -.039

r (104)= -.060

r (104)= -.054

r (104)= .118

r (104)= -.162

r (42)= -.155

r (42)= .170

r (43)= .020

r (43)= .076

r (120)= -.056

r (120)= .007

r (120)= .213*

r (120)= -.177+

r (10)= -.226

r (11)= -.516

r (11)= .540+

r (11)= -.145

r (36)= -.063

r (36)= -.001

r (36)= .205

r (36)= -.287+

r (5)= .368

r (5)= .414

r (5)= .169

r (5)= -.817+

r (17)= .035

r (17)= .064

r (17)= -.514*

r (17)= .461+

Mean Percentage of Network Rated as Generally Supportive

AUDIT Score

r (43)= -.108 r (42)= -.127

Percent Drinking Days Mean Drinks Per Drinking Day Mean Drinks Per Week Timeframe to Quit / Cut Down Days of Self-Help in Past Year

+ p < .10 *p < .05 ** p < .01

Mean Percentage of Network Rated as Abstainers


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS Table 7: Social Network Substance Use and Support by Gender and Sexual Orientation Variable Group Mean and Standard Statistics Deviation r ()= Percentage Male (n = 41) M = 80.00; SD = 25.30 t(159) = -1.54, p = .126 Generally Female (n = 120) M = 86.00; SD = 20.18 Supportive Heterosexual (n = 118) M = 85.08; SD = 21.43 t(159) = .59, p = .554 LGB (n = 43) M = 82.79; SD = 22.50 Percentage Male (n = 41) M = 44.39; SD = 37.29 t(159) = -1.84, p = .068 Supportive of Female (n = 120) M = 57.33; SD = 39.49 Treatment Seeking Heterosexual (n = 118) M = 54.91; SD = 38.98 t(159) = .47, p = .640 LGB (n = 43) M = 51.63; SD = 40.29 Percentage Male (n = 41) M = 39.51; SD = 31.78 t(160) = .60, p = .553 Rated as Female (n = 121) M = 36.53; SD = 26.29 Moderate/Heavy Drinkers Heterosexual (n = 118) M = 36.27; SD = 27.79 t(160) = -.76, p = .448 LGB (n = 44) M = 40.00; SD = 27.62 More than Half Moderate/Heavy Drinkers Percentage Male (n = 41) M = 23.41; SD = 26.42 t(160) = .284, p = .777 Rated as Female (n = 121) M = 36.53; SD = 26.29 Abstainers from Alcohol Heterosexual (n = 118) M = 25.25; SD = 26.33 t(160) = 2.878, p = .005 LGB (n = 44) M = 15.00; SD = 17.32 No Abstainers in Reported

29


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS Network

30


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

23

Table 3: Substance Use Pattern Comparisons Substance Use Measure

Percent drinking days

Male Gay/Bisexual (n = 16)

Female

Heterosexual (n = 48)

Statistical Comparisons

Lesbian/ Bisexual (n = 46) M = 32.50 SD = 29.82 n = 40 M = 3.98 SD = 2.68 n = 40 M = 10.36 SD = 15.76 n = 43 M = 9.98 SD = 7.59 n = 46 37% (n = 16 of 43)

Heterosexual (n = 108)

Statistical Comparisons

M = 32.37 SD = 30.84 n = 97 M = 3.74 SD = 2.18 n = 87 M = 9.08 SD = 10.75 n = 102 M = 9.56 SD = 7.75 n = 108 46% (n = 47 of 102)

t(135) = -0.023

M = 50.04 SD = 37.73 n = 15 M = 3.69 SD = 1.61 n = 14 M = 14.45 SD = 13.39 n = 15 M = 11.19 SD = 6.25 n = 16 60% (n = 9 of 15)

M = 43.26 SD = 32.38 n = 43 M = 4.70 SD = 2.52 n = 39 M = 16.10 SD = 16.43 n = 45 M = 12.58 SD = 7.90 n = 48 64% (n = 29 of 45)

t(56) = -0.668

AUDIT Score 8+

69% (n = 11 of 16)

Z = -0.158

52% (n = 24 of 46)

56% (n = 61 of 108)

Z = -0.492

Lifetime drug use

86% (n = 12 of 14)

Z = 1.418

90% (n = 37 of 41)

78% (n = 79 of 101)

Z = 1.679+

Past Year Drug Use

86% (n = 12 of 14)

Z = 2.427*

59% (n = 22 of 37)

53% (n = 51 of 97)

Z = 0.715

Past Year Marijuana Only

17% (n = 2 of 12)

Z = -1.289

50% (n = 11 of 22)

39% (n = 20 of 51)

Z = 0.855

Past Month Drug Use

58% (n = 7 of 12)

71% (n = 34 of 48) 66% (n = 29 of 44) 49% (n = 21 of 43) 38% (n = 8 of 21) 33% (n = 14 of 42)

Z = 1.567

38% (n = 14 of 37)

33% (n = 31 of 93)

Z = 0.487

Mean drinks per drinking day Mean drinks per week AUDIT score Unsafe Drinking on AUDIT

t(51) = 1.390

t(58) = 0.352

t(62) = 0.642

Z = -0.309

t(63) = -0.506

t(143) = -0.568

t(152) = -0.312

Z =-0.984


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

Past Month Marijuana Only

29% (n = 2 of 7)

Prior SUD Treatment

23% (n = 3 of 13)

Prior Self-Help

25% (n = 3 of 12)

Currently Considering Treatment

8% (n = 1 of 13)

Other Suggested Treatment

8% (n = 1 of 13)

Trying To Cut Down

75% (n = 3 of 4)

Abstinence Goal

17% (n = 2 of 12)

36% (n = 5 of 14) 16% (n = 5 of 32) 13% (n = 4 of 32) 13% (n = 4 of 32) 16% (n = 5 of 32) 62% (n = 8 of 13) 14% (n = 4 of 29)

24

Z = -0.327

64% (n = 9 of 14)

65% (n = 20 of 31)

Z = -0.015

Z = 0.593

3% (n = 1 of 37)

16% (n = 15 of 93)

Z = -2.103*

Z = 1.01

8% (n = 3 of 37)

17% N = 16 of 92)

Z = -1.346

Z = -0.465

0

9% (n = 8 of 94)

Z = -1.831+

Z = -0.710

11% (n = 4 of 36)

10% (n = 9 of 94)

Z = 0.261

Z = 0.493

30% (n = 3 of 10)

83% (n = 24 of 29)

Z = -3.117**

Z = 0.237

3% (n = 1 of 35)

11% (n = 10 of 89)

Z = -1.477

*The percentages reported for drug use are the percentage of the group including missing data. The numbers reported are the numbers of respondents that reported drug use out of the number with data for that variable.


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

25

Table 4: Social Network and Social Support Comparisons Social Network Variable

Currently in a Romantic Relationship Current Relationship Length Less than 6 months 6 – 12 months 1 – 5 years 5 – 10 years 10 – 25 years 25+ years

Cohabiting With Partner

Gay/Bisexual (n = 16) 69% (n = 11 of 16) 9% (n = 1 of 11) 0% (n = 0) 36% (n = 4 of 11) 36% (n = 4 of 11) 9% (n = 1 of 11) 9% (n = 1 of 11) 82% (n = 9 of 11)

Married

36% (n = 4 of 11)

Children

21% (n = 3 of 14)

Number of people reporting 3 or 0 fewer important people (max of 5)

Male Heterosexual (n = 48)

Female Statistical Comparisons

Lesbian/ Bisexual (n = 46) 63% (n = 29 of 46) (n = 7 of 29) (n = 1 of 29) (n = 8 of 29) (n = 7 of 29) (n = 4 of 29) (n = 2 of 29)

Heterosexual (n = 108)

Statistical Comparisons

75% (n = 81 of 108) (n = 11 of 81) (n = 11 of 81) (n = 36 of 81) (n = 7 of 81) (n = 10 of 81) (n = 6 of 81)

Z = -1.503

65% (n = 31 of 48) (n = 5 of 31) (n = 7 of 31) (n = 12 of 31) (n = 2 of 31) (n = 3 of 31) (n = 2 of 31)

Z = 0.304

42% (n = 13 of 31) 33% (n = 10 of 30) 30% (n = 14 of 47) 0

Z = 2.275*

72% (n = 21 of 29)

58% (n = 47 of 81)

Z = 1.39

Z = 0.181

25% (n = 7 of 28)

35% (n = 28 of 81)

Z = -0.935

Z = -0.612

33% (n = 13 of 39)

24% (n = 25 of 103)

Z = 1.089

-----

5% (n = 2 of 40)

3% (n = 3 of 99)

Z = 0.565


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

26

Mean Percentage of Network Rated as Generally Supportive

M = 81.82 SD = 20.89 n = 11

M = 79.33 SD = 27.03 n = 30

t(39) = -0.275

M = 83.13 SD = 23.34 n = 32

M = 87.05 SD = 18.95 n = 88

t(118) = 0.940

More than Half of Network Rated Generally Supportive

91% (n = 10 of 11)

90% (n = 27 of 30)

Z = 0.087

91% (n = 29 of 32)

93% (n = 82 of 88)

Z = -0.470

Mean Percentage of Network Rated as Supportive of SUD Treatment More than Half of Network Rated as Supportive of SUD Treatment Mean Percentage of Network Rated as Moderate/Heavy Drinkers More than Half of Network Rated as Moderate/Heavy Drinkers Mean Percentage of Network Rated as Abstainers from Alcohol Percentage Reporting No Abstainers in Network

M = 47.27 SD = 41.25 n = 11

M = 43.33 SD = 36.42 n = 30

t(39) = -0.296

M = 53.13 SD = 40.52 n = 32

M = 58.86 SD = 39.23 n = 88

t(118) = 0.703

46% (n = 5 of 11)

43% (n = 13 of 30)

Z = 0.121

56% (n = 18 of 32)

59% (n = 52 of 88)

Z = -0.279

M = 41.82 SD = 35.16 n = 11

M = 38.67 SD = 31.04 n = 30

t(39) = -0.278

M = 39.39 SD = 25.24 n = 33

M = 35.45 SD = 26.73 n = 88

t(119) = -0.733

46% (n = 5 of 11)

33% (n = 10 of 30)

Z = 0.714

36% (n = 12 of 33)

26% (n = 23 of 88)

Z = 1.105

M = 18.18 SD = 20.89 n = 11

M = 25.33 SD = 28.25 n = 30

t(39) = 0.794

M = 13.94 SD = 16.19 n = 33

M = 25.23 SD = 25.82 n = 88

t(91) = 2.865**

46% (n = 5 of 11)

37% (n = 11 of 30)

Z = 0.511

49% (n = 16 of 33)

36% (n = 32 of 88)

Z = 1.214


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS Table 5: Perceived Social Support Comparisons Percentage of Network Rated as Generally Supportive M = 84.38 Entire Sample SD = 21.71 n = 160 M = 80.00 SD = 25.30 n = 41 M = 85.88 SD = 20.23 n = 119 M = 85.08 SD = 21.43 n =118 M = 82.38 SD = 22.61 n =42

Males Females Heterosexual LGB

27

Percentage of Network Rated as Supportive of SUD Treatment

Paired Sample t-test

M = 54.00 SD = 39.35 n = 160 M = 44.39 SD = 37.29 n = 41 M = 57.31 SD = 39.65 n = 119 M = 54.92 SD = 38.98 n =118 M = 51.43 SD = 40.76 n =42

t(159) = 9.373*** t(41) = 5.647*** t(118) = 7.562*** t(117) = 8.094*** t(41) = 4.679***


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

28

Table 6: Pearson Correlations Between Social Network Variables and Personal Substance Use

LGB Respondents

AUDIT Score Percent Drinking Days Mean Drinks Per Drinking Day Mean Drinks Per Week Timeframe to Quit / Cut Down Days of Self-Help in Past Year

Mean Percentage of Network Rated as: Generally Supportive of Moderate / Supportive SUD Treatment Heavy Drinkers r (43)= -.108 r (43)= .368* r (44)= .081 r (42)= -.127 r (42)= .114 r (43)= .039

Heterosexual Respondents Abstainers r (44)= .070

Generally Supportive r (120)= -.137

Supportive of Moderate / SUD Treatment Heavy Drinkers r (120)= .088 r (120)= .244**

Abstainers

r (43)= -.014

r (117)= -.067

r (117)= .013

r (117)= .292**

r (120)= .252** r (117)= -.210*

r (41)= -.138

r (41)= .313*

r (42)= .224

r (42)= -.039

r (104)= -.060

r (104)= -.054

r (104)= .118

r (104)= -.162

r (42)= -.155

r (42)= .170

r (43)= .020

r (43)= .076

r (120)= -.056

r (120)= .007

r (120)= .213*

r (120)= -.177+

r (10)= -.226

r (11)= -.516

r (11)= .540+

r (11)= -.145

r (36)= -.063

r (36)= -.001

r (36)= .205

r (36)= -.287+

r (5)= .368

r (5)= .414

r (5)= .169

r (5)= -.817+

r (17)= .035

r (17)= .064

r (17)= -.514*

r (17)= .461+

+ p < .10 *p < .05 ** p < .01


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS Table 7: Differences in Alcohol and Drug Use by Sex and Sexual Orientation Variable Group Mean and Standard Statistics Deviation r ()= Percentage Male (n = 41) M = 80.00; SD = 25.30 t(159) = -1.54, p = .126 Generally Female (n = 120) M = 86.00; SD = 20.18 Supportive Heterosexual (n = 118) M = 85.08; SD = 21.43 t(159) = .59, p = .554 LGB (n = 43) M = 82.79; SD = 22.50 Percentage Male (n = 41) M = 44.39; SD = 37.29 t(159) = -1.84, p = .068 Supportive of Female (n = 120) M = 57.33; SD = 39.49 Treatment Seeking Heterosexual (n = 118) M = 54.91; SD = 38.98 t(159) = .47, p = .640 LGB (n = 43) M = 51.63; SD = 40.29 Percentage Male (n = 41) M = 39.51; SD = 31.78 t(160) = .60, p = .553 Rated as Female (n = 121) M = 36.53; SD = 26.29 Moderate/Heavy Drinkers Heterosexual (n = 118) M = 36.27; SD = 27.79 t(160) = -.76, p = .448 LGB (n = 44) M = 40.00; SD = 27.62 More than Half Moderate/Heavy Drinkers Percentage Male (n = 41) M = 23.41; SD = 26.42 t(160) = .284, p = .777 Rated as Female (n = 121) M = 36.53; SD = 26.29 Abstainers from Alcohol Heterosexual (n = 118) M = 25.25; SD = 26.33 t(160) = 2.878, p = .005 LGB (n = 44) M = 15.00; SD = 17.32 No Abstainers in Reported

29

Ryan Wheeler 10/20/2016 10:27 AM Deleted: Social Network Substance Use and Support by Gender and Sexual Orientation


Running Head: SUBSTANCE USE IN LGB SOCIAL NETWORKS Network

30


Page 11: [1] Deleted

Ryan Wheeler

12/9/16 11:24 AM

ANCOVAs were used to evaluate differences between groups based on sex and sexual orientation, using covariates selected in univariate correlational analyses described above. The drinking frequency (PDD) ANCOVA[1] revealed sex differences but no sexual orientation differences. Men reported more frequent drinking than women, but there was no interaction with sexual orientation. Page 11: [2] Comment [18]

Kelly E. Green

12/9/16 11:17 AM

Do we have tables for the ANCOVA data? If not, we need to either include the stats in the text of remove those portions and report on the t-tests and z-tests in the tables. At this point, we haven’t discussed any correlations, so this data may just need to be moved further down. Page 11: [3] Deleted

Ryan Wheeler

12/9/16 11:33 AM

More frequent drinking also was reported by those without children in the home. Page 11: [4] Deleted

Ryan Wheeler

12/9/16 11:31 AM

The drinking intensity (MDPDD) ANCOVA did not reveal differences between sexes or sexual orientations. The interaction between sex and sexual orientation was not significant. For mean drinks per week (MDPW), [2]ANCOVA showed differences between sexes, but not between sexual orientations or their interaction. The ANCOVA for AUDIT scores evidenced differences between sexes, but not between sexual orientations. The sex by sexual orientation interaction was not significant. Page 11: [5] Comment [19]

seu

12/9/16 11:17 AM

Couldn’t you just write the acronym since it’s been stated before? Page 11: [6] Comment [22]

Kelly E. Green

12/9/16 11:17 AM

We need to add z-test results for all of these % comparisons Page 11: [7] Comment [23]

Kelly E. Green

12/9/16 11:13 AM

Kelly E. Green

12/9/16 11:16 AM

We need to add the z test info here Page 11: [8] Comment [28]

We need a bit more summary of drug use pattern data.


