Styles of analysis of GENACIS data: the relation between alcohol - - PowerPoint PPT Presentation

▶
styles of analysis of genacis data the relation between
SMART_READER_LITE
LIVE PREVIEW

Styles of analysis of GENACIS data: the relation between alcohol - - PowerPoint PPT Presentation

May 02, 2023 599 likes •985 views

Styles of analysis of GENACIS data: the relation between alcohol consumption and alcohol problems Robin Room AER Centre for Alcohol Policy Research, Turning Point Alcohol & Drug Centre, and School of Population Health, University of Melbourne

slide-1
SLIDE 1

Styles of analysis of GENACIS data: the relation between alcohol consumption and alcohol problems

Robin Room

AER Centre for Alcohol Policy Research, Turning Point Alcohol & Drug Centre, and School of Population Health, University of Melbourne

Kate Graham

Centre for Addiction and Mental Health, London, Ontario and Department of Psychology, University of Western Ontario and National Drug Research Institute, Curtin University, Western Australia Presentation at GENACIS plenary, June 1, 2008, Victoria, British Columbia

slide-2
SLIDE 2

Relationships:

predictors  consumption  problems

  • A. Predictors:

demographics, social context, personality ...

  • B. Consumption:

volume, heavy

  • ccasions
  • C. Alcohol problems:

chronic health, injury, social problems

  • 1. BC: Relation of amount & patterns of consumption to alcohol problems
  • 2. AC: relation of predictors to alcohol problems
  • 3. AC controlling for B: relation of predictors to the risk of problems for a given

consumption

  • 4. BC controlling for A: context etc. intermediating relation of consumption

& problems

slide-3
SLIDE 3

BC

Relation of amount & patterns of consumption

 alcohol problems

slide-4
SLIDE 4

1.a. Relation of amount and pattern of consumption to problems: problem score

Pia Mäkelä et al., Drinking habits in the Nordic Countries. Oslo: SIFA Report No. 2/99, 1999.

slide-5
SLIDE 5

1.b. Relation of amount and pattern of consumption to problems: number of harms from drinking

Room, R., Bondy, S.J. & Ferris, J. The risk of harm to oneself from drinking, Canada, 1989. Addiction 90:499‐513, 1995. (mislabeled: should be “Percent reporting each harm”)

slide-6
SLIDE 6

How much drinking is too much? It depends on the problem

(examples of different patterns of relationship for

headache/nausea, finances and fighting)

From Graham, K. (March, 2007). Drinking and drinking problems among Canadian men. The Canadian Conference on Men’s Health: Raising the Standard, Victoria, BC.

1.c. Relation of amount and pattern of consumption to problems: different aspects of drinking  specific types of problem consequences

slide-7
SLIDE 7

Percent of men who had a headache or nausea after drinking

10 20 30 40 50 60 70 80 90 100

1 to <2 2 to <3 3 to <4 4 to <5 5 to <6 6 to <7 7 or more

10 20 30 40 50 60 70 80 90 100

1 to <2 2 to <3 3 to <4 4 to <5 5 to <6 6 to <7 7 to <8 8 to <9 9 to <10 10 to <11 11 or more

10 20 30 40 50 60 70 80 90 100

never 5+ <1 month 1-3 days/month 1-2 days/week 3-4 days/week 5-7 days/week

usual number of drinks per occasion maximum number of drinks per occasion frequency of drinking 5 or more drinks

slide-8
SLIDE 8

Percent of men reporting harmful effects on finances from drinking

5 10 15 20 25 30 35 40

1 to <2 2 to <3 3 to <4 4 to <5 5 to <6 6 to <7 7 or more

5 10 15 20 25 30 35 40

1 t

  • <

2 2 t

  • <

3 3 t

  • <

4 4 t

  • <

5 5 t

  • <

6 6 t

  • <

7 7 t

  • <

8 8 t

  • <

9 9 t

  • <

1 1 t

  • <

1 1 1 1

  • r

m

  • r

e

5 10 15 20 25 30 35 40

never 5+ <1 month 1-3 days/month 1-2 days/week 3-4 days/week 5-7 days/week

usual number of drinks per occasion maximum number of drinks per occasion frequency of drinking 5 or more drinks

slide-9
SLIDE 9

Percent of men who had gotten into a physical fight when drinking

5 10 15 20

1 to <2 2 to <3 3 to <4 4 to <5 5 to <6 6 to <7 7 or more

5 10 15 20

1 to <2 2 to <3 3 to <4 4 to <5 5 to <6 6 to <7 7 to <8 8 to <9 9 to <10 10 to <11 11 or more 5 10 15 20

never 5+ <1 month 1-3 days/month 1-2 days/week 3-4 days/week 5-7 days/week

usual number of drinks per occasion maximum number of drinks per occasion frequency of drinking 5 or more drinks

slide-10
SLIDE 10

1.d. Relation of amount and pattern of consumption to problems: amount crossed with pattern

