Web appendix
International Journal of Public Health
The effects of alcohol-related harms to others on self perceived mental well-being in a Canadian sample
Authors:
Candace Lewis-Laietmark 1, 2 (),
Ashley Wettlaufer 2 (),
Kevin D Shield 2,3 (),
Norman Giesbrecht 2,4 ()
Nicole April 5 ()
Mark Asbridge 6 ()
Colleen Dell 7 ()
Jürgen Rehm 2,3,4,8 ()
Tim Stockwell 9 ()
Affiliations:
1) School of Community Health, Portland State University, Portland, Oregon, United States
2) Social and Epidemiological Research Department, Centre for Addiction and Mental Health (CAMH), Toronto, Ontario, Canada
3) Institute of Medical Science, University of Toronto, Toronto, Ontario, Canada
4) Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada
5) Institut national de santé publique du Québec, Québec City, Québec, Canada
6) Dalhousie University, Department of Community Health and Epidemiology and Department of Emergency Medicine, Halifax, Nova Scotia, Canada
7) University of Saskatchewan, Saskatoon, Saskatchewan, Canada
8) Institute of Clinical Psychology and Psychotherapy, Technische Universität Dresden, Germany
9) Centre for Addictions Research of BC, University of Victoria, Victoria, BC, Canada
Corresponding author and author designated to check proofs:
Norman Giesbrecht, PhD. Centre for Addiction & Mental Health, Toronto, Ontario M5S2S1 Canada. Email: ; telephone: 416-535-8501 x 36895; fax 416-595-6899
Survey methodology
Recruitment took place on weekdays and Sundays. For the weekday dialing, calls were attempted during the day from 11:00 am to 3:00 pm and during the evening from 5:00 pm to 9:00 pm. In total 5,174 numbers were called; 2,156 of these numbers were eligible to participate, 1,600 of these numbers were not eligible to participate, and 1,272 of these numbers had an unknown eligibility (i.e. no answer was obtained during the call attempts). Of the numbers that were eligible to participate, 384 people were interviewed for the survey (7.4% of all numbers). The number of calls attempted before a household participated in the study, taking into account the phone numbers where the reviewer received no answer/answering machine/busy, are outlined in Figure A1. Table A1 outlines the numbers called by eligibility status. Data on response rates were not available by province.
The survey was fairly representative of Canadians 18 years of age and older when compared to the Canadian Census (see Table A2); however, level of educational attainment was skewed towards those with a higher education when compared to the Canadian population. Data for the Canadian population were obtained from the 2006 Canadian census, the 2011 Canadian census and the 2011 National Household survey and were imputed for 2012 by Statistics Canada [Statistics Canada, 2016a; Statistics Canada 2016b; Statistics Canada 2016c; Statistics Canada 2016d; Statistics Canada 2016e]. Furthermore, the sample of people by province was in all cases above 30 individuals (see Table A3) and, therefore, provincial comparisons do not violate small sample size requirements (i.e. samples below 30 individuals) (Groves 2004).
Survey introduction
“Good morning/afternoon/evening, my name is <___ >. I’m calling from the University of Waterloo Survey Research Centre on behalf of the Centre for Addiction and Mental Health, which is conducting a survey about health issues. On average, the survey will take about 15 minutes to complete. We are not asking for money or selling anything.”
Figure A1. Number of calls attempted before a household participated in the study, taking into account the phone numbers where the reviewer received no answer/answering machine/busy*, **.
* No phone number received the maximum number of call-backs
** Takes into account the estimated fraction of non-respondents that would not be eligible or would have refused to take part in the survey
Table A1. The dispositions for all call attempt records by phone number status
Phone number status / n / PercentInterviewed / 384 / 7.4%
Eligible, non-interview / 1,772 / 34.2%
Refusal and break-off, eligible / 1,757 / 34.0%
Respondent not available during data collection time period / 15 / 0.3%
Unknown eligibility / 1,272 / 24.6%
Not eligible / 1,600 / 30.9%
Total calls / 5,028
Table A2. Socio-demographic and mental health characteristics of survey respondents
Socio-demographic factors / Men / Women / TotalHTO survey / 2012 Canada / HTO survey / 2012 Canada / HTO survey / 2012 Canada
Province
British Columbia / 14.6% / 16.5% / 16.6% / 16.3% / 15.5% / 16.4%
Nova Scotia / 3.6% / 3.4% / 3.4% / 3.5% / 3.5% / 3.4%
Ontario / 46.9% / 47.0% / 42.7% / 47.7% / 45.0% / 47.4%
Quebec / 31.5% / 29.3% / 32.6% / 28.9% / 32.0% / 29.1%