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

1 Kelly E. Green 11/2/2016 8:51 PM Formatted: Left Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt Kelly E. Green 11/2/2016 8:49 PM Formatted

Substance Use in Social Networks of Lesbian, Gay, Bisexual, and Heterosexual Worried Drinkers

Kelly E. Green 11/2/2016 8:36 PM Deleted: : Comparisons

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

Kelly E. Green (1,2), Barbara S. McCrady (2,3,4), and Elizabeth E. Epstein (2,5) Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

1 St. Edward’s University

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

2 Rutgers, The State University of New Jersey

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Ryan Wheeler 11/2/2016 10:20 PM

3 University of New Mexico 4 Center on Alcoholism, Substance Abuse, and Addictions University of Massachusetts, Amherst

Comment [1]: Why underlined?

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:(Default) Arial, 11 pt Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:11 pt Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 6:26 PM Formatted: Normal, Line spacing: single

Corresponding author: Kelly E. Green, (voice) 512-448-8708, (email) kellygr@stedwards.edu, St. Edward’s University, School of Behavioral and Social Sciences, 3001 South Congress Avenue, Austin, TX 78704 Author note: This research was funded in part by a training grant from the National Institute on Alcohol Abuse and Alcoholism awarded to Barbara S. McCrady, PhD (T32 AA07569). The author greatly appreciates Drs. Barbara McCrady, PhD, Elizabeth Epstein, PhD, Donald Bux Jr., PhD, and Diana Sanchez, PhD for their guidance on this dissertation project. The author would like to thank Drs. McCrady and Epstein for their ongoing mentorship throughout her graduate education and beyond. Parts of this paper were presented at the 2015 Annual Convention for the Research Society on Alcoholism, San Antonio, TX. All authors were previously affiliated with Rutgers University, Department of Psychology, where this project was initiated and data collected as part of Kelly E. Green’s doctoral dissertation.

Kelly E. Green 11/2/2016 8:35 PM Formatted: Line spacing: single Kelly E. Green 11/2/2016 9:14 PM Formatted: Default Paragraph Font, Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

2

Abstract This web-based study evaluated the substance use in social networks of

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt

lesbian/gay/bisexual (LGB) and heterosexual drinkers who were worried about their drinking

Kelly E. Green 11/2/2016 8:51 PM Formatted: Left

habits. Objectives were: 1) to evaluate differences in social network substance use reported in a

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt

sample of LGB and heterosexual individuals who reported concern about their drinking, and 2)

Deleted:

to evaluate correlates of social network substance use reported by LGB and heterosexual

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

Kelly E. Green 11/2/2016 8:51 PM

respondents. The sample (n = 218) was ethnically diverse (76% White) and mostly middleaged, employed, and college educated. The sample included 22% heterosexual males, 50% heterosexual females, 7% gay/bisexual males, and 21% lesbian/bisexual females. In addition to assessment of personal substance use patterns, respondents were asked to select up to 5

Kelly E. Green 11/2/2016 8:43 PM Moved (insertion) [1]

people who had been important/influential to them in the past six months and then rate each

Kelly E. Green 11/2/2016 8:44 PM Moved (insertion) [2]

person’s social support and alcohol use. The percentage of the assessed social network rated as moderate/heavy drinkers was correlated with respondents’ drinking severity, younger age, shorter relationship length, and not having children or stepchildren. There were no sex differences or sexual orientation differences in reported levels of the assessed network rated as moderate or heavy drinkers, but LGB respondents reported significantly fewer abstainers in their close network than heterosexual participants (15% vs. 25%). The percentage of the assessed social network supportive of alcohol treatment correlated significantly with history of past or current treatment, and there was a trend for women to report more social support for alcohol treatment than men. Findings indicate relationships between personal social roles, personal alcohol use patterns, and social network alcohol use in a sample of persons worried about their drinking. LGB social networks appear to contain fewer abstainers, potentially related to different social role patterns, and there seems to be less support for alcohol treatment than general social support, likely related to stigma. These domains should be explored in treatment and subsequent research.

Kelly E. Green 11/2/2016 8:44 PM Deleted: ddiction

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 8:39 PM Deleted: The sample (n = 218) was ethnically diverse, middle-aged, employed, college educated, 71% female, and 72% heterosexual. Results showed a significant correlation between respondents’ drinking severity and reported drinking in their social network. There were no sex differences or sexual orientation differences i

Kelly E. Green 11/2/2016 8:43 PM Moved up [1]: There were no sex differences or sexual orientation differences in reported levels of the assessed network rated as moderate or heavy drinkers, but LGB respondents reported significantly fewer abstainers in their close network than heterosexual participants (15% vs. 25%).

Kelly E. Green 11/2/2016 8:44 PM Moved up [2]: The percentage of the assessed social network supportive of alcohol treatment correlated significantly with history of past or current treatment, and there was a trend for women to report more social support for addiction treatment than men.

Kelly E. Green 11/2/2016 8:44 PM Deleted: ocial support for addiction treatment than men. The percentage of the assessed network rated as moderate or heavy drinkers correlated significantly with younger age, shorter relationship length, and not having children or stepchildren.


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

3 Kelly E. Green 11/2/2016 8:51 PM Formatted: Left

Keywords: alcohol use disorder, problem drinking, drinking patterns, social network, peer influence, sexual minority, lesbian, gay, bisexual

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Italic Kelly E. Green 11/2/2016 8:45 PM Formatted: Indent: First line: 0.5" Kelly E. Green 11/2/2016 8:45 PM Deleted: LGB

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 8:36 PM Deleted: Additional research questions:

... [1]


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

4

Substance Use in Social Networks of Lesbian, Gay, Bisexual, and Heterosexual Worried

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar

Drinkers

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt Kelly E. Green 11/2/2016 8:51 PM Formatted: Left

Research clearly supports the role of peer and partner drinking in shaping individual

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt

patterns of use (e.g. McCrady, 2004; Roberts & Leonard, 1998) and treatment of alcohol use disorders often includes close examination of social network drinking. Although the notion of Kelly E. Green 11/2/2016 8:53 PM

“peer pressure� to drink or use drugs often is thought of only in the context of adolescent

Deleted: The

substance use, the influence of peer substance use patterns clearly extends into adulthood.

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

Research has shown that shortly after marriage, the drinking habits of each partner become

Deleted: . However,

more similar. More specifically, within the first year of marriage, partners decrease the

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

frequency of their drinking and drink less often without each other (Demers, Bisson, & Palluy,

Deleted: drinking habits

1999; Leonard & Eiden, 1999; Leonard & Rothbard, 1999; Roberts & Leonard, 1997; Roberts &

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

Kelly E. Green 11/2/2016 8:53 PM

Leonard, 1998). Additionally, there is some evidence that heavy drinking and alcohol dependence of one spouse represent risk factors to the other spouse, with the assumption that imitation or social contagion contributes to this elevated risk (McLeod, 1993). The effects of

Kelly E. Green 11/2/2016 8:52 PM

seu 11/2/2016 10:20 PM Comment [2]: Word choice

Kelly E. Green 11/2/2016 8:53 PM Deleted: s

Ryan Wheeler 9/27/2016 9:00 PM Deleted: assumpt

social network substance use are not limited to romantic partners or spouses. There is a strong

seu 11/2/2016 10:20 PM Comment [3]: This seems unclear

correlation between drinking habits and perceptions of peer drinking in college samples (e.g. Fromme & Ruela, 1994). Research has shown that support for drinking, often quantified based

Kelly E. Green 11/2/2016 8:55 PM Deleted: changes to social networks to exclude users and include

on the drinking status of important peer and family relations, is associated with poorer treatment

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

outcomes (Beattie & Longabaugh, 1997). In line with these tendencies, treatment for substance

Deleted: and often thought to be centered on

use problems often includes increasing support for sobriety, potentially by changing social

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

network composition to limit contact with current users and increase contact with abstainers.

Deleted: are

Ryan Wheeler 9/27/2016 9:20 PM

Kelly E. Green 11/2/2016 8:56 PM

Working from a social learning perspective leads researchers and clinicians to predict

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Ryan Wheeler 9/27/2016 9:20 PM

differences between LGB and heterosexual drinking for several reasons. First, the LGB

Deleted: that involve

community is unique because social activities often are centered on drinking and drug use (e.g.

Deleted: quality

bars, circuit parties). This unique aspect of this population could lead to social networks of LGB

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

Kelly E. Green 11/2/2016 8:57 PM


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

5

individuals that consist of heavier substance users than those of heterosexual individuals.

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [3]

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [2]

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [4]

Kelly E. Green 11/2/2016 8:58 PM

Recent research has indicated that there is greater overall substance use in LGB communities

Deleted: Although empirical data on drinking ... [5]

compared to heterosexual communities (Kerr et al.; McCabe et al.) The present study

Deleted: come out consistently showing

addressed this gap in the literature by examining the relationship between respondent

Deleted: homosexual

substance use and the reported substance use patterns in their close social networks in a

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [6]

sample of lesbian, gay, bisexual, and heterosexual individuals who were worried about their

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [7]

drinking habits.

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [8]

The Present Study The present study had two primary aims: 1) to evaluate differences in social network

Kelly E. Green 11/2/2016 8:58 PM Kelly E. Green 11/2/2016 8:58 PM

Kelly E. Green 11/2/2016 10:20 PM Comment [4]: Need year and all authors for...first [9]

Kelly E. Green 11/2/2016 10:20 PM Comment [5]: We need something here about ... [11] the

Kelly E. Green 11/2/2016 8:59 PM

substance use reported in a sample of LGB and heterosexual individuals who reported concern

Deleted: .

about their drinking, and 2) to evaluate correlates of social network substance use reported by

Kelly E. Green 11/2/2016 9:14 PM Formatted

LGB and heterosexual respondents. Based on previous research and social learning theory, it was hypothesized that higher rates of moderate/heavy drinking in respondents’ close social

... [10]

Ryan Wheeler 9/26/2016 11:21 PM Deleted: However

Ryan Wheeler 9/26/2016 11:21 PM Deleted: , the majority evidence about the ... [13]

networks would be correlated with respondents’ drinking severity . Other variables (e.g.

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [12]

relationship satisfaction, social support for sobriety/treatment seeking, demographic

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [14]

characteristics) were examined in an exploratory manner.

Kelly E. Green 11/2/2016 9:01 PM Deleted: the direct effect of close

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [15]

Kelly E. Green 11/2/2016 9:02 PM Deleted: and substance use reported by

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [16]

Kelly E. Green 11/2/2016 9:04 PM Deleted: 1) to evaluate differences in

... [17]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [18]

seu 11/2/2016 10:20 PM Comment [6]: Unclear sentence

seu 11/2/2016 10:20 PM Comment [7]: Based on previous research...and [20]

Kelly E. Green 11/2/2016 9:04 PM Deleted: that drinking

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [19]

Kelly E. Green 11/2/2016 9:05 PM Deleted: reported by respondents

Kelly E. Green 11/2/2016 9:05 PM Deleted: since previous research has not ... [21]


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

6 Kelly E. Green 11/2/2016 8:51 PM Formatted

Method Participants and Recruitment Participants were recruited from 5/19/2007 to 5/31/2008 by posting a link to the webbased survey on health-related, alcohol-related, and LGB-related websites. Gay/lesbian/bisexual participants were actively recruited by posting advertisements and sending flyers to 146 LGB-related websites and community organizations across the United States. The advertisements said, “Are you worried that you drink too much? Take this online survey to get free personalized feedback on your drinking habits.” The only requirement for participation was

... [22]

Kelly E. Green 11/2/2016 8:51 PM Formatted: Left Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Ryan Wheeler 11/2/2016 10:20 PM Comment [8]: Passive

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [23]

Ryan Wheeler 11/2/2016 10:20 PM Comment [9]: Passive

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [24]

that respondents were age 18 or older; this was not verified objectively, but was based on reported age only. Potential participants were assured of anonymity and provided passive

Ryan Wheeler 11/2/2016 10:20 PM Comment [10]: Passive voice.

consent by entering the website. Upon completion of the survey and as incentive for

Kelly E. Green 11/2/2016 9:14 PM Formatted

participation, respondents received individualized feedback (based on their reported drinking

Kelly E. Green 11/2/2016 9:06 PM

patterns) and treatment referral information.

Deleted: Participants were recruited from 5/19/2007 to 5/31/2008… During the ... [26]

. During the recruitment period, 792 respondents initiated the survey, and 543 (69%) of

... [25]

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:07 PM

those failed to answer a single question. Two-hundred forty-nine respondents (31%) provided

Deleted: . Of those,

demographic data. Of those 249 respondents, seven respondents (3%) were excluded because

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [27]

Kelly E. Green 11/2/2016 9:07 PM

they reported their gender as something other than male or female, and ten respondents (4%)

Deleted:

were excluded because they reported their sexual orientation as “Questioning/Undecided”,

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

“Asexual”, or “None of the above.” Four other entries were excluded because they were judged

Deleted: …were excluded because they... [28]

to be duplicate data. Thus the sample used for data analysis consisted of 218 respondents.

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

Overall, the sample was ethnically diverse (76% White), middle-aged, employed, and college

Kelly E. Green 11/2/2016 9:08 PM

seu 11/2/2016 10:20 PM

educated. Table 1 summarizes demographic characteristics. Based on self-reported sexual

Comment [11]: Maybe provide more on the diversity? Although there were many ethnicities, the majority was still Caucasian.

orientation, the sample included 22% heterosexual males (n = 48), 50% heterosexual females

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM

(n = 108), 7% gay/bisexual males (n = 16), and 21% lesbian/bisexual females (n = 46). Seventy percent of respondents reported that they were married or in a committed relationship and 25% of respondents reported that they had children living in their home.