Room, R., Bondy, S.J. & Ferris, J. The risk of harm to oneself from drinking, Canada, 1989. Addiction 90:499‐513, 1995.

slide-11
SLIDE 11

A  C

Relation of nondrinking predictors

 alcohol problems

slide-12
SLIDE 12

2.a. Relation of predictors to alcohol problems ‐‐ gender and age

Graham, K., Demers, A., Bernards, S. with Dell, C., George, A., Kairouz, S., Nadeau, L., Poulin, C., Rehm, J., Wells, S. (November, 2005). Regional, gender and demographic variations in alcohol problems in Canada: Results from the GENACIS Canada Survey. Paper presented at “Issues of Substance” Canadian Centre on Substance Abuse National Conference, Markham, Ontario.

slide-13
SLIDE 13

Percent of all respondents scoring 8 or more on the AUDIT by gender and age group (abstainers and light infrequent=0)

10 20 30 40 50 60 70 80 90 100

18 to 20 21 to 25 26 to 30 31 to 35 36 to 40 41 to 45 46 to 50 51 to 55 56 to 60 61 to 65 66 to 70 71 to 76

Percentage male female

slide-14
SLIDE 14

Had enough to drink so that you felt the effects of the alcohol for example, your speech was slurred or you had trouble walking steadily (by gender and age group)

10 20 30 40 50 60 70 80 90 100

18 to 20 21 to 25 26 to 30 31 to 35 36 to 40 41 to 45 46 to 50 51 to 55 56 to 60 61 to 65 66 to 70 71 to 76

Percentage male female

slide-15
SLIDE 15

10 20 30 40 50 60 70 80 90 100

18 to 20 21 to 25 26 to 30 31 to 35 36 to 40 41 to 45 46 to 50 51 to 55 56 to 60 61 to 65 66 to 70 71 to 76

Percentage male female

Found that you were not able to stop drinking once you had started (by gender and age group)

slide-16
SLIDE 16

10 20 30 40 50 60 70 80 90 100

18 to 20 21 to 25 26 to 30 31 to 35 36 to 40 41 to 45 46 to 50 51 to 55 56 to 60 61 to 65 66 to 70 71 to 76

Percentage male female

The following persons tried to get you to cut down on your drinking –partner, child or health worker/doctor (by gender and age group)

slide-17
SLIDE 17

10 20 30 40 50 60 70 80 90 100

18 to 20 21 to 25 26 to 30 31 to 35 36 to 40 41 to 45 46 to 50 51 to 55 56 to 60 61 to 65 66 to 70 71 to 76

Percentage male female

Tried to cut down or quit drinking (by gender and age group)

slide-18
SLIDE 18

10 20 30 40 50 60 70 80 90 100

18 to 20 21 to 25 26 to 30 31 to 35 36 to 40 41 to 45 46 to 50 51 to 55 56 to 60 61 to 65 66 to 70 71 to 76

Percentage male female

Driven after having 2 or more drinks in the previous hour (by gender and age group)

slide-19
SLIDE 19

10 20 30 40 50 60 70 80 90 100

18 to 20 21 to 25 26 to 30 31 to 35 36 to 40 41 to 45 46 to 50 51 to 55 56 to 60 61 to 65 66 to 70 71 to 76

Percentage male female

Gotten into a physical fight while drinking (by gender and age group)

slide-20
SLIDE 20

A │ BC

Relation of different measures of consumption

 problems in predictor categories

slide-21
SLIDE 21
  • 2b. Relation of different measures of consumption

to problems in predictor categories – example of correlations between different measures of drinking and depression for males compared to females

Graham, K., Massak, A., Demers, A, & Rehm, J. (2007) Does the association between alcohol consumption and depression depend on how they are measured? Alcoholism: Clinical and Experimental Research 31:78‐88.

slide-22
SLIDE 22

Depression and drinking:

  • 1. Average correlation between depression and four types of alcohol measures

(frequency, volume, usual and maximum quantity and heavy episodic drinking) for males and females