Saskatchewan / 3.5% / 3.8% / 4.7% / 3.6% / 4.0% / 3.7%
Age (years)
18-34 / 37.8% / 29.9% / 22.9% / 28.4% / 30.8% / 29.1%
35-44 / 16.3% / 17.1% / 15.1% / 16.5% / 15.7% / 16.8%
45-64 / 34.6% / 36.1% / 40.5% / 34.9% / 37.4% / 35.5%
65+ / 11.2% / 16.9% / 21.5% / 20.2% / 16.1% / 18.6%
Gender
Male / - / - / - / - / 53.1% / 49.1%
Female / - / - / - / - / 46.9% / 50.9%
Highest Education Attainment**
Secondary School Graduation or Less / 24.8% / - / 29.0% / - / 26.8% / 49.3%
Some College or Some Trade School / 16.4% / - / 13.9% / - / 15.2% / 28.1%
Diploma/Certificate for Trade School/Community College/ CEGEP / 18.6% / - / 24.5% / - / 21.4%
Bachelor Degree or Higher / 40.2% / - / 32.6% / - / 36.7% / 22.6%
Marital Status*
Married/ Living with a partner/common-law marriage / 50.3% / 60.6% / 63.7% / 57.6% / 56.6% / 59.1%
Widowed, Separated or Divorced / 11.5% / 10.6% / 18.5% / 19.7% / 14.8% / 15.2%
Single, never married / 38.2% / 28.8% / 17.8% / 22.8% / 28.6% / 25.7%
Occupational Status
Employed part-time, full-time or self-employed / 70.9% / - / 58.6% / - / 65.2% / 61.7%
Retired / 14.9% / - / 22.2% / - / 18.3% / 38.3%
Not working / 12.0% / - / 15.2% / - / 13.5%
Other/refused / 2.3% / - / 3.9% / - / 3.0% / -
Household Income**
Less than $20,000 / 7.5% / - / 6.7% / - / 7.1% / 13.3%
$20,000-$49,999 / 23.6% / - / 36.2% / - / 29.5% / 27.6%
$50,000-$79,999 / 24.7% / - / 27.2% / - / 25.9% / 22.3%
$80,000-$99,999 / 5.2% / - / 9.5% / - / 7.2% / 10.9%
$100,000 + / 39.1% / - / 20.4% / - / 30.3% / 25.9%
* Data for Canada is for 2006 from the 2006 Canadian census
** Data for Canada is for 2010 from the 2011 National Household survey
Table A3. Survey sample (n) by province and gender
Socio-demographic factors / Men / Women / TotalProvince
British Columbia / 32 / 26 / 58
Nova Scotia / 34 / 25 / 59
Ontario / 32 / 23 / 55
Quebec / 39 / 28 / 67
Saskatchewan / 41 / 22 / 63
Power analysis
The power of a chi-square test given a total sample size of 302 and an average design effect of 3 was assessed using a Cohen’s w of 0.1 (small effect), 0.3 (medium effect) and 0.5 (large effect) and various degrees of freedom (see Table A4). The power to detect a difference in mental-well being given a total sample size of 302, and a design effect of 3 through a Quasi-Poisson regression model, was estimated using 10,000 simulations assuming a normal mental well-being score of 8.40 (standard deviation of 1.43) and varying mental health scores in people who are affected by harms to others (assuming the same standard deviations as in the normal mental health score), and varying degrees of the prevalence of harm (see Table A5).
Table A4. Power analysis (chi-squared) assuming a sample size of 302 and an average design effect of 3
Degrees of Freedom / Small effect (w = 0.1) / Medium effect (w = 0.3) / Large effect (w = 0.5)2 / 0.13 / 0.78 / >0.99
3 / 0.12 / 0.72 / 0.99
4 / 0.11 / 0.67 / 0.99
5 / 0.10 / 0.63 / 0.99
Table A5. Power analysis (Quasi-Poisson regression) assuming a sample size of 302 and an average design effect of 3, a normal mental well-being score of 8.40 (standard deviation of 1.43)
Mental well-being score (for people experiencing a harm) / Prevalence (%) of the harm5% / 10% / 20% / 30% / 40% / 50%
8.40 (REF) / - / - / - / - / - / -
8.20 / 0.06 / 0.07 / 0.08 / 0.08 / 0.11 / 0.09
8.00 / 0.08 / 0.12 / 0.19 / 0.24 / 0.26 / 0.25
7.80 / 0.16 / 0.23 / 0.35 / 0.47 / 0.51 / 0.53
7.60 / 0.24 / 0.37 / 0.56 / 0.72 / 0.77 / 0.80
7.40 / 0.32 / 0.54 / 0.78 / 0.87 / 0.92 / 0.93
7.20 / 0.45 / 0.67 / 0.89 / 0.97 / 0.98 / 0.98
7.00 / 0.56 / 0.81 / 0.97 / 0.99 / >0.99 / >0.99
References
Cohen, J. (1992). A power primer. Psychological bulletin, 112(1), 155.
Groves RM (2004) Survey errors and survey costs. John Wiley and Sons Ltd, Chichester, UK
Statistics Canada (2016a). Table 051-0001 - Estimates of population, by age group and sex for July 1, Canada, provinces and territories, annual (persons unless otherwise noted). Statistics Canada; Ottawa, Canada.
Statistics Canada (2016b). Table 051-0042 Estimates of population, by marital status or legal marital status, age and sex for July 1, Canada, provinces and territories, annual (persons). Statistics Canada; Ottawa, Canada.
Statistics Canada (2016c). Table 282-0089 Labour force survey estimates (LFS), employment by class of worker and sex, seasonally adjusted and unadjusted, monthly (persons x 1,000). Statistics Canada; Ottawa, Canada.
Statistics Canada (2016d). 2011 National Household Survey: Data tables. Statistics Canada; Ottawa, Canada.
Statistics Canada (2016e). Table 051-0042 Estimates of population, by marital status or legal marital status, age and sex for July 1, Canada, provinces and territories, annual (persons). Statistics Canada; Ottawa, Canada.