Deleted: With regards to sexual orientation, 48 (72% of men) men identified as heterosexual, 14 (21%) men identified as gay, and 2 (3%) men identified as bisexual; 108 (68%) women identified as heterosexual, 34 (21%) women identified as lesbian, and 12 (8%) women identified as bisexual.…Seventy percent of ... [29]


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

7 Kelly E. Green 11/2/2016 8:51 PM Formatted: Left

Measures An Internet survey was developed to assess sexual orientation, drinking patterns, drug use history, social network substance use, relationship satisfaction, perceived barriers to treatment-seeking, and treatment preferences of adult problem drinkers (see Green, 2011 for findings related to treatment barriers and preferences). The measures listed below were

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [30]

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt Ryan Wheeler 11/2/2016 10:20 PM Comment [12]: Passive

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [31]

Kelly E. Green 11/2/2016 10:20 PM

adapted for online administration and some were modified for use in the LGB population. The

Comment [13]: We need to add this to the reference list

Internet survey was pilot tested (n = 16 peers and colleagues) and adjustments to phrasing and

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

format were made prior to data collection. Demographics. Respondents were assessed on the following demographic variables: sex, age, ethnicity, education, employment status, geographic location, relationship status,

Kelly E. Green 11/2/2016 9:16 PM Formatted

... [32]

length of relationship, presence of children in the home, self-reported sexual orientation, and gender of past year sexual partners. Kelly E. Green 11/2/2016 9:16 PM

Alcohol Use Disorder Identification Test (AUDIT;

Deleted: ,

Daily Drinking Questionnaire-Revised (DDQ-R; Kruse, Corbin, & Fromme, 2005;

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 10:20 PM

Collins, Parks, & Marlatt, 1985). The DDQ-R more closely resembles validated retrospective

Comment [14]: We need to add a description of this measure, the citation, and add it to the ref list.

interviews (e.g. Timeline Followback, Form 90) than other quantity-frequency measures of

Kelly E. Green 11/2/2016 9:28 PM Formatted: Font:Not Italic

substance use patterns. The DDQ-R disaggregates quantity and frequency indices to yield more

Kelly E. Green 11/2/2016 9:16 PM Formatted

... [33]

accurate estimates of drinking frequency and intensity. The DDQ-R first assesses frequency of alcohol consumption on each day of the week during the past three months, then assesses typical consumption levels (i.e., number of standard drinks) for days of the week when any drinking was reported. Responses to the DDQ-R were used to calculate estimates of percent Kelly E. Green 11/2/2016 9:18 PM

drinking days (PDD), mean drinks per drinking day (MDPDD), and mean drinks per week

Deleted: drinking quantity and frequency

(MDPW). These variables were used as the primary measurements of respondents’ drinking

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

patterns.

Kelly E. Green 11/2/2016 9:18 PM Formatted

History of Drug Use. Three questions were asked to assess history of drug use, and these questions were based on the 2006 National Survey on Drug Use and Health interview

... [34]

Kelly E. Green 11/2/2016 9:18 PM Deleted: (see Appendix D, page 129). These

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [35]


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

8

developed by SAMHSA. The items assessed lifetime, past year, and past month use of 17

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar

groups of illicit drugs. Respondents were presented with a list of drugs and asked to select

Kelly E. Green 11/2/2016 8:51 PM Formatted: Left

which ones they had used ever, used within the past year, and used within the past month.

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt

Respondents were coded dichotomously on whether they had used any of the illicit substances

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt Kelly E. Green 11/2/2016 10:20 PM

in the stated time period. Further distinction was made by coding those who reported drug use as either having used marijuana only, or other drugs either alone or in conjunction with

Comment [15]: We need to cite this here and add to ref list

Kelly E. Green 11/2/2016 9:19 PM Deleted: conducted

marijuana. Important People and Activities (IPA; Longabaugh & Zywiak, 1999; Zywiak,

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:20 PM Deleted: and

Longabaugh, & Wirtz, 2002). The IPA is an interview designed to gather information about an

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

individual’s social network, and it was adapted to a self-report format for this study. Additional

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

questions/answers relevant to LGB individuals were added for the present study. Respondents were asked to name up to 5 people who had been important/influential to them in the past six months. Each person listed was then rated on type of relationship, gender, sexual orientation, amount of contact, drinking/drug use status, support for alcohol/drug use, and support for sobriety/treatment-seeking. Responses were used to calculate the size of respondents’ social networks, percentage of the network who were rated as abstainers from alcohol or moderate/heavy drinkers, percentage of the network that was rated as generally supportive of the respondents, and percentage of the network that respondents reported would be supportive of treatment-seeking for alcohol problems. . Procedures

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:20 PM Formatted: Font:Arial, Italic Kelly E. Green 11/2/2016 9:20 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:21 PM Deleted: . This interview

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:21 PM Deleted: and a

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:21 PM

All procedures were approved by the Institutional Review Board at Rutgers, the State

Deleted: (see Appendix D, page 132)

Kelly E. Green 11/2/2016 9:23 PM

University of New Jersey. Prior to data analyses, comparisons of respondents’ IP addresses

Deleted: or drug use

were used to identify duplicate data. Several steps were taken to minimize missing data.

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

Respondents were informed that complete data were necessary for valid feedback, and a pop-

Comment [16]: May not be needed

up reminder appeared if an item was skipped. Respondents were permitted to continue the

Deleted: (unique to a computer and automatically collected as part of the survey)

survey with items skipped, and they were able to terminate the survey by simply closing the

seu 11/2/2016 10:20 PM Kelly E. Green 11/2/2016 9:23 PM


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

9

website at any time. As incentive to complete the survey, respondents received personalized

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar

feedback upon completion. Feedback included estimations of typical drinking patterns,

Kelly E. Green 11/2/2016 8:51 PM Formatted: Left

estimates of money spent on alcohol, and risks associated with reported level of drinking.

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt

Feedback also included information on perceived social support for drinking/abstinence and

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt seu 11/2/2016 10:20 PM

information about treatment resources.

Comment [17]: Confusing. “that were consistent with web-based research� ?

Kelly E. Green 11/2/2016 9:26 PM Deleted: Despite these efforts and consistent with web-based research, there was a significant proportion of incomplete surveys.

seu 11/2/2016 10:20 PM Comment [18]: May not be needed

Kelly E. Green 11/2/2016 9:26 PM Deleted: (based on their responses to the survey)

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:26 PM Deleted: consumption

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial seu 11/2/2016 10:20 PM Comment [19]: Repeated

seu 11/2/2016 10:20 PM Comment [20]: Repeated

Kelly E. Green 11/2/2016 9:25 PM Deleted: Several steps were taken to minimize missing data. First, respondents were informed that complete data were necessary for valid feedback. Additionally, if a respondent did not complete a question, a pop-up reminder appeared reminding them that complete data were necessary for accurate feedback. Respondents were permitted to continue the survey with items skipped, and they were able to terminate the survey by simply closing the website. There was a significant proportion of incomplete surveys. ... [36]


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

10

Results Alcohol Use Patterns For the whole sample, percent drinking days ranged from 0% to 100% (M = 36.16; SD =

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [37]

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [38]

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [39]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [40]

Kelly E. Green 11/2/2016 9:39 PM Deleted: Substance

31.86), the mean drinks per drinking day ranged from 1 to 10+ (M = 4.00; SD = 2.35), and the

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [41]

mean drinks per week (MDPW) ranged from 0.00 to 70+ (M = 11.28; SD = 13.69). About 50% of

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [42]

Kelly E. Green 11/2/2016 9:30 PM

respondents had MDPW greater than the safe drinking levels recommended by the National

Deleted:

Institute of Health (7 drinks per week for women, 14 for men). It is important to note that there

Comment [21]: Better fit for the discussion ... [43]

was a false ceiling on these estimates because the highest option for respondents to report was

Kelly E. Green 11/2/2016 9:34 PM Moved (insertion) [3]

Kelly E. Green 11/2/2016 10:20 PM

... [44]

Kelly E. Green 11/2/2016 9:30 PM

“10 or more” drinks for a given day. Overall, AUDIT scores indicated that this sample was at

Deleted: In terms of group differences with ... [45]

moderate risk for alcohol use disorders (M = 10.43; SD = 7.70). Respondents were asked to

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [46]

classify their drinking pattern with the prompt, “Do you consider yourself:”. Sixteen (7%) of

Kelly E. Green 11/2/2016 9:34 PM Formatted

... [47]

respondents reported that they considered themselves abstainers, 134 (62%) considered themselves non-problem/normal/social drinkers, 38 (17%) considered themselves problem

Kelly E. Green 11/2/2016 9:34 PM Moved up [3]: Tables 2 and 3 detail

... [48]

Kelly E. Green 11/2/2016 9:34 PM Deleted:

... [49]

drinkers, and 18 (8%) described themselves as alcoholics. Therefore, only 56 respondents

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [50]

(26%) considered themselves problem drinkers or alcoholics. This finding leads to questions

Kelly E. Green 11/2/2016 10:20 PM

about why the remaining respondents chose to complete a survey that advertised for people

Comment [22]: We need to add the the z test info ... [51]

Ryan Wheeler 11/2/2016 10:20 PM Comment [23]: (15.7 vs 9.5)

who were worried about their drinking habits. Unfortunately, reason for response was not

Kelly E. Green 11/2/2016 9:14 PM Formatted

evaluated so one is left to speculation. Tables 2 and 3 detail substance use patterns in this

Kelly E. Green 11/2/2016 9:32 PM

... [52]

Deleted: there was a trend for men to report ... [53]

sample. Gender differences. Men were more likely to consider themselves problem drinkers or

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [54]

Kelly E. Green 11/2/2016 9:33 PM Deleted: and

alcoholics than women (38% vs. 23%; z = 2.133, CI = 96.7) and heterosexual respondents were

Kelly E. Green 11/2/2016 10:20 PM Comment [24]: Also need z test here

more likely to be abstainers than LGB respondents (16% vs. 0%; z = 2.318, CI = 98%). Men

Kelly E. Green 11/2/2016 9:14 PM Formatted

reported more frequent drinking than women (45% versus 32% drinking days), ,men scored

Kelly E. Green 11/2/2016 10:20 PM

significantly higher on the AUDIT than women (12.23 versus 9.68), and there was a trend for men to report more mean drinks per week (16 versus 11 drinks). Additionally, a higher proportion of men exceeded the safe drinking guidelines than women (63% versus 43%).

... [55]

Comment [25]: Add z test here as well

Ryan Wheeler 11/2/2016 10:20 PM Comment [26]: (15.7 vs 9.5)

Ryan Wheeler 11/2/2016 10:20 PM Comment [27]: Represented in poster.

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [56]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [57]

Kelly E. Green 11/2/2016 10:20 PM Comment [28]: Add z test info

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [58]


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

11

Sexual orientation differences. Due to a limited number of bisexual respondents,

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar

sexual orientation was examined by comparing heterosexual respondents to a a collapsed of

Kelly E. Green 11/2/2016 8:51 PM Formatted: Left

sexual minority respondents (ie., gay/bisexual men and lesbian/bisexual women). There were

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt

no differences in drinking patterns reported by lesbian/bisexual women compared to

Kelly E. Green 11/2/2016 9:35 PM Formatted: Font:Bold, Italic

heterosexual women. ANCOVAs were used to evaluate differences between groups based on

Deleted: T

sex and sexual orientation, using covariates selected in univariate correlational analyses

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

described above. The drinking frequency (PDD) ANCOVA revealed sex differences but no

Deleted: and

sexual orientation differences. Men reported more frequent drinking than women, but there was

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

Kelly E. Green 11/2/2016 9:35 PM

no interaction with sexual orientation. More frequent drinking also was reported by those without

Kelly E. Green 11/2/2016 9:35 PM

Kelly E. Green 11/2/2016 10:20 PM

between sexes or sexual orientations. The interaction between sex and sexual orientation was

Comment [29]: Do we have tables for the ANCOVA data? If not, we need to either include the stats in the text of remove those portions and report on the t-tests and z-tests in the tables. At this point, we haven’t discussed any correlations, so this data may just need to be moved further down.

not significant. For mean drinks per week (MDPW), ANCOVA showed differences between

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

children in the home. The drinking intensity (MDPDD) ANCOVA did not reveal differences

sexes, but not between sexual orientations or their interaction. The ANCOVA for AUDIT scores

seu 11/2/2016 10:20 PM Comment [30]: Couldn’t you just write the acronym since it’s been stated before?

evidenced differences between sexes, but not between sexual orientations. The sex by sexual orientation interaction was not significant. Drug Use Patterns Illicit drug use was reported at fairly high rates in the current sample; see table 3 for

Kelly E. Green 11/2/2016 9:39 PM Formatted: Font:Bold

details by group. In order to discriminate use of marijuana alone from use of marijuana and/or

Kelly E. Green 11/2/2016 9:39 PM Formatted: Indent: First line: 0"

other drugs, proportions were examined in terms of any reported drug use, and use of

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

marijuana only. Lifetime drug use was reported by 157 respondents (72%), and 43 of those (20% of whole sample) reported only using marijuana. Past year drug use was reported by 106 respondents (56%), and 41 of those (19% of whole sample) used marijuana only. Past month drug use was reported by 66 respondents (30%) and 36 of those (17% of whole sample) used marijuana only. Sexual orientation differences. LGB respondents reported lifetime drug use (89% versus 74%) and past year drug use (67% versus 50%) more often than heterosexual

Kelly E. Green 11/2/2016 10:20 PM Comment [31]: z-test info

Kelly E. Green 11/2/2016 9:42 PM Formatted: Font:Bold, Italic


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

12

respondents, but there was no difference in past month use. Significantly more gay/bisexual men reported past year drug use than heterosexual men (86% versus 49%), and there was a trend for gay/bisexual men to report more past month drug use (58% versus 33%). These trends were not seen for differences between heterosexual and lesbian/bisexual women for past

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [59]

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [60]

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [61]

Kelly E. Green 11/2/2016 9:46 PM Deleted: g

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [62]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [63]

seu 11/2/2016 10:20 PM

month or past year drug use, but there was a trend for lesbian/bisexual women to report lifetime

Comment [32]: This part is a little confusing

drug use more often than their heterosexual counterparts (90% versus 78%).

Comment [33]: We need to add z-test results for ... [64]

Kelly E. Green 11/2/2016 10:20 PM Kelly E. Green 11/2/2016 9:47 PM

Gender differences. Interestingly, there was a trend for women to report lifetime drug

Deleted: significantly

Kelly E. Green 11/2/2016 9:47 PM

use more often than men (82% versus 70%), although this did not extend to past year or past

Deleted:

month use.

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [65]

Social Networks

Kelly E. Green 11/2/2016 9:43 PM Formatted

... [66]

Respondents were asked to name up to 5 people who had been important/influential to them in the past six months and then rate each important person on several different items. Ninety-three percent of respondents reported the maximum of 5 important people.

Kelly E. Green 11/2/2016 9:43 PM Deleted: when the entire sample was

Comment [34]: We need a bit more summary ... [68] of

Ryan Wheeler 11/2/2016 10:20 PM Comment [35]: In poster

Respondents’ ratings for each of the social network members were used to calculate the

Ryan Wheeler 11/2/2016 10:20 PM

percentage of the network rated certain ways (e.g. percentage rated as generally supportive,

Ryan Wheeler 11/2/2016 10:20 PM

percentage rated as moderate/heavy drinkers). Table 4 summarizes the social network ratings

... [67]

Kelly E. Green 11/2/2016 10:20 PM

Comment [36]: • Add à Higher proportions ... [72] Comment [37]: In poster.