0.00 0.10 0.15

  • 0.04

0.06 0.15 0.21 0.08

  • 0.1
  • 0.05

0.05 0.1 0.15 0.2 0.25 Frequency Volume Quantity HED Male Female

slide-23
SLIDE 23

Percent of males and females who met the criteria for a diagnosis of major depression based on the CIDI by usual quantity per drinking day for past 12 months

8.1 12.0 8.2 10.6 16.6 13.9 19.4 21.2 24.5 3.9 8.1 7.4 5.1 5 4.9 4.4 5.9 5.0

5 10 15 20 25 lifetime abstainer former drinker 1 2 3 4 5 6 7 or more drinks Usual quantity (past year)

% meeting criteria for depression

Male Female

slide-24
SLIDE 24

Percent of males and females who met the criteria for a diagnosis of major depression based on the CIDI by frequency of drinking 5 or more drinks per occasion in the past year

19.4 30.8 7.7 7.3 4.7 6.5 4.0 5.9 3.9 28.6 17.5 17.1 13.9 8.7 12.0 5.0

5 10 15 20 25 30 35 lifetime abstainer former drinker drinker but never +5 less than once a month 1-3 days a month 1-2 days a week 3-4 days a week everyday or nearly everyday Frequency of drinking 5+

% meeting criteria for depression

Male Female

slide-25
SLIDE 25

AC controlling for B Relation of predictors to the risk of problems at a given consumption

slide-26
SLIDE 26

3.a. Relation of predictors to the risk of problems for a given consumption: ratio of harm per litre/per binge

Hazardous behaviour score per drinking volume and occasions drinking five or more drinks, by age and sex (ratio with males aged 40 to 44 as baseline), National Drug Strategy Household Survey, 2004 (Livingston & Room, working paper, 2008)

0.5 1 1.5 2 2.5 3 12 to 14 15 to 17 18 to 19 20 to 24 25 to 29 30 to 34 35 to 39 40 to 44 45 to 49 50 to 54 55 to 59 60 to 64 65 to 69 70 to 74 75 to 79 8 Age group Ratio (40-44 year old males as base) Hazardous behaviour per standard drink, male Hazardous behaviour per standard drink, female Hazardous behaviour per risky drinking occasion, male Hazardous behaviour per risky drinking occasion, female

slide-27
SLIDE 27

3.b. AC controlling for B:

does the relation between a predictor and rate of problems disappear when consumption is controlled?

Gender differences in problem rates controlling for alcohol consumption

– Gender difference disappears for some problems but not others

Graham, K. (March, 2007). Drinking and drinking problems among Canadian men. The Canadian Conference on Men’s Health: Raising the Standard, Victoria, BC.

slide-28
SLIDE 28

Odds of experiencing problem consequences from drinking for men compared to women

1 2 3

Felt guilt or remorse after drinking? Had enough to drink that you had a headache or felt nauseated Been unable to remember what happened the night before because you had been drinking Tried to cut down or quit drinking? Felt sick or found yourself shaking when you cut down or stopped drinking Unable to stop drinking once you had started

The extended bar shows the raw odds; the dark green bar shows the odds ratio controlling for drinking pattern and context

slide-29
SLIDE 29

Odds of reporting perceived harmful effect of drinking for men compared to women

1 2 3

Marriage/intimate relationships Work, studies or employment

  • pportunities

Physical health, not counting hangovers Finances Housework or chores around the house Outlook on life Friendships Relationships with family

  • ther than spouse/partner

The extended bar shows the raw odds; the dark green bar shows the odds ratio controlling for drinking pattern and contextv

slide-30
SLIDE 30

Odds ratios for consequences that were more likely to be reported by men than by women controlling for level of alcohol consumption 1 2 3 4 5 6 7 8

Driven after having 2 or more drinks in the previous hour Gotten into a physical fight while drinking Needed a drink in the morning after a heavy drinking session

The extended bar shows the raw odds; the dark green bar shows the odds ratio controlling for drinking pattern and context

slide-31
SLIDE 31

BC controlling for A

Relationship between consumption and problems controlling for predictors such as demographics and context

slide-32
SLIDE 32

4.a. BC controlling for A: Relationship between consumption and specific types of problems (drinking to intoxication, guilt or remorse, driving after drinking and trying to cit down or quit) ‐ controlling for (a) drinking context and (b) respondent characteristics

