Edited by: Shervin Assari, University of Michigan, USA
Reviewed by: Arezoo Shajiei, Mashhad University of Medical Sciences, Iran; Navvab Shamspour, Iranain Red Crescent Society, Iran; Mahshid Taj, World Health Organization, Egypt; Anahita Bassirnia, Mount Sinai Beth Israel, USA
This article was submitted to Public Mental Health, a section of the journal Frontiers in Public Health.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
In adolescents and young adults, acute consequences like injuries account for a substantial proportion of alcohol-related harm, especially in risky single-occasion (RSO) drinkers. The primary aim of the study was to characterize different drinking profiles in RSO drinkers according to drinking locations and their relationship to negative, alcohol-related consequences. The sample consisted of 2746 young men from the Cohort Study on Substance Use Risk Factors who had reported drinking six or more drinks on a single-occasion at least monthly over the preceding 12 months. Principal component analysis on the frequency and amount of drinking at 11 different locations was conducted, and 2 distinguishable components emerged: a
From the public health perspective, risky single-occasion drinking (RSOD) is considered one of the major problems stemming from alcohol consumption among adolescents and young adults. RSOD has been consistently identified as a stronger predictor of negative alcohol-related consequences among young adults than total drinking volume (
To investigate the relationship between drinking and alcohol-related consequences, not only drinking pattern – including the frequency and volume of drinking over a given time period – but also the demographics (i.e., male gender, age), psychological state (i.e., personality, attitudes), and social context (i.e., peer-influence, drinking locations) of drinkers must be considered (
The primary aim of the present study was to characterize drinking profiles according to drinking location and their differential impact on negative alcohol-related consequences in a population-based Swiss sample. It can be assumed that specific drinking locations and occasions are associated with a lack of supervision and social control, thereby raising the likelihood of negative consequences. Findings should have further implications for public health interventions aimed at lowering problematic alcohol consumption in young people, which is one of the main strategic goals of the National Alcohol Program 2013–2016 of the Swiss Federal Office of Public Health (
All analyses conducted for this study were based on data collected for the Cohort Study on Substance Use Risk Factors (C-SURF). The study design and study sample for C-SURF have been detailed elsewhere (
Out of the 15,074 young men who showed up at the recruitment centers, 1829 (12.1%) were not seen by the research staff. Among the 13,245 conscripts who were informed about the study, 7563 (57.1%) gave written informed consent to participate, among whom 5990 (79.2%) subsequently filled out the baseline questionnaire. For the purposes of this study, only those men who reported having consumed six or more standard drinks on a single-occasion at least monthly over the preceding 12 months were included in analysis [2746 (45.8%) of the 5990 C-SURF participants], as risky single-occasion (RSO) drinking contributes substantially to the risk of alcohol-related harm in young men. Fifty participants had to be excluded due to missing data on drinking locations. Hence, ultimately, the sample consisted of 2696 subjects.
Drinking volume (the average number of drinks per week) was calculated by multiplying the usual frequency (rescaled to days per week) and the usual quantity (number of standard drinks per occasion) of the past 12 months. RSOD was defined in accordance with Murgraff et al. (
Drinking locations were categorized according to the 2000 New Zealand National Alcohol Tracking Survey (
Participants were asked whether or not they had experienced negative consequences associated with drinking alcohol over the previous 12 months. To date, no standardized grouping of consequences from drinking exists for epidemiological surveys (
The following sociodemographic factors were assessed by means of self-report:
This personality trait was measured using the Brief Sensation Seeking Scale (BSSS) (
Three different personality traits were measured, in accordance with the Zuckerman–Kuhlmann Personality scale (ZKPQ-50-cc) (
Finally, subjects were asked whether any of their closest friends had what they would call a “significant drinking problem” – meaning one that did or should have led to treatment. Answer categories were as follows: “no one,” “one or two,” “some of them,” and “most of them.”