Kelly E. Green 11/2/2016 9:45 PM Deleted: , and LGB respondents reported ... [69]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [70]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [71]

assessed social network; over 80% were rated as generally supportive. There were lower levels

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [73]

of support for the decision to seek treatment; about 50% of the assessed social network was

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [74]

rated as supportive of addiction treatment. Across groups, there was significantly less perceived

Kelly E. Green 11/2/2016 9:49 PM

as well as comparisons of these variables between groups. Social support. Overall, respondents reported high levels of social support from their

Deleted: assessing social support and alcohol ... [75]

social support for SUD treatment than perceived general social support (53% vs 84%) [t(166) = 9.706, p < .001]. The percentage of the assessed social network supportive of SUD treatment correlated significantly with history of past or current SUD treatment. There was a trend for women to report more social support for treatment [57% vs 44%; t(159) = 1.84, p = .068], but no difference by sexual orientation.

seu 11/2/2016 10:20 PM Comment [38]: The rating each respondent gave ... [77]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [76]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [78]

Kelly E. Green 11/2/2016 9:54 PM Deleted: social support variables

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [79]

Kelly E. Green 11/2/2016 10:10 PM Formatted

... [80]

Kelly E. Green 11/2/2016 10:11 PM Formatted

... [81]

Kelly E. Green 11/2/2016 10:10 PM Formatted

... [82]

Kelly E. Green 11/2/2016 9:55 PM Kelly E. Green 11/2/2016 10:10 PM Formatted

... [83]

Kelly E. Green 11/2/2016 10:20 PM ... [84]

Kelly E. Green 11/2/2016 10:10 PM Formatted

... [85]


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

13

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [87]

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [86]

members were rated as moderate or heavy drinkers, whereas only 23% were rated as

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [88]

abstainers from alcohol. Twenty-nine percent of respondents did not classify any of their

Kelly E. Green 11/2/2016 9:55 PM Formatted

... [89]

selected social network members as abstainers from alcohol. There were no sex difference in

Ryan Wheeler 11/2/2016 10:20 PM

Social network substance use. A little over a third of the assessed social network

Comment [40]: Does this statement represent ... [90]

reported levels of general social support, moderate/heavy drinkers, or abstainers. Interestingly,

Kelly E. Green 11/2/2016 9:14 PM Formatted

there was a trend for women to report more social support for alcohol treatment seeking than

Ryan Wheeler 11/2/2016 10:20 PM

... [91]

Comment [41]: Reported in poster

men. There were no sexual orientation differences in the percentages of the network rated as

Kelly E. Green 11/2/2016 9:14 PM Formatted

generally supportive, supportive of treatment seeking, or moderate/heavy drinkers. However,

Kelly E. Green 11/2/2016 10:20 PM

... [92]

Comment [42]: We need to add the statistics for ... [93]

heterosexual respondents reported higher percentages of abstainers in their assessed social

Kelly E. Green 11/2/2016 9:14 PM Formatted

networks than LGB respondents reported (25% vs. 15%). In fact, 21 (34%) LGB respondents

Kelly E. Green 11/2/2016 10:20 PM

reported that none of their social network members were abstainers. Importantly, 38 respondents (24%) reported that none of their assessed social network would support treatment seeking for alcohol problems. When this subgroup was examined more closely, there was a range of alcohol problem severity – PDD ranged from 0 – 90 (M = 24.32; SD = 22.40), MDPDD ranged from 1 – 10+ (M = 3.36; SD = 2.13), MDPW ranged from 0 – 39 (M = 6.89; SD = 8.84), and AUDIT scores ranged from 0 – 16 (M = 6.26; SD = 4.05). Correlates of Social Network Variables Exploratory analyses examined the correlations between demographic variables and

... [94]

Comment [43]: Add the % and z tests

Ryan Wheeler 11/2/2016 10:20 PM Comment [44]: •ADD à 29% of respondents ... [96]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [95]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [97]

Kelly E. Green 11/2/2016 9:59 PM Deleted: Interestingly

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [98]

Kelly E. Green 11/2/2016 10:01 PM Deleted: but many consumed modest

... [99]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [100]

Ryan Wheeler 11/2/2016 10:20 PM

social network variables. The percentage of the assessed social network reported to be

Comment [45]: ADDà Lesbian/bisexual...women [101]

supportive of alcohol treatment was correlated significantly only with the percentage of the

Kelly E. Green 11/2/2016 9:14 PM Formatted

network generally supportive [rho(160) = .244, p = .002] and history of past or current alcohol

Kelly E. Green 11/2/2016 10:01 PM Formatted ... [103]

treatment [rho(149) = .170, p = .039]. The percentage of the assessed network rated as

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [102]

... [104]

Kelly E. Green 11/2/2016 10:02 PM

moderate/heavy drinkers was correlated significantly with age [rho(161) = -.178, p = .024],

Deleted: characteristics of social support ... [105]

relationship length [rho(117) = -.262, p = .004], and having children in the home [rho(154) = -

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [106]

Kelly E. Green 11/2/2016 10:03 PM

.172, p = .033]. It is important to note that children could have been listed as part of the

Deleted: or

assessed social network, so the relationship between children in the home and social network

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [107]


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

14

drinking should be interpreted with caution. There were no gender or sexual orientation differences in rates of marriage or having children. Results supported the hypothesis that the social network alcohol use would correlate with respondent alcohol use. The percentage of the

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [109]

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [108]

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [110]

Ryan Wheeler 11/2/2016 10:20 PM Comment [46]: •ADDà there were no...gender [111]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [112]

Kelly E. Green 11/2/2016 10:07 PM

assessed social network rated as moderate/heavy drinkers was correlated significantly with

Deleted: It was

AUDIT scores [rho(162) = .280, p <.001], PDD [rho(158) = .294, p < .001], and MDPDD

Deleted: ized

[rho(145) = .198, p = .017].

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [113]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [114]

Kelly E. Green 11/2/2016 10:07 PM

Discussion Based on social learning theory and previous research, it was hypothesized that drinking severity would be related to drinking in the social network. This hypothesis was supported for

Kelly E. Green 11/2/2016 10:09 PM Deleted: percentage of respondents’ social ... [115]

Kelly E. Green 11/2/2016 9:14 PM Formatted

PDD, MDPDD, and AUDIT scores. These findings are consistent with previous findings that

Deleted: or

peer and partner drinking shape an individual’s drinking patterns (e.g. McCrady, 2004; Roberts

Kelly E. Green 11/2/2016 9:14 PM Formatted

& Leonard, 1998). These findings are also consistent with research showing some patterns of elevated substance use and increased risk of substance use problems in LGB communities (Green & Feinstein, 2015; Kerr et al.; Lanfear et al.; McCabe et al.) There were no sex

... [116]

Kelly E. Green 11/2/2016 10:09 PM

... [117]

Kelly E. Green 11/2/2016 10:20 PM Comment [47]: We need to add some about ... [118] the

Ryan Wheeler 11/2/2016 10:20 PM Comment [48]: ADDàAcross groups, there ... [121] was

Kelly E. Green 11/2/2016 10:12 PM Deleted: Incorporate analyses from RSA ... [120]

differences or sexual orientation differences in reported levels of the assessed network rated as

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [119]

moderate or heavy drinkers. There were, however, sexual orientation differences in the

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [122]

proportions of the assessed networks rated as abstainers; LGB respondents reported

Kelly E. Green 11/2/2016 10:21 PM Comment [49]: This discussion needs a bit ... [123] more

significantly fewer abstainers than heterosexual participants, and a third of LGB respondents

Kelly E. Green 11/2/2016 9:14 PM Formatted

rated none of their assessed network members as alcohol abstainers. These findings are

Kelly E. Green 11/2/2016 10:20 PM

... [124]

Comment [50]: We need to add this to the ... ref [128]

consistent with the finding that LGB respondents were less likely to abstain from drinking than

Kelly E. Green 11/2/2016 10:20 PM Comment [51]: These citations need the year ... [127] as

heterosexual respondents, and this pattern has implications for treatment and prevention efforts.

Kelly E. Green 11/2/2016 10:14 PM Deleted: increased rate and risk of substance ... [125]

It is commonly posited that LGB substance use is higher partially due to the impact of social network substance use, and some research does support that notion. However, this study suggests that there may be no differences in rates of moderate/heavy drinking, but that LGB individuals have fewer abstainers in their networks. Results from this study showed stronger correlations for ________________. However, the lack of empirical support for the differences

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [126]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [129]

Ryan Wheeler 9/26/2016 11:40 PM Deleted:

Kelly E. Green 11/2/2016 10:24 PM Deleted: (15% vs. 25%)… and a third34% ... [130]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [131]

Kelly E. Green 11/2/2016 10:24 PM Deleted: eported that

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [132]

Kelly E. Green 11/2/2016 10:24 PM Kelly E. Green 11/2/2016 9:14 PM Formatted

... [133]

Kelly E. Green 11/2/2016 10:25 PM Kelly E. Green 11/2/2016 9:14 PM Formatted

... [134]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [135]


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

15

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [137]

between moderate/heavy drinking social network members in LGB and heterosexual

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [136]

respondents may be due to insufficient power or the result of the sample being composed of

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [138]

worried drinkers rather than problem drinkers or individuals with alcohol use disorders. Further

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [139]

studies should examine social network differences in clinical samples as well as general

Kelly E. Green 11/2/2016 10:32 PM

population samples. Overall, the sample reported high rates of general social support and somewhat lower

Deleted: It is possible that heterosexual... [140]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [141]

Kelly E. Green 11/2/2016 10:33 PM Deleted: Other aspects of social network ... [142]

rates of support for treatment seeking; 24% (38 respondents) reported that none of their

Kelly E. Green 11/2/2016 9:14 PM Formatted

assessed network would support treatment seeking for alcohol problems. This finding is

Kelly E. Green 11/2/2016 10:33 PM

... [143]

Deleted: there was a wide range in drinking ... [144]

particularly troublesome. Why is it that the decision to seek treatment for a drinking problem is not perceived as being supported as much as other aspects of life? One possibility is that stigma associated with treatment could lead individuals to think that those close to them would not be supportive of treatment seeking for their drinking. An alternate explanation for this finding is that many of the people who reported no support for treatment were not drinking at very high levels. However, drinking at unsafe levels (more than 7 drinks per week for women or 14 drinks per week for men) was reported by 29% of the respondents who reported that none of their assessed network would be supportive of their decision to seek alcohol treatment. It makes sense that individuals with low or moderate rates of drinking would not perceive much social support for treatment seeking, but for individuals with heavy drinking patterns, that perceived

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [145]

Kelly E. Green 11/2/2016 10:34 PM Deleted: addiction

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [146]

Kelly E. Green 11/2/2016 10:35 PM Deleted: because most of the people who ... [147]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [148]

Kelly E. Green 11/2/2016 10:35 PM Deleted: those i

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [149]

Kelly E. Green 11/2/2016 10:35 PM Deleted: whose drinking patterns are more ... [150]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [151]

Kelly E. Green 11/2/2016 10:35 PM

lack of social support for treatment-seeking could pose an additional barrier to accessing

Deleted: e

services. Interestingly, there was a trend for women to report more social support for addiction

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [152]

Kelly E. Green 11/2/2016 10:36 PM

treatment than men, and this conflicts with previous findings that women report social barriers to

Deleted: treatment seeking.

treatment seeking and opposition to the decision to seek treatment more often than do men

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [153]

Kelly E. Green 11/2/2016 10:36 PM

(Beckman & Amaro, 1986). However, research on barriers has not been conducted in samples

Deleted: . T

with large proportions of LGB participants, and it is possible that this sex difference is driven by

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [154]

Kelly E. Green 11/2/2016 10:38 PM

a sexual orientation difference (see Green 2011 for additional findings from this study on

Deleted: In other words, i

barriers and treatment preferences). It is possible that LGB communities are more supportive of

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [155]


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

16 Kelly E. Green 11/2/2016 8:51 PM Formatted: Left

addiction treatment than heterosexual communities, and this possibility is supported by previous findings that lesbian women are more likely to seek addiction treatment than heterosexual

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar

women (Drabble et al., 2005). Future research should examine further perceived social support

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt

for psychotherapy and addiction treatment in the LGB community.

Deleted: . This notion

Kelly E. Green 11/2/2016 10:38 PM

Exploratory analyses revealed that the percentage of the assessed social network rated as moderate/heavy drinkers correlated significantly with younger age, shorter relationship length, and not having children or stepchildren. The inverse relationship between age and social network drinking is understandable from a “maturing out” perspective suggested by previous research and supported by considerable evidence; older individuals are less likely to be heavy drinkers (e.g. Labouvie, 1996; ). Consistent with that pattern, this study found a significant inverse correlation between age and mean drinks per drinking day. The inverse relationship between relationship length and social network drinking is consistent with research suggesting drinking patterns of partners tend to become more similar over time and relationships have a protective function against problematic drinking (Demers, Bisson, & Palluy, 1999; Leonard &

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 10:39 PM Deleted: examined the correlations between demographic variables and characteristics of social support networks. The percentage of the assessed social network supportive of alcohol treatment correlated significantly with history of past or current treatment and with the percentage of the network that was generally supportive. T

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [156]

Kelly E. Green 11/2/2016 10:40 PM Deleted: or

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 10:40 PM Deleted: negative

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [157]

Kelly E. Green 11/2/2016 10:41 PM

Eiden, 1999; Leonard & Rothbard, 1999; McLeod, 1993; Roberts & Leonard, 1997; Roberts &

Comment [52]: Let’s add a citation for the most recent NSDUH study here. Also, there was a...recent [158]

Leonard, 1998). This finding also could be related to age in that individuals with longer

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

relationships are likely to be older, and therefore further along in the maturing out process. With regards to the finding that social network drinking was negatively correlated with having children, it is important to note that underage children could have been included in the assessed social network, therefore lowering the percentage of drinking-aged people in the assessed social network. However, if this finding is not an artifact of the methodology, it suggests that

Kelly E. Green 11/2/2016 10:42 PM Deleted: In fact, for the current sample there ... [159]

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 10:42 PM Deleted: negative

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 10:42 PM Deleted: negative

individuals with children have social networks composed of fewer moderate/heavy drinkers. This finding would make sense in terms of both a maturing out perspective (parents are likely to be

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 10:43 PM Deleted: makes sense in light of

older and also have parenting responsibilities that could conflict with heavy drinking) and in

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [160]

terms of selecting social network members who limit their heavy drinking, which would increase the safety and family-friendly social interactions. Future research should examine further the

Kelly E. Green 11/2/2016 10:45 PM Deleted: nature of …ocial interactions. Future ... [161]


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

17 Kelly E. Green 11/2/2016 8:51 PM Formatted

... [163]

impact of relationship length, relationship satisfaction, and childrearing on substance use

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [162]

patterns in LGB samples.

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [164]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [165]

Study Strengths and Limitations The present study has several strengths. Internet administration provided a level of

Kelly E. Green 11/2/2016 10:47 PM

anonymity that could have made respondents more comfortable reporting accurate estimates of their drinking and drug use, and also made the survey available to people at various stages of

Deleted: that make it unique in the field

Kelly E. Green 11/2/2016 10:48 PM Deleted: . Additionally, Internet administration

concern about their drinking. Traditional samples often are limited to those who seek treatment

Kelly E. Green 11/2/2016 10:48 PM Deleted: all over the world

through traditional channels, but the present study accessed respondents who largely had no

Kelly E. Green 11/2/2016 9:14 PM Formatted

previous contact with the treatment community. The present sample also was composed of high

Ryan Wheeler 11/2/2016 10:20 PM

proportions of women, who often are underrepresented in treatment samples. The inclusion of men, women, heterosexual, and LGB participants in the same study and the use of validated measures permitted direct comparisons between groups on the constructs under investigation. Uniquely, this study specifically examined close members of social networks rather than the general social networks or community-level comparisons. This allowed a more discrete look at the relationship between social network variables and substance use patterns. Despite these strengths, the present study has several limitations. First, the use of an Internet survey poses several threats to data validity and sample generalizability. The current sample was quite different from nationally representative samples; women and those with a

... [166]

Comment [53]: Is this still true?