Graham, K., Demers, A., Bernards, S. with Dell, C., George, A., Kairouz, S., Nadeau, L., Poulin, C., Rehm, J., Wells, S. (November, 2005). Regional, gender and demographic variations in alcohol problems in Canada: Results from the GENACIS Canada Survey. Paper presented at “Issues of Substance” Canadian Centre on Substance Abuse National Conference, Markham, Ontario.

slide-33
SLIDE 33

Drinking pattern, drinking context and respondent characteristics predicting consequences of drinking to intoxication (odds ratios)

Alcohol consumption variables Frequency of drinking (days per year) 1.003*** Usual number of drinks consumed 1.080*** Maximum number of drinks consumed in a day 1.355*** Drinking context variables Proportion of alcohol consumed that was wine .959 Proportion of drinking done with meals .876† Proportion of drinking done at a bar 1.660*** Proportion of drinking done at a restaurant 1.175† Respondent characteristics Male gender (female comparison) .839** Age .956*** Level of education 1.058*** Whether born in Canada (not born in Canada comparison) 1.398*** Population of place of residence of respondent 1.028***

† p<.10, *p<.05, **p<.01, ***p<.001

slide-34
SLIDE 34

Drinking pattern, drinking context and respondent characteristics predicting feelings of guilt or remorse after drinking

Alcohol consumption variables Frequency of drinking (days per year) 1.003*** Usual number of drinks consumed 1.065*** Maximum number of drinks consumed in a day 1.093*** Drinking context variables Proportion of alcohol consumed that was wine .826 Proportion of drinking done with meals .881 Proportion of drinking done at a bar 1.571*** Proportion of drinking done at a restaurant .927 Respondent characteristics Male gender (female comparison) .702*** Age .979*** Level of education 1.044*** Whether born in Canada (not born in Canada comparison) 1.272* Population of place of residence of respondent

1.022†

† p<.10, *p<.05, **p<.01, ***p<.001

slide-35
SLIDE 35

Drinking pattern, drinking context and respondent characteristics predicting driving after drinking

Alcohol consumption variables Frequency of drinking (days per year)

1.004***

Usual number of drinks consumed

.995

Maximum number of drinks consumed in a day

1.099***

Drinking context variables Proportion of alcohol consumed that was wine

.765*

Proportion of drinking done with meals

1.022

Proportion of drinking done at a bar

1.676***

Proportion of drinking done at a restaurant

1.456***

Respondent characteristics Male gender (female comparison)

2.163***

Age

1.003

Level of education

1.019†

Whether born in Canada (not born in Canada comparison)

1.340**

Population of place of residence of respondent

.973**

† p<.10, *p<.05, **p<.01, ***p<.001

slide-36
SLIDE 36

Drinking pattern, drinking context and respondent characteristics predicting trying to cut down or quit drinking

Alcohol consumption variables Frequency of drinking (days per year)

1.004***

Usual number of drinks consumed

1.061***

Maximum number of drinks consumed in a day

1.045***

Drinking context variables Proportion of alcohol consumed that was wine

.913

Proportion of drinking done with meals

.811*

Proportion of drinking done at a bar

1.053

Proportion of drinking done at a restaurant

.939

Respondent characteristics Male gender (female comparison)

.831*

Age

.988***

Level of education

.953***

Whether born in Canada (not born in Canada comparison)

.985

Population of place of residence of respondent

1.008

† p<.10, *p<.05, **p<.01, ***p<.001

slide-37
SLIDE 37

Some conclusions

  • Relationship between consumption and problems varies with the measure of

consumption (and the measure & type of problem) – multicollinearity between consumption measures as an issue

  • Predictors of problems may or may not be the same as predictors of

consumption

  • Controlling consumption may “wipe out” relations between predictors and

problems Two step prediction (Lee Robins): demographics predict heavy drinking; personal history/personality predicts who among heavy drinkers get in trouble?? Whether you want to control for consumption depends on intended use fo data:

You would not want to control if the objective is to identify high risk groups You would want to control if the objective is to understand the relationship between demographics etc. and problem consequences from drinking

‐‐

slide-38
SLIDE 38

Some conclusions continued ‐‐ “Trouble per litre” or per heavy occasion as a way of presenting differential risk of harm for different demographics $ tendencies: gender differences washed out, age differences remain ‐‐ or show different regression models side by side, with and without consumption $ Controlling for context, etc. may intermediate relation between consumption and problems $ There are a number of ways of presenting results from analyses across the three domains