Continuous variable data are presented as medians and inter-quartile ranges (IQR), and categorical variable data as percentages. Principle component analysis (PCA) of drinking location was conducted to identify variable combinations. Each of the resulting components was then classified into three categories, according to their factor scores. For this purpose, two cut-off values at the 33.3 and the 66.6 percentiles were applied, and the components divided into low (≤33.3 percentile), intermediate (>33.3 and ≤66.6 percentile), and high (>66.6 percentile), as per their factor scores. Logistic regression models were used to evaluate the associations between SAC and the drinking profiles identified by PCA. SAC served as the dependent variable, and the drinking profiles as independent. Adjustments were made for sociodemographic variables (age, type of residence, linguistic region, family affluence, highest achieved education of the participant, and highest achieved education of the father), personality factors (sensation seeking, anxiety/neuroticism, aggression/hostility, and sociability), and peer-influence in logistic regression models. Classification of cases served to evaluate model adequacy. Statistical analyses were performed using SPSS version 21.0.
To investigate whether drinking locations might cluster in a way as to identify different dimensions of drinking profiles, PCA with varimax rotation was conducted on the frequency and amount of drinking at the 11 different types of location. The factor analysis was exploratory, and two distinguishable components emerged. The two resulting components explained 56.1% of the total variance in the grouping items. The first component was termed as a
Item | Non-party- dimension | Party- dimension |
---|---|---|
Theater/cinema | 0.86 | 0.12 |
Sport clubs (e.g., football, hockey, gymnastics) | 0.82 | 0.24 |
Other clubs/societies (orchestra, choir, chess club etc.) | 0.78 | 0.25 |
Restaurants | 0.70 | 0.34 |
Sports events | 0.65 | 0.38 |
Pubs/bars | 0.13 | 0.74 |
Someone else’s home | 0.25 | 0.73 |
Discos/nightclubs | 0.21 | 0.69 |
Outdoor public places (e.g., parks, swimming pools, streets) | 0.25 | 0.62 |
Special events (e.g., festivals, street parties, exhibitions) | 0.16 | 0.58 |
Home | 0.30 | 0.51 |
The factor scores from the two resulting PCA components were each divided into three categories (low, intermediate, and high). Consequently, subjects were allocated to one of four profile categories: (1) subjects who had low to intermediate factor scores for both the
Baseline characteristics | LL | NH | PH | HH |
---|---|---|---|---|
( |
( |
( |
( |
|
Rural | 669 (60.0) | 476 (69.7) | 425 (62.3) | 154 (71.3) |
Urban | 446 (40.0) | 207 (30.3) | 257 (37.7) | 62 (28.7) |
German | 526 (47.2) | 380 (55.6) | 254 (37.2) | 103 (47.7) |
French | 589 (52.8) | 303 (44.4) | 428 (62.8) | 113 (52.3) |
Above average income | 517 (46.4) | 333 (48.8) | 308 (45.2) | 99 (45.8) |
Average income | 455 (40.8) | 259 (37.9) | 297 (43.5) | 89 (41.2) |
Below average income | 143 (12.8) | 91 (13.3) | 77 (11.3) | 28 (13.0) |
Primary school | 581 (52.8) | 340 (50.7) | 310 (46.5) | 95 (44.4) |
Higher vocational school | 266 (24.1) | 208 (31.0) | 217 (32.5) | 77 (36.0) |
High school/university | 254 (23.1) | 123 (18.3) | 140 (21.0) | 42 (19.6) |
No secondary education | 89 (8.1) | 49 (7.3) | 44 (6.5) | 15 (7.0) |
Secondary education | 570 (51.6) | 374 (55.4) | 370 (54.7) | 123 (57.2) |
Tertiary education | 445 (40.3) | 252 (37.3) | 263 (38.8) | 77 (35.8) |
Personality traits | LL | NH | PH | HH |
---|---|---|---|---|
(median; IQR) | ( |
( |
( |
( |
Sensation seeking | 3.3 (2.9–3.8) | 3.1 (2.6–3.8) | 3.6 (3.0–4.0) | 3.6 (3.0–4.3) |
Anxiety/neuroticism | 1 (0–3) | 1 (1–3) | 1 (0–3) | 1 (0–3) |
Aggression/hostility | 4 (3–6) | 4 (3–6) | 5