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [167]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [168]

Kelly E. Green 11/2/2016 10:49 PM Deleted: T

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [169]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [170]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [171]

Kelly E. Green 11/2/2016 10:51 PM Deleted: greater community, narrowing ... the [172]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [173]

college education were overrepresented in the current sample. However this disparity is not a

Kelly E. Green 11/2/2016 10:53 PM Moved (insertion) [4] ... [174]

critical flaw since the goal of the current study was not to assess the epidemiology of alcohol

Deleted: R

and drug use, but to explore social network substance use reported by individuals who were

Kelly E. Green 11/2/2016 9:14 PM Formatted

Kelly E. Green 11/2/2016 10:53 PM

... [175]

Ryan Wheeler 11/2/2016 10:20 PM

concerned about their drinking. Other limitations to internet research include that respondents

Comment [54]: Fictitious answers? Untrue ... [177]

could terminate the survey before completion, complete the survey more than once, or provide

Deleted: bogus

innaccurate answers in order to finish more quickly. Efforts were made to minimize these

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [176]

possibilities by providing participants with an estimate of completion time, providing pop-up

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [178]

alerts when an item was skipped, and collecting IP addresses to search for duplicates. The

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [179]

Kelly E. Green 11/2/2016 10:18 PM


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

18

sample was not a random sample and could therefore be biased towards LGB individuals who

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar

were more comfortable with their sexual orientation, more comfortable with the Internet, and

Kelly E. Green 11/2/2016 8:51 PM Formatted: Left

those who have Internet access. Although these potential biases are difficult to eliminate, efforts

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt Kelly E. Green 11/2/2016 10:19 PM

were made to recruit individuals from diverse backgrounds for this study (see participants and

Deleted: Little can be done to minimize these biases, but

recruitment section). There was no way to evaluate respondents’ understanding of the survey,

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

and those not fluent in English may have misunderstood questions or answers, therefore limiting

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

the validity of their answers. Finally, the modest sample size resulted in limited power for some Kelly E. Green 11/2/2016 10:54 PM

analyses, so null findings must be considered with caution. Despite these limitations, this study is the first to examine respondent-level social network variables for LGB and heterosexual worried drinkers, and findings provide additional insight in to the complex relationships between social network variables and alcohol use patterns.

Deleted: the use of the Internet for data collection presents an additional threat to sample generalizability.

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 10:53 PM Moved up [4]: The current sample was quite different from nationally representative samples; women and those with a college education were overrepresented in the current sample. However this disparity is not a critical flaw since the goal of the current study was not to assess the epidemiology of alcohol and drug use, but to explore social network substance use reported by individuals who were concerned about their drinking.

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 10:52 PM Deleted: Future Directions

Ryan Wheeler 11/2/2016 10:20 PM Comment [55]: Is this the final copy?


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

19

Acknowledgments

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar Kelly E. Green 11/2/2016 8:51 PM Formatted: Left

This research was funded in part by a training grant from the National Institute on Alcohol Abuse

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt

and Alcoholism awarded to Barbara S. McCrady, Ph.D. (T32 AA07569). The author greatly

Kelly E. Green 11/2/2016 10:20 PM

appreciates Barbara McCrady, Ph.D., Elizabeth Epstein, Ph.D., Donald Bux Jr., Ph.D. and

Comment [56]: Does APA 6th include this in the Author Note section on the title page? Or do we still need this Acknowledgments page?

Diana Sanchez, Ph.D. for their guidance on this dissertation project. The author would like to

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:(Default) Arial, 11 pt

thank Barbara and Beth for their ongoing mentorship throughout her graduate education and

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

beyond.

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

20

References Babor, T. F., Biddle-Higgins, J. C., Sauders, J. B., & Montiero, M. G. (2001). AUDIT: The Alcohol Use Disorders Identification Test: Guidelines for Use in Primary Health Care. Geneva, Switzerland: World Health Organization. Beatty, R.L., Geckle, M.O., Huggins, J., Kapner, C., Lewis, K., & Sandstrom, D.J. (1999). Gay men, lesbians, and bisexuals. In B.S. McCrady & E.E. Epstein (Eds.), Addictions: A comprehensive guidebook (pp. 542-551). New York: Oxford University Press. Beckman, L.J., & Amaro, H. (1986). Personal and social difficulties faced by women and men entering alcoholism treatment. Journal of Studies on Alcohol, 47, 135-145. Beckman, L.J., & Kocel, K.M. (1982). The treatment-delivery system and alcohol abuse in women: Social policy implications. Journal of Social Issues, 38, 139-151. Burckell, L.A., & Goldfried, M.R. (2005, November). Therapist qualities preferred by lesbian, gay male, and bisexual individuals. Presentation at the 39th annual convention of the Association for Behavioral and Cognitive Therapies, Washington D.C. Burgard, S.A., Cochran, S.D., & Mays, V.M. (2005). Alcohol and tobacco use patterns among heterosexually and homosexually experienced California women. Drug & Alcohol Dependence, 77, 61-70. Bux, D.A. (1996). The epidemiology of problem drinking in gay men and lesbians: A critical review. Clinical Psychology Review, 16, 277-298. CASAA Research Division (1995). Readiness Ruler. Available at: http://www.casaa.unm.edu/instr Cochran, S.D., Keenan, C., Schober, C., & Mays, V.M. (2000). Estimates of alcohol use and clinical treatment needs among homosexually active men and women in the U.S. population. Journal of Consulting and Clinical Psychology, 68, 1062-1071. Collins, R.L., Park, G.E., & Marlatt, G.A. (1985). Social determinants of alcohol consumption: The effects of social interaction and model status on the self-administration of alcohol. Journal of Consulting and Clinical Psychology, 53, 189-200. Driscoll, R. (1982). A gay-identified alcohol treatment program: A follow-up study. Journal of Homosexuality, 7, 71-80. Finnegan, D.G., & McNally, E.B. (2002). Counseling Lesbian, Gay, Bisexual, and Transgender Substance Abusers: Dual Identities. New York: Haworth Press. Green, K.E., & Iverson, K.M. (2009). Computerized cognitive-behavioral therapy in a stepped care model of treatment. Professional Psychology: Research and Practice, 40, 96-103.. Kerr, Dianne L. et al. "Substance Use of Lesbian, Gay, Bisexual and Heterosexual College Students." American Journal of Health Behavior, vol. 38, no. 6, 2014, pp. 951-962, doi:10.5993/ajhb.38.6.17. Kruse, M.I., Fromme, K., & Corbin, W.R. (2005). Improving the accuracy of self-report measures of drinking: Disaggregating quantity and frequency indices of alcohol consumption. Alcoholism: Clinical and Experimental Research, 29 supp., 118A.Kerr, Dianne L. et al. "Substance Use of Lesbian, Gay, Bisexual and Heterosexual College Students." American Journal of Health Behavior, vol. 38, no. 6, 2014, pp. 951-962, doi:10.5993/ajhb.38.6.17. Lanfear, Charles et al. "Examining the Relationship of Substance Use and Sexual Orientation." Deviant Behavior, vol. 34, no. 7, 2013, pp. 586-597, doi:10.1080/01639625.2012.749149. McCabe, Sean Esteban et al. "Sexual Orientation, Substance Use Behaviors and Substance Dependence in the United States." Addiction, vol. 104, no. 8, 2009, pp. 1333-1345, doi:10.1111/j.1360-0443.2009.02596.x.

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar Kelly E. Green 11/2/2016 8:51 PM Formatted: Left Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt Kelly E. Green 11/2/2016 10:57 PM Formatted: Font:Arial, Bold, Not Italic Kelly E. Green 11/2/2016 10:58 PM Comment [57]: Does ref list need to be doublespaced for APA 6th?

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:(Default) Arial, 11 pt Ryan Wheeler 10/20/2016 10:24 AM Formatted: EndNote Bibliography Kelly E. Green 11/2/2016 10:58 PM Comment [58]: These EndNote citations are not in APA 6th format.

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:(Default) Arial, 11 pt Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial, English (US), Do not check spelling or grammar Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:(Default) Arial, 11 pt


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

21 Kelly E. Green 11/2/2016 8:51 PM Formatted: Left

McCabe, Sean Esteban et al. "Sexual Orientation, Substance Use Behaviors and Substance Dependence in the United States." Addiction, vol. 104, no. 8, 2009, pp. 1333-1345, doi:10.1111/j.1360-0443.2009.02596.x. Meyers, R.J., Miller, W.R., Smith, J.E., & Tonigan,J.S. (2002). A randomized trial of two methods for engaging treatment-refusing drug users through concerned significant others. Journal of Consulting and Clinical Psychology, 70, 1182-1185. Miller, W.R. & Tonigan, J.S. (1995). Barriers Questionnaire. Available from CASAA Research Division at: http://www.casaa.unm.edu/instr. Peavy, K.M., Cochran, B.N, Kelley, T., Markin, C., Schneider, K., & Gibson, L. (2004). What types of programs are offered for sexual minorities in substance abuse treatment? Presentation at the 38th Annual Convention of the Association for Advancement of Behavior Therapy, New Orleans, LA, November, 2004. Saunders, J.B, Assland, O.G., Babor, T.F., De La Fuente, J.R., & Grant, M. (1993). Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO Collaborative Project on Early Detection of Persons with Harmful Alcohol ConsumptionII. Addiction, 81, 791-804. Schober, R. & Annis, H.M (1996). Barriers to help-seeking for change in drinking: A genderfocused review of the literature. Addictive Behaviors, 21, 81-92. Stall, R., Paul, J.P., Greenwood, G., Pollack, L.M., Bein, E., et al. (2001). Alcohol use, drug use and alcohol-related problems among men who have sex with men: the Urban Men’s Health Study. Addiction, 96, 1589-1601. Thom, B. (1986). Sex differences in help-seeking for alcohol problems – 1. The barriers to helpseeking. British Journal of Addiction, 81, 777-788.

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt Ryan Wheeler 10/20/2016 10:25 AM Deleted:

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:(Default) Arial, 11 pt Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial, English (US), Do not check spelling or grammar Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

22

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [181]

Table 1: Demographic Characteristics of Respondents Included in Analyses (n = 218)

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [180]

Characteristic

Sample Mean or Distribution

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [182]

M = 30.61 SD = 11.91

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [184]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [183]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [185]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [187]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [186]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [188]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [189]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [190]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [191]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [192]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [193]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [194]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [195]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [196]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [197]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [198]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [199]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [200]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [201]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [202]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [203]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [204]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [205]

* Age (n = 218) * Gender (n = 218) Male Female Sexual Orientation (n = 218) Heterosexual Male Female Lesbian/Gay/Bisexual Male Female Ethnicity (n =218) Caucasian only African-American/Black Hispanic/Latino Asian/Pacific Islander Native American Other Missing Education (n = 218) Less than High School High School Only Trade School Some College 2-year College Degree 4-year College Degree Advanced Degree Employment Status (n = 218) Full-time employment Part-time employment Full-time student Homemaker Retired Unemployed (on disability) Unemployed Occupation (n = 214) Higher executives, major professionals Executives, less professionals Administrative, minor professionals Clerical, sales, technicians, servicemen Skilled manual employees Semiskilled employees Unskilled employees Student Homemaker

* 64 (29%) 154 (71%) * 156 (72%) 48 (22%) 108 (50%) 62 (28%) 16 (7%) 46 (21%) * 166 (76%) 7 (3%) 13 (6%) 21 (10%) 1 (1%) 6 (3%) 2 (1%) * 1 (0.5%) 31 (14%) 2 (1%) 68 (31%) 16 (7%) 42 (19%) 58 (27%) * 96 (44%) 31 (14%) 72 (33%) 7 (3%) 2 (1%) 2 (1%) 8 (4%) * 19 (9%) 26 (12%) 29 (14%) 27 (12%) 5 (2%) 4 (2%) 4 (2%) 78 (36%) 6 (3%)


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

23 Kelly E. Green 11/2/2016 8:51 PM Formatted: Left

Retired Unemployed Others, not classifiable Household Income (n = 218) No answer $0 – 15,000 $15,001 – 30,000 $30,001 – 45,000 $45,001 – 60,000 $60,001 – 75,000 $75,001 – 100,000 $100,001 – 125,000 $125,001 – 150,000 $150,001 and above Region of Residence (n =216 ) Outside United States United States New England Middle Atlantic East North Central West North Central South Atlantic West South Central Mountain Pacific Missing Urbanicity (n = 218) Metropolitan/Urban Area Suburban Area Country/Rural Area Missing Relationship Characteristics In a Committed Relationship (n = 218) Married, Civil Union, or Commitment Ceremony (n = 150) With children or stepchildren (n =203)

2 (1%) 9 (4%) 5 (2%)

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:(Default) Arial, 11 pt, Check spelling and grammar

* 8 (4%) 47 (22%) 26 (12%) 26 (12%) 27 (12%) 19 (9%) 25 (12%) 15 (7%) 11 (5%) 14 (6%)

Kelly E. Green 11/2/2016 8:51 PM Formatted: Font:10 pt Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial, Font color: Auto Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

* Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

26 (12%) 190 (87%) 11 (6%) 30 (16%) 12 (6%) 9 (5%) 27 (14%) 15 (8%) 16 (8%) 17 (9%) 53 (28%)

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial, Font color: Auto Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial

* 94 (43%) 90 (41%) 33 (15%) 1 (0.5%)

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial, Font color: Auto

* 152 (70%) * 49 (23% of whole sample) 55 (25% of whole sample)

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial, Font color: Auto Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial, Font color: Auto Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 11:00 PM Comment [59]: These numbers are off, and the formatting is different than other boxes

Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:Arial Kelly E. Green 11/2/2016 9:14 PM Formatted: Font:(Default) Arial, Italic Ryan Wheeler 10/20/2016 10:27 AM Formatted: Normal


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

24

Table 2: Substance Use Differences and by Sex and Sexual Orientation Vari able Perc ent Drin king Days Mea n Drin ks Per Drin king Day Mea n Drin ks Per Wee k AUD IT Scor e Lifeti me Drug Use Past Year Drug Use Past Mont h Drug Use

Group * Male (n = 58) Female (n = 137) * Heterosexual (n = 140) LGB (n = 55) Male (n = 53) Female (n = 127) * Heterosexual (n = 126) LGB (n = 54)

Mean (SD) M = 45.02; SD = 33.63 M = 32.41; SD = 30.44 * M = 35.71; SD = 31.61 M = 37.28; SD = 32.77 M = 4.43; SD = 2.34 M = 3.82; SD = 2.34 * M = 4.04; SD = 2.32 M = 3.91; SD = 2.44

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [207]

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [206]

Kelly E. Green 11/2/2016 8:51 PM Formatted

... [208]

Kelly E. Green 11/2/2016 11:01 PM

Statistics

Deleted: Overall Social Network

t(193) = 2.65, p = .011 * * t(193) = -.31, ns

Kelly E. Green 11/2/2016 11:01 PM Formatted ... [210] Deleted: Differences in Alcohol and Drug ... [211] Use

t(178) = 1.60, ns

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [212]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [214]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [213]

Kelly E. Green 11/2/2016 9:14 PM Formatted

* * t(178) = .33, ns

... [209]

Ryan Wheeler 10/20/2016 10:26 AM

Kelly E. Green 11/2/2016 11:03 PM Comment [60]: I'd like to change "male"...to[216] "men

Male (n = 60) Female (n = 145) * Heterosexual (n = 147) LGB (n = 58) Male (n = 64) Female (n = 154) * Heterosexual (n = 156) LGB (n = 62) Male (n = 59) Female (n = 147) * Heterosexual (n = 148) LGB (n = 56) Male (n = 58) Female (n = 138) * Heterosexual (n = 143) LGB (n = 52) Male (n = 55) Female (n = 134) * Heterosexual(n = 138) LGB (n = 50)

M = 15.69; SD = 15.64 M = 9.46; SD = 12.41 * M = 11.23; SD = 13.11 M = 11.42; SD = 15.18 M = 12.23; SD = 7.50 M = 9.68; SD = 7.68 * M = 10.49; SD = 7.90 M = 10.29; SD = 7.24 41 (70%) 121 (82%) * 109 (74%) 50 (89%) 33 (57%) 76 (55%) * 72 (50%) 35 (67%) 21 (38%) 47 (35%) * 45 (33%) 21 (42%)

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [215]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [217]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [218]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [220]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [222]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [219]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [221]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [223]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [224]

* * Z = 2.104, p = .035

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [229]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [225]

Z = 0.40, ns

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [227]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [226]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [228]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [230]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [231]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [236]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [232]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [234]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [233]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [235]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [237]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [238]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [243]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [239]

Kelly E. Green 11/2/2016 9:14 PM Formatted

... [241]

t(203) = 2.75, p = .007 * * t(203) = -.09, ns

t(216) = 2.25, p = .025 * * t(216) = .17, ns Z = -2.03, p = .04 * * Z = -2.40 p = .02 Z = 0.24, ns

* * Z = -1.192, ns


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

1

Substance Use in Social Networks of Lesbian, Gay, Bisexual, and Heterosexual Worried Drinkers

Kelly E. Green (1,2), Barbara S. McCrady (2,3,4), and Elizabeth E. Epstein (2,5)

1 St. Edward’s University 2 Rutgers, The State University of New Jersey Ryan Wheeler 12/9/2016 11:13 AM

3 University of New Mexico 4 Center on Alcoholism, Substance Abuse, and Addictions University of Massachusetts, Amherst

Corresponding author: Kelly E. Green, (voice) 512-448-8708, (email) kellygr@stedwards.edu, St. Edward’s University, School of Behavioral and Social Sciences, 3001 South Congress Avenue, Austin, TX 78704 Author note: This research was funded in part by a training grant from the National Institute on Alcohol Abuse and Alcoholism awarded to Barbara S. McCrady, PhD (T32 AA07569). The author greatly appreciates Drs. Barbara McCrady, PhD, Elizabeth Epstein, PhD, Donald Bux Jr., PhD, and Diana Sanchez, PhD for their guidance on this dissertation project. The author would like to thank Drs. McCrady and Epstein for their ongoing mentorship throughout her graduate education and beyond. Parts of this paper were presented at the 2015 Annual Convention for the Research Society on Alcoholism, San Antonio, TX. All authors were previously affiliated with Rutgers University, Department of Psychology, where this project was initiated and data collected as part of Kelly E. Green’s doctoral dissertation.

Comment [1]: Why underlined?


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

2

Abstract This web-based study evaluated the substance use in social networks of lesbian/gay/bisexual (LGB) and heterosexual drinkers who were worried about their drinking habits. Objectives were: 1) to evaluate differences in social network substance use reported in a sample of LGB and heterosexual individuals who reported concern about their drinking, and 2) to evaluate correlates of social network substance use reported by LGB and heterosexual respondents. The sample (n = 218) was ethnically diverse (76% White) and mostly middleaged, employed, and college educated. The sample included 22% heterosexual males, 50% heterosexual females, 7% gay/bisexual males, and 21% lesbian/bisexual females. In addition to assessment of personal substance use patterns, respondents were asked to select up to 5 people who had been important/influential to them in the past six months and then rate each person’s social support and alcohol use. The percentage of the assessed social network rated as moderate/heavy drinkers was correlated with respondents’ drinking severity, younger age, shorter relationship length, and not having children or stepchildren. There were no sex differences or sexual orientation differences in reported levels of the assessed network rated as moderate or heavy drinkers, but LGB respondents reported significantly fewer abstainers in their close network than heterosexual participants (15% vs. 25%). The percentage of the assessed social network supportive of alcohol treatment correlated significantly with history of past or current treatment, and there was a trend for women to report more social support for alcohol treatment than men. Findings indicate relationships between personal social roles, personal alcohol use patterns, and social network alcohol use in a sample of persons worried about their drinking. LGB social networks appear to contain fewer abstainers, potentially related to different social role patterns, and there seems to be less support for alcohol treatment than general social support, likely related to stigma. These domains should be explored in treatment and subsequent research.


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

3

Keywords: alcohol use disorder, problem drinking, drinking patterns, social network, peer influence, sexual minority, lesbian, gay, bisexual


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

4

Substance Use in Social Networks of Lesbian, Gay, Bisexual, and Heterosexual Worried Drinkers

Research clearly supports the role of peer and partner drinking in shaping individual patterns of use (e.g. McCrady, 2004; Roberts & Leonard, 1998) and treatment of alcohol use disorders often includes close examination of social network drinking. Although the notion of “peer pressure� to drink or use drugs often is thought of only in the context of adolescent substance use, the influence of peer substance use patterns clearly extends into adulthood. Research has shown that shortly after marriage, the drinking habits of each partner become more similar. More specifically, within the first year of marriage, partners decrease the frequency of their drinking and drink less often without each other (Demers, Bisson, & Palluy, 1999; Leonard & Eiden, 1999; Leonard & Rothbard, 1999; Roberts & Leonard, 1997; Roberts & Leonard, 1998). Additionally, there is some evidence that heavy drinking and alcohol dependence of one spouse represent risk factors to the other spouse, with the assumption that seu 12/9/2016 11:13 AM

imitation or social contagion contributes to this elevated risk (McLeod, 1993). The effects of

Comment [2]: Word choice

social network substance use are not limited to romantic partners or spouses. There is a strong correlation between drinking habits and perceptions of peer drinking in college samples (e.g. Fromme & Ruela, 1994). Research has shown that support for drinking, often quantified based on the drinking status of important peer and family relations, is associated with poorer treatment seu 12/9/2016 11:13 AM

outcomes (Beattie & Longabaugh, 1997). In line with these tendencies, treatment for substance use problems often includes increasing support for sobriety, potentially by changing social network composition to limit contact with current users and increase contact with abstainers. Working from a social learning perspective leads researchers and clinicians to predict differences between LGB and heterosexual drinking for several reasons. First, the LGB community is unique because social activities often are centered on drinking and drug use (e.g. bars, circuit parties). This unique aspect of this population could lead to social networks of LGB

Comment [3]: This seems unclear


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

5

individuals that consist of heavier substance users than those of heterosexual individuals. Recent research has indicated that there is greater overall substance use in LGB communities compared to heterosexual communities (Kerr, Ding, & Chaya, 2014; McCabe, Hughes, Bostwick, West, & Boyd, 2009) However, these studies were not studying social network as Kelly E. Green 12/9/2016 11:13 AM

specifically as the current study. These studies were attempting to gauge the affect of the overall homosexual community. The present study addressed this gap in the literature by examining the relationship between respondent substance use and the reported substance use patterns in their close social networks in a sample of lesbian, gay, bisexual, and heterosexual individuals who were worried about their drinking habits.

Comment [4]: Need year and all authors for first time cited for each of these

Ryan Wheeler 12/9/2016 8:29 PM Comment [5]: 4 DONE

Kelly E. Green 12/9/2016 11:13 AM Comment [6]: We need something here about the way that social network substance use was defined/measured in those studies… is this one the first to examine the individual level social network as opposed to the global trends in the community vs heterosexual community?

The Present Study The present study had two primary aims: 1) to evaluate differences in social network substance use reported in a sample of LGB and heterosexual individuals who reported concern about their drinking, and 2) to evaluate correlates of social network substance use reported by LGB and heterosexual respondents. Based on previous research and social learning theory, it was hypothesized that higher rates of moderate/heavy drinking in respondents’ close social networks would be correlated with respondents’ drinking severity . Other variables (e.g. seu 12/9/2016 11:13 AM

relationship satisfaction, social support for sobriety/treatment seeking, demographic

Comment [7]: Unclear sentence

characteristics) were examined in an exploratory manner.

Comment [8]: Based on previous research and social learning theory, it was hypothesized that the social networks respondents’ percentage of drinking would be correlated with drinking severity reported by respondents.

seu 12/9/2016 11:13 AM


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

6

Method Participants and Recruitment Participants were recruited from 5/19/2007 to 5/31/2008 by posting a link to the webbased survey on health-related, alcohol-related, and LGB-related websites. Gay/lesbian/bisexual participants were actively recruited by posting advertisements and sending flyers to 146 LGB-related websites and community organizations across the United States. The advertisements said, “Are you worried that you drink too much? Take this online survey to get free personalized feedback on your drinking habits.” The only requirement for participation was that respondents were age 18 or older; this was not verified objectively, but was based on reported age only. Potential participants were assured of anonymity and provided passive consent by entering the website. Upon completion of the survey and as incentive for participation, respondents received individualized feedback (based on their reported drinking patterns) and treatment referral information. . During the recruitment period, 792 respondents initiated the survey, and 543 (69%) of those failed to answer a single question. Two-hundred forty-nine respondents (31%) provided demographic data. Of those 249 respondents, seven respondents (3%) were excluded because they reported their gender as something other than male or female, and ten respondents (4%) were excluded because they reported their sexual orientation as “Questioning/Undecided”, “Asexual”, or “None of the above.” Four other entries were excluded because they were judged to be duplicate data. Thus the sample used for data analysis consisted of 218 respondents. Overall, the sample was ethnically diverse (76% White), middle-aged, employed, and college seu 12/9/2016 11:13 AM

educated. Table 1 summarizes demographic characteristics. Based on self-reported sexual orientation, the sample included 22% heterosexual males (n = 48), 50% heterosexual females (n = 108), 7% gay/bisexual males (n = 16), and 21% lesbian/bisexual females (n = 46). Seventy percent of respondents reported that they were married or in a committed relationship and 25% of respondents reported that they had children living in their home.

Comment [9]: Maybe provide more on the diversity? Although there were many ethnicities, the majority was still Caucasian.


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

7

Measures An Internet survey was developed to assess sexual orientation, drinking patterns, drug use history, social network substance use, relationship satisfaction, perceived barriers to treatment-seeking, and treatment preferences of adult problem drinkers (see Green, 2011 for Kelly E. Green 12/9/2016 11:13 AM

findings related to treatment barriers and preferences). The measures listed below were adapted for online administration and some were modified for use in the LGB population. The Internet survey was pilot tested (n = 16 peers and colleagues) and adjustments to phrasing and

Comment [10]: We need to add this to the reference list

Ryan Wheeler 12/9/2016 11:13 AM Comment [11]: Do I have access to this? Can I cite it?

format were made prior to data collection. Demographics. Respondents were assessed on the following demographic variables: sex, age, ethnicity, education, employment status, geographic location, relationship status, length of relationship, presence of children in the home, self-reported sexual orientation, and gender of past year sexual partners. Alcohol Use Disorder Identification Test (AUDIT) (Babor, 2001) The AUDIT was Ryan Wheeler 12/9/2016 8:32 PM

devised by the World Health Association to help to help healthcare professionals identify

Comment [12]: 13 DONE

harmful or hazardous drinking patterns. It uses a 10-item questionnaire with a 5-point Likert

Comment [13]: We need to add a description of this measure, the citation, and add it to the ref list.

Kelly E. Green 12/9/2016 11:13 AM

scale to identify amount and frequency of drinking, alcohol dependence, and problematic alcohol use. Daily Drinking Questionnaire-Revised (DDQ-R; Kruse, Corbin, & Fromme, 2005; Collins, Parks, & Marlatt, 1985). The DDQ-R more closely resembles validated retrospective interviews (e.g. Timeline Followback, Form 90) than other quantity-frequency measures of substance use patterns. The DDQ-R disaggregates quantity and frequency indices to yield more accurate estimates of drinking frequency and intensity. The DDQ-R first assesses frequency of alcohol consumption on each day of the week during the past three months, then assesses typical consumption levels (i.e., number of standard drinks) for days of the week when any drinking was reported. Responses to the DDQ-R were used to calculate estimates of percent drinking days (PDD), mean drinks per drinking day (MDPDD), and mean drinks per week


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

8

(MDPW). These variables were used as the primary measurements of respondents’ drinking patterns. History of Drug Use. Three questions were asked to assess history of drug use, and these questions were based on the 2006 National Survey on Drug Use and Health interview developed by SAMHSA. (Center for Behavioral Health Statistics and Quality. 2015) The items Kelly E. Green 12/9/2016 11:13 AM

assessed lifetime, past year, and past month use of 17 groups of illicit drugs. Respondents were presented with a list of drugs and asked to select which ones they had used ever, used within

Comment [14]: We need to cite this here and add to ref list

Ryan Wheeler 12/9/2016 11:13 AM Comment [15]: DONE

the past year, and used within the past month. Respondents were coded dichotomously on whether they had used any of the illicit substances in the stated time period. Further distinction was made by coding those who reported drug use as either having used marijuana only, or other drugs either alone or in conjunction with marijuana. Important People and Activities (IPA; Longabaugh & Zywiak, 1999; Zywiak, Longabaugh, & Wirtz, 2002). The IPA is an interview designed to gather information about an individual’s social network, and it was adapted to a self-report format for this study. Additional questions/answers relevant to LGB individuals were added for the present study. Respondents were asked to name up to 5 people who had been important/influential to them in the past six months. Each person listed was then rated on type of relationship, gender, sexual orientation, amount of contact, drinking/drug use status, support for alcohol/drug use, and support for sobriety/treatment-seeking. Responses were used to calculate the size of respondents’ social networks, percentage of the network who were rated as abstainers from alcohol or moderate/heavy drinkers, percentage of the network that was rated as generally supportive of the respondents, and percentage of the network that respondents reported would be supportive of treatment-seeking for alcohol problems. . Procedures All procedures were approved by the Institutional Review Board at Rutgers, the State University of New Jersey. Prior to data analyses, comparisons of respondents’ IP addresses

seu 12/9/2016 11:13 AM Comment [16]: May not be needed


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

9

were used to identify duplicate data. Several steps were taken to minimize missing data. Respondents were informed that complete data were necessary for valid feedback, and a popup reminder appeared if an item was skipped. Respondents were permitted to continue the survey with items skipped, and they were able to terminate the survey by simply closing the website at any time. As incentive to complete the survey, respondents received personalized feedback upon completion. Feedback included estimations of typical drinking patterns, estimates of money spent on alcohol, and risks associated with reported level of drinking. Feedback also included information on perceived social support for drinking/abstinence and information about treatment resources.


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

10

Results Overall Alcohol and Drug Use Patterns For the whole sample, percent drinking days ranged from 0% to 100% (M = 36.16; SD = 31.86), the mean drinks per drinking day ranged from 1 to 10+ (M = 4.00; SD = 2.35), and the mean drinks per week (MDPW) ranged from 0.00 to 70+ (M = 11.28; SD = 13.69). About 50% of respondents had MDPW greater than the safe drinking levels recommended by the National Institute of Health (7 drinks per week for women, 14 for men). It is important to note that there was a false ceiling on these estimates because the highest option for respondents to report was “10 or more” drinks for a given day. Overall, AUDIT scores indicated that this sample was at moderate risk for alcohol use disorders (M = 10.43; SD = 7.70). Respondents were asked to classify their drinking pattern with the prompt, “Do you consider yourself:”. Sixteen (7%) of respondents reported that they considered themselves abstainers, 134 (62%) considered themselves non-problem/normal/social drinkers, 38 (17%) considered themselves problem drinkers, and 18 (8%) described themselves as alcoholics. Therefore, only 56 respondents (26%) considered themselves problem drinkers or alcoholics. Tables 2 and 3 detail substance Kelly E. Green 12/9/2016 11:13 AM

use patterns in this sample. Illicit drug use was reported at fairly high rates in the current sample; see table 3 for details by group. In order to discriminate use of marijuana alone from use of marijuana and/or other drugs, proportions were examined in terms of any reported drug use, and use of marijuana only. Lifetime drug use was reported by 157 respondents (72%), and 43 of those (20% of whole sample) reported only using marijuana. Past year drug use was reported by 106 respondents (56%), and 41 of those (19% of whole sample) used marijuana only. Past month drug use was reported by 66 respondents (30%) and 36 of those (17% of whole sample) used marijuana only. Sexual orientation differences with drugs and alcohol use. Due to a limited number of bisexual respondents, sexual orientation was examined by comparing heterosexual

Comment [17]: Better fit for the discussion section??

Ryan Wheeler 12/9/2016 8:37 PM Deleted: This finding leads to questions about why the remaining respondents chose to complete a survey that advertised for people who were worried about their drinking habits. Unfortunately, reason for response was not evaluated so one is left to speculation.


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

11

respondents to a collapsed of sexual minority respondents (ie., gay/bisexual men and lesbian/bisexual women). There were no significant differences in drinking patterns reported by lesbian/bisexual women compared to heterosexual women or for gay and bisexual men and heterosexual men. There were also no significant differences in AUDIT scores for lesbian or heterosexual women or gay and heterosexual men.

Kelly E. Green 12/9/2016 11:17 AM Comment [18]: Do we have tables for the

... [2]

Ryan Wheeler 12/9/2016 11:24 AM Deleted: ANCOVAs were used to evaluate ... [1]

Ryan Wheeler 12/9/2016 11:33 AM Deleted: More frequent drinking also was... [3]

seu 12/9/2016 11:17 AM Comment [19]: Couldn’t you just write the... [5]

Kelly E. Green 12/9/2016 11:17 AM Comment [20]: z-test info

LGB respondents reported lifetime drug use (Z = -2.40 p = .02) and past year drug use

Ryan Wheeler 12/9/2016 11:31 AM Deleted: The drinking intensity (MDPDD) ... [4]

(Z = 2.104, p = .035) more often than heterosexual respondents, but there was no difference in

Ryan Wheeler 12/9/2016 11:37 AM Deleted: 89% versus 74%)

past month use. Significantly more gay/bisexual men reported past year drug use than

Ryan Wheeler 12/9/2016 11:38 AM Deleted: 67% versus 50%

heterosexual men (Z = 2.427), and there was a trend for gay/bisexual men to report more past month drug use (Z = 1.567) These trends were not seen for differences between heterosexual and lesbian/bisexual women for past month or past year drug use, but there was a trend for lesbian/bisexual women to report lifetime drug use more often than their heterosexual counterparts (Z = 1.679+) Gender differences with drugs and alcohol use. Men were more likely to consider themselves problem drinkers or alcoholics than women (38% vs. 23%; z = 2.133, CI = 96.7) and

seu 12/9/2016 11:17 AM Comment [21]: This part is a little confusing

Kelly E. Green 12/9/2016 11:17 AM Comment [22]: We need to add z-test results ... [6] for

Ryan Wheeler 12/9/2016 11:39 AM Deleted: 86% versus 49%)

Ryan Wheeler 12/9/2016 11:42 AM Deleted: 58% versus 33%).

Ryan Wheeler 12/9/2016 11:42 AM Deleted: 90% versus 78%).

Kelly E. Green 12/9/2016 11:13 AM Comment [23]: We need to add the z test info ... [7]

Ryan Wheeler 12/9/2016 11:50 AM

heterosexual respondents were more likely to be abstainers than LGB respondents (16% vs.

Deleted: 45% versus 32%

0%; z = 2.318, CI = 98%). Men reported more frequent drinking than women (t(193) = 2.65, p =

Deleted: ,

.011 drinking days), men scored significantly higher on the AUDIT than women (t(216) = 2.25, p

Comment [24]: Also need z test here

= .025), and there was a trend for men to report more mean drinks per week (t(203) = 2.75, p =

Deleted: 12.23 versus 9.68

.007). Additionally, a higher proportion of men exceeded the safe drinking guidelines than

Comment [25]: Add z test here as well

women (63% versus 43%).

Deleted: 16 versus 11 drinks

Ryan Wheeler 12/9/2016 11:50 AM Kelly E. Green 12/9/2016 11:13 AM Ryan Wheeler 12/9/2016 11:51 AM Kelly E. Green 12/9/2016 11:13 AM Ryan Wheeler 12/9/2016 11:52 AM Kelly E. Green 12/9/2016 11:13 AM

Interestingly, there was a trend for women to report lifetime drug use more often than

Comment [26]: Add z test info

Ryan Wheeler 12/9/2016 11:53 AM

men (Z = -2.03 p = .04), although this did not extend to past year or past month use. There was

Comment [27]: I cant find this stat in table.

a significant difference where LGB men and women had a higher percentage of lifetime drug

Comment [28]: We need a bit more summary ... [8] of

use and than heterosexual individuals (Z = -2.40 p = .02). There were no significant findings

Deleted: (

with past month drug use.

Deleted: 82% versus 70%

Kelly E. Green 12/9/2016 11:16 AM Ryan Wheeler 12/9/2016 6:55 PM Ryan Wheeler 12/9/2016 6:45 PM


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

12

Social Networks Respondents were asked to name up to 5 people who had been important/influential to them in the past six months and then rate each important person on several different items. Ninety-three percent of respondents reported the maximum of 5 important people. Respondents’ ratings for each of the social network members were used to calculate the percentage of the network rated certain ways (e.g. percentage rated as generally supportive, percentage rated as moderate/heavy drinkers). Table 4 summarizes the social network ratings as well as comparisons of these variables between groups. Social support. Overall, respondents reported high levels of social support from their assessed social network; over 80% were rated as generally supportive. There were lower levels of support for the decision to seek treatment; about 50% of the assessed social network was rated as supportive of addiction treatment. Across groups, there was significantly less perceived social support for SUD treatment than perceived general social support (53% vs 84%) [t(166) = 9.706, p < .001]. The percentage of the assessed social network supportive of SUD treatment correlated significantly with history of past or current SUD treatment. There was a trend for women to report more social support for treatment [57% vs 44%; t(159) = 1.84, p = .068], but no Kelly E. Green 12/9/2016 11:13 AM

difference by sexual orientation.

Comment [29]: We need to check these stats because t-tests don’t correspond to % comparisons

Social network substance use. A little over a third of the assessed social network members were rated as moderate or heavy drinkers, whereas only 23% were rated as abstainers from alcohol. Twenty-nine percent of respondents did not classify any of their selected social network members as abstainers from alcohol. There were no sex difference in reported levels of general social support, moderate/heavy drinkers, or abstainers. Interestingly, there was a trend for women to report more social support for alcohol treatment seeking than men . There were no sexual orientation differences in the percentages of the network rated as generally supportive, supportive of treatment seeking, or moderate/heavy drinkers. However, heterosexual respondents reported higher percentages of abstainers in their assessed social

Ryan Wheeler 12/9/2016 11:13 AM Comment [30]: Does this statement represent this point in the poster? à 29% of respondents did not classify a single person in their reported social network as an abstainer

Kelly E. Green 12/9/2016 11:13 AM Comment [31]: We need to add the statistics for this (% and z tests)

Ryan Wheeler 12/9/2016 7:10 PM Comment [32]: Table 4 Does this refer to “mean percentage of network rated as generally supportive” à (t (39) = 0.275) OR “mean percentage of network rated as supportive of SUD treatment (t (39) = 0.296)


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

13

networks than LGB respondents reported (25% vs. 15%). In fact, 21 (34%) LGB respondents reported that none of their social network members were abstainers. Importantly, 38 Kelly E. Green 12/9/2016 11:13 AM

respondents (24%) reported that none of their assessed social network would support treatment

Comment [33]: Add the % and z tests

Ryan Wheeler 12/9/2016 11:13 AM

seeking for alcohol problems. When this subgroup was examined more closely, there was a range of alcohol problem severity – PDD ranged from 0 – 90 (M = 24.32; SD = 22.40), MDPDD ranged from 1 – 10+ (M = 3.36; SD = 2.13), MDPW ranged from 0 – 39 (M = 6.89; SD = 8.84),

Comment [34]: •ADD à 29% of respondents did not classify a single person in their reported social network as an abstainer

Ryan Wheeler 12/9/2016 7:14 PM Comment [35]: Cant find this info in tables.

and AUDIT scores ranged from 0 – 16 (M = 6.26; SD = 4.05). Ryan Wheeler 12/9/2016 11:13 AM

Correlates of Social Network Variables Exploratory analyses examined the correlations between demographic variables and

Comment [36]: ADDà Lesbian/bisexual women reported significantly fewer abstainers in their networks than heterosexual women (14% vs 25%; t(91) = 2.865***).

social network variables. The percentage of the assessed social network reported to be supportive of alcohol treatment was correlated significantly only with the percentage of the network generally supportive [rho(160) = .244, p = .002] and history of past or current alcohol treatment [rho(149) = .170, p = .039]. The percentage of the assessed network rated as moderate/heavy drinkers was correlated significantly with age [rho(161) = -.178, p = .024], relationship length [rho(117) = -.262, p = .004], and having children in the home [rho(154) = .172, p = .033]. It is important to note that children could have been listed as part of the

Ryan Wheeler 12/9/2016 11:13 AM Comment [37]: •ADDà there were no gender or sexual orientation differences in rates of marriage or having children.

assessed social network, so the relationship between children in the home and social network Kelly E. Green 12/9/2016 11:13 AM

drinking should be interpreted with caution. There were no gender or sexual orientation differences in rates of marriage or having children. Results supported the hypothesis that the social network alcohol use would correlate with respondent alcohol use. The percentage of the assessed social network rated as moderate/heavy drinkers was correlated significantly with AUDIT scores [rho(162) = .280, p <.001], PDD [rho(158) = .294, p < .001], and MDPDD

Comment [38]: We need to add some about the LGB and gender differences in these correlations… just1-2 sentences and then reference to the tables.

Ryan Wheeler 12/9/2016 7:18 PM Comment [39]: Not sure if these are supposed to be Rho or Pearsons. Found these results in table 6 reported as pearsons.. Not sure how you want to handle this. LGB: r (44) = .381 Hetro: r (120)= .244

Kelly E. Green 12/9/2016 11:13 AM

[rho(145) = .198, p = .017]. Discussion Based on social learning theory and previous research, it was hypothesized that drinking severity would be related to drinking in the social network. This hypothesis was supported for

Comment [40]: This discussion needs a bit more re-working… we need to eliminate all the numbers and focus on the big picture… relating back to the lit review… potential implications for treatment and prevention efforts… etc.

Ryan Wheeler 12/9/2016 11:13 AM Comment [41]: I Don’t feel like I can make these changes.


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

14

PDD, MDPDD, and AUDIT scores. These findings are consistent with previous findings that peer and partner drinking shape an individual’s drinking patterns (e.g. McCrady, 2004; Roberts & Leonard, 1998). These findings are also consistent with research showing some patterns of elevated substance use and increased risk of substance use problems in LGB communities (Kerr et al., 2014; Lanfear, Akins, & Mosher, 2013; McCabe et al., 2009) There were no sex Kelly E. Green 12/9/2016 11:13 AM

differences or sexual orientation differences in reported levels of the assessed network rated as moderate or heavy drinkers. There were, however, sexual orientation differences in the proportions of the assessed networks rated as abstainers; LGB respondents reported significantly fewer abstainers than heterosexual participants, and a third of LGB respondents rated none of their assessed network members as alcohol abstainers. These findings are

Comment [42]: We need to add this to the ref list, and also include in the intro.

Ryan Wheeler 12/9/2016 7:19 PM Comment [43]: 44: These are referenced in the intro.

Kelly E. Green 12/9/2016 11:13 AM Comment [44]: These citations need the year as well as all authors for the Lanfear since it’s the first time cited in this paper

Ryan Wheeler 12/9/2016 7:19 PM Comment [45]: 46 DONE

consistent with the finding that LGB respondents were less likely to abstain from drinking than heterosexual respondents, and this pattern has implications for treatment and prevention efforts. It is commonly posited that LGB substance use is higher partially due to the impact of social network substance use, and some research does support that notion. However, this study suggests that there may be no differences in rates of moderate/heavy drinking, but that LGB individuals have fewer abstainers in their networks. Results from this study showed stronger correlations for ________________. However, the lack of empirical support for the differences between moderate/heavy drinking social network members in LGB and heterosexual respondents may be due to insufficient power or the result of the sample being composed of worried drinkers rather than problem drinkers or individuals with alcohol use disorders. Further studies should examine social network differences in clinical samples as well as general population samples. Overall, the sample reported high rates of general social support and somewhat lower rates of support for treatment seeking; 24% (38 respondents) reported that none of their assessed network would support treatment seeking for alcohol problems. This finding is particularly troublesome. Why is it that the decision to seek treatment for a drinking problem is


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

15

not perceived as being supported as much as other aspects of life? One possibility is that stigma associated with treatment could lead individuals to think that those close to them would not be supportive of treatment seeking for their drinking. An alternate explanation for this finding is that many of the people who reported no support for treatment were not drinking at very high levels. However, drinking at unsafe levels (more than 7 drinks per week for women or 14 drinks per week for men) was reported by 29% of the respondents who reported that none of their assessed network would be supportive of their decision to seek alcohol treatment. It makes sense that individuals with low or moderate rates of drinking would not perceive much social support for treatment seeking, but for individuals with heavy drinking patterns, that perceived lack of social support for treatment-seeking could pose an additional barrier to accessing services. The finding that only 20% of participants consider their habits to be harmful leads to questions about why the remaining respondents chose to complete a survey that advertised for people who were worried about their drinking habits. Unfortunately, reason for response was not evaluated so one is left to speculation. Interestingly, there was a trend for women to report more Kelly E. Green 12/9/2016 8:39 PM

social support for addiction treatment than men, and this conflicts with previous findings that women report social barriers to treatment seeking and opposition to the decision to seek treatment more often than do men (Beckman & Amaro, 1986). However, research on barriers has not been conducted in samples with large proportions of LGB participants, and it is possible that this sex difference is driven by a sexual orientation difference (see Green 2011 for additional findings from this study on barriers and treatment preferences). It is possible that LGB communities are more supportive of addiction treatment than heterosexual communities, and this possibility is supported by previous findings that lesbian women are more likely to seek addiction treatment than heterosexual women (Drabble et al., 2005). Future research should examine further perceived social support for psychotherapy and addiction treatment in the LGB community.

Comment [46]: Better fit for the discussion section??

Ryan Wheeler 12/9/2016 8:39 PM Comment [47]: Is this sufficient.


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

16

Exploratory analyses revealed that the percentage of the assessed social network rated as moderate/heavy drinkers correlated significantly with younger age, shorter relationship length, and not having children or stepchildren. The inverse relationship between age and social network drinking is understandable from a “maturing out� perspective suggested by previous research and supported by considerable evidence; older individuals are less likely to be heavy drinkers (e.g. Labouvie, 1996; ). Consistent with that pattern, this study found a significant Kelly E. Green 12/9/2016 11:13 AM

inverse correlation between age and mean drinks per drinking day. The inverse relationship between relationship length and social network drinking is consistent with research suggesting drinking patterns of partners tend to become more similar over time and relationships have a protective function against problematic drinking (Demers, Bisson, & Palluy, 1999; Leonard & Eiden, 1999; Leonard & Rothbard, 1999; McLeod, 1993; Roberts & Leonard, 1997; Roberts & Leonard, 1998). This finding also could be related to age in that individuals with longer relationships are likely to be older, and therefore further along in the maturing out process. With regards to the finding that social network drinking was negatively correlated with having children, it is important to note that underage children could have been included in the assessed social network, therefore lowering the percentage of drinking-aged people in the assessed social network. However, if this finding is not an artifact of the methodology, it suggests that individuals with children have social networks composed of fewer moderate/heavy drinkers. This finding would make sense in terms of both a maturing out perspective (parents are likely to be older and also have parenting responsibilities that could conflict with heavy drinking) and in terms of selecting social network members who limit their heavy drinking, which would increase the safety and family-friendly social interactions. Future research should examine further the impact of relationship length, relationship satisfaction, and childrearing on substance use patterns in LGB samples. Study Strengths and Limitations

Comment [48]: Let’s add a citation for the most recent NSDUH study here. Also, there was a recent NSDUH report on sexual orientation that we should add to the intro as well.


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

17

The present study has several strengths. Internet administration provided a level of anonymity that could have made respondents more comfortable reporting accurate estimates of their drinking and drug use, and also made the survey available to people at various stages of concern about their drinking. Traditional samples often are limited to those who seek treatment through traditional channels, but the present study accessed respondents who largely had no previous contact with the treatment community. The present sample also was composed of high proportions of women, who often are underrepresented in treatment samples. The inclusion of Ryan Wheeler 12/9/2016 11:13 AM

men, women, heterosexual, and LGB participants in the same study and the use of validated

Comment [49]: Is this still true?

measures permitted direct comparisons between groups on the constructs under investigation. Uniquely, this study specifically examined close members of social networks rather than the general social networks or community-level comparisons. This allowed a more discrete look at the relationship between social network variables and substance use patterns. Despite these strengths, the present study has several limitations. First, the use of an Internet survey poses several threats to data validity and sample generalizability. The current sample was quite different from nationally representative samples; women and those with a college education were overrepresented in the current sample. However this disparity is not a critical flaw since the goal of the current study was not to assess the epidemiology of alcohol and drug use, but to explore social network substance use reported by individuals who were concerned about their drinking. Other limitations to internet research include that respondents could terminate the survey before completion, complete the survey more than once, or provide innaccurate answers in order to finish more quickly. Efforts were made to minimize these Ryan Wheeler 12/9/2016 11:13 AM

possibilities by providing participants with an estimate of completion time, providing pop-up alerts when an item was skipped, and collecting IP addresses to search for duplicates. The sample was not a random sample and could therefore be biased towards LGB individuals who were more comfortable with their sexual orientation, more comfortable with the Internet, and those who have Internet access. Although these potential biases are difficult to eliminate, efforts

Comment [50]: Fictitious answers? Untrue answers? Unreliable answers?


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

18

were made to recruit individuals from diverse backgrounds for this study (see participants and recruitment section). There was no way to evaluate respondents’ understanding of the survey, and those not fluent in English may have misunderstood questions or answers, therefore limiting the validity of their answers. Finally, the modest sample size resulted in limited power for some analyses, so null findings must be considered with caution. Despite these limitations, this study is the first to examine respondent-level social network variables for LGB and heterosexual worried drinkers, and findings provide additional insight in to the complex relationships between social network variables and alcohol use patterns.

Ryan Wheeler 12/9/2016 11:13 AM Comment [51]: Is this the final copy?


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

19

Acknowledgments Kelly E. Green 12/9/2016 11:13 AM

This research was funded in part by a training grant from the National Institute on Alcohol Abuse and Alcoholism awarded to Barbara S. McCrady, Ph.D. (T32 AA07569). The author greatly appreciates Barbara McCrady, Ph.D., Elizabeth Epstein, Ph.D., Donald Bux Jr., Ph.D. and Diana Sanchez, Ph.D. for their guidance on this dissertation project. The author would like to thank Barbara and Beth for their ongoing mentorship throughout her graduate education and beyond.

Comment [52]: Does APA 6th include this in the Author Note section on the title page? Or do we still need this Acknowledgments page?


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

20

References Babor, T. F., Biddle-Higgins, J. C., Sauders, J. B., & Montiero, M. G. (2001). AUDIT: The Kelly E. Green 12/9/2016 11:13 AM

Alcohol Use Disorders Identification Test: Guidelines for Use in Primary Health Care.

Comment [53]: Does ref list need to be doublespaced for APA 6th?

Geneva, Switzerland: World Health Organization. Babor, Thomas F.; de la Fuente, J.R.; Saunders, J.; Grant, Marcus; World Health Organization Ryan Wheeler 12/9/2016 11:13 AM

Beatty, R.L., Geckle, M.O., Huggins, J., Kapner, C., Lewis, K., & Sandstrom, D.J. (1999). Gay men, lesbians, and bisexuals. In B.S. McCrady & E.E. Epstein (Eds.), Addictions: A comprehensive guidebook (pp. 542-551). New York: Oxford University Press. Beckman, L.J., & Amaro, H. (1986). Personal and social difficulties faced by women and men entering alcoholism treatment. Journal of Studies on Alcohol, 47, 135-145. Beckman, L.J., & Kocel, K.M. (1982). The treatment-delivery system and alcohol abuse in women: Social policy implications. Journal of Social Issues, 38, 139-151. Burckell, L.A., & Goldfried, M.R. (2005, November). Therapist qualities preferred by lesbian, gay male, and bisexual individuals. Presentation at the 39th annual convention of the Association for Behavioral and Cognitive Therapies, Washington D.C. Burgard, S.A., Cochran, S.D., & Mays, V.M. (2005). Alcohol and tobacco use patterns among heterosexually and homosexually experienced California women. Drug & Alcohol Dependence, 77, 61-70. Bux, D.A. (1996). The epidemiology of problem drinking in gay men and lesbians: A critical review. Clinical Psychology Review, 16, 277-298. CASAA Research Division (1995). Readiness Ruler. Available at: http://www.casaa.unm.edu/instr Cochran, S.D., Keenan, C., Schober, C., & Mays, V.M. (2000). Estimates of alcohol use and clinical treatment needs among homosexually active men and women in the U.S. population. Journal of Consulting and Clinical Psychology, 68, 1062-1071.

Comment [54]: Above citation was already here. Do you want this one too?


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

21

Collins, R.L., Park, G.E., & Marlatt, G.A. (1985). Social determinants of alcohol consumption: The effects of social interaction and model status on the self-administration of alcohol. Journal of Consulting and Clinical Psychology, 53, 189-200. Driscoll, R. (1982). A gay-identified alcohol treatment program: A follow-up study. Journal of Homosexuality, 7, 71-80. Finnegan, D.G., & McNally, E.B. (2002). Counseling Lesbian, Gay, Bisexual, and Transgender Substance Abusers: Dual Identities. New York: Haworth Press. Green, K.E., & Iverson, K.M. (2009). Computerized cognitive-behavioral therapy in a stepped care model of treatment. Professional Psychology: Research and Practice, 40, 96-103.. Kerr, Dianne L. et al. "Substance Use of Lesbian, Gay, Bisexual and Heterosexual College Students." American Journal of Health Behavior, vol. 38, no. 6, 2014, pp. 951-962, doi:10.5993/ajhb.38.6.17. Kelly E. Green 12/9/2016 11:13 AM

Kruse, M.I., Fromme, K., & Corbin, W.R. (2005). Improving the accuracy of self-report measures of drinking: Disaggregating quantity and frequency indices of alcohol consumption. Alcoholism: Clinical and Experimental Research, 29 supp., 118A.Kerr, D. L., Ding, K., & Chaya, J. (2014). Substance Use of Lesbian, Gay, Bisexual and Heterosexual College Students. American Journal of Health Behavior, 38(6), 951-962. doi:10.5993/ajhb.38.6.17 Lanfear, C., Akins, S., & Mosher, C. (2013). Examining the Relationship of Substance Use and Sexual Orientation. Deviant Behavior, 34(7), 586-597. doi:10.1080/01639625.2012.749149 McCabe, S. E., Hughes, T. L., Bostwick, W. B., West, B. T., & Boyd, C. J. (2009). Sexual orientation, substance use behaviors and substance dependence in the United States. Addiction, 104(8), 1333-1345. doi:10.1111/j.1360-0443.2009.02596.x McCabe, Sean Esteban et al. "Sexual Orientation, Substance Use Behaviors and Substance Dependence in the United States." Addiction, vol. 104, no. 8, 2009, pp. 1333-1345, doi:10.1111/j.1360-0443.2009.02596.x. Meyers, R.J., Miller, W.R., Smith, J.E., & Tonigan,J.S. (2002). A randomized trial of two methods for engaging treatment-refusing drug users through concerned significant others. Journal of Consulting and Clinical Psychology, 70, 1182-1185.

Comment [55]: These EndNote citations are not in APA 6th format.


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

22

Miller, W.R. & Tonigan, J.S. (1995). Barriers Questionnaire. Available from CASAA Research Division at: http://www.casaa.unm.edu/instr. Peavy, K.M., Cochran, B.N, Kelley, T., Markin, C., Schneider, K., & Gibson, L. (2004). What types of programs are offered for sexual minorities in substance abuse treatment? Presentation at the 38th Annual Convention of the Association for Advancement of Behavior Therapy, New Orleans, LA, November, 2004. Center for Behavioral Health Statistics and Quality. (2015). Behavioral health trends in the United States: Results from the 2014 National Survey on Drug Use and Health (HHS Publication No. SMA 15-4927, NSDUH Series H-50). Retrieved from http://www.samhsa.gov/data/ Ryan Wheeler 12/9/2016 11:13 AM

Saunders, J.B, Assland, O.G., Babor, T.F., De La Fuente, J.R., & Grant, M. (1993). Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO Collaborative Project on Early Detection of Persons with Harmful Alcohol ConsumptionII. Addiction, 81, 791-804. Schober, R. & Annis, H.M (1996). Barriers to help-seeking for change in drinking: A genderfocused review of the literature. Addictive Behaviors, 21, 81-92. Stall, R., Paul, J.P., Greenwood, G., Pollack, L.M., Bein, E., et al. (2001). Alcohol use, drug use and alcohol-related problems among men who have sex with men: the Urban Men’s Health Study. Addiction, 96, 1589-1601. Thom, B. (1986). Sex differences in help-seeking for alcohol problems – 1. The barriers to helpseeking. British Journal of Addiction, 81, 777-788.

Comment [56]: Do you want to inc. the website?


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

23

Table 1: Demographic Characteristics of Respondents Included in Analyses (n = 218) Characteristic

Sample Mean or Distribution

Age (n = 218)

M = 30.61 SD = 11.91

Gender (n = 218) Male Female Sexual Orientation (n = 218) Heterosexual Male Female Lesbian/Gay/Bisexual Male Female Ethnicity (n =218) Caucasian only African-American/Black Hispanic/Latino Asian/Pacific Islander Native American Other Missing Education (n = 218) Less than High School High School Only Trade School Some College 2-year College Degree 4-year College Degree Advanced Degree Employment Status (n = 218) Full-time employment Part-time employment Full-time student Homemaker Retired Unemployed (on disability) Unemployed Occupation (n = 214) Higher executives, major professionals Executives, less professionals Administrative, minor professionals Clerical, sales, technicians, servicemen Skilled manual employees Semiskilled employees Unskilled employees Student Homemaker

64 (29%) 154 (71%) 156 (72%) 48 (22%) 108 (50%) 62 (28%) 16 (7%) 46 (21%) 166 (76%) 7 (3%) 13 (6%) 21 (10%) 1 (1%) 6 (3%) 2 (1%) 1 (0.5%) 31 (14%) 2 (1%) 68 (31%) 16 (7%) 42 (19%) 58 (27%) 96 (44%) 31 (14%) 72 (33%) 7 (3%) 2 (1%) 2 (1%) 8 (4%) 19 (9%) 26 (12%) 29 (14%) 27 (12%) 5 (2%) 4 (2%) 4 (2%) 78 (36%) 6 (3%)


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS Retired Unemployed Others, not classifiable Household Income (n = 218) No answer $0 – 15,000 $15,001 – 30,000 $30,001 – 45,000 $45,001 – 60,000 $60,001 – 75,000 $75,001 – 100,000 $100,001 – 125,000 $125,001 – 150,000 $150,001 and above Region of Residence (n =216 ) Outside United States United States New England Middle Atlantic East North Central West North Central South Atlantic West South Central Mountain Pacific Missing Urbanicity (n = 218) Metropolitan/Urban Area Suburban Area Country/Rural Area Missing Relationship Characteristics In a Committed Relationship (n = 218) Married, Civil Union, or Commitment Ceremony (n = 150) With children or stepchildren (n =203)

24 2 (1%) 9 (4%) 5 (2%) 8 (4%) 47 (22%) 26 (12%) 26 (12%) 27 (12%) 19 (9%) 25 (12%) 15 (7%) 11 (5%) 14 (6%) 26 (12%) 190 (87%) 11 (6%) 30 (16%) 12 (6%) 9 (5%) 27 (14%) 15 (8%) 16 (8%) 17 (9%) 53 (28%) 94 (43%) 90 (41%) 33 (15%) 1 (0.5%) 152 (70%)

49 (23% of whole sample) 55 (25% of whole sample) Kelly E. Green 12/9/2016 11:13 AM Comment [57]: These numbers are off, and the formatting is different than other boxes


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

25

Table 2: Substance Use Differences and by Sex and Sexual Orientation Variable

Group

Mean (SD)

Statistics

Percent Drinking Days

Male (n = 58) Female (n = 137)

M = 45.02; SD = 33.63 M = 32.41; SD = 30.44

t(193) = 2.65, p = .011

Heterosexual (n = 140) LGB (n = 55)

Mean Drinks Per Drinking Day Mean Drinks Per Week

Male (n = 53) Female (n = 127) Heterosexual (n = 126) LGB (n = 54) Male (n = 60) Female (n = 145) Heterosexual (n = 147) LGB (n = 58)

M = 35.71; SD = 31.61 M = 37.28; SD = 32.77 M = 4.43; SD = 2.34 M = 3.82; SD = 2.34 M = 4.04; SD = 2.32 M = 3.91; SD = 2.44 M = 15.69; SD = 15.64 M = 9.46; SD = 12.41

Kelly E. Green 12/9/2016 11:13 AM

t(193) = -.31, ns

t(178) = 1.60, ns t(178) = .33, ns t(203) = 2.75, p = .007 t(203) = -.09, ns

Male (n = 64) Female (n = 154)

M = 11.23; SD = 13.11 M = 11.42; SD = 15.18 M = 12.23; SD = 7.50 M = 9.68; SD = 7.68 M = 10.49; SD = 7.90 M = 10.29; SD = 7.24 41 (70%) 121 (82%)

t(216) = .17, ns

Lifetime Drug Use

Heterosexual (n = 156) LGB (n = 62) Male (n = 59) Female (n = 147)

109 (74%) 50 (89%)

Z = -2.40 p = .02

Past Year Drug Use

Heterosexual (n = 148) LGB (n = 56) Male (n = 58) Female (n = 138)

33 (57%) 76 (55%)

Z = 0.24, ns

72 (50%) 35 (67%)

Z = 2.104, p = .035

Past Month Drug Use

Heterosexual (n = 143) LGB (n = 52) Male (n = 55) Female (n = 134)

21 (38%) 47 (35%)

Z = 0.40, ns

Heterosexual(n = 138) LGB (n = 50)

45 (33%) 21 (42%)

Z = -1.192, ns

AUDIT Score

t(216) = 2.25, p = .025

Z = -2.03, p = .04

Comment [58]: I'd like to change "male" to "men and "female" to "women" for all of these tables and in the text


Running head: SUBSTANCE USE IN LGB SOCIAL NETWORKS

26


Turn static files into dynamic content formats.

Create a flipbook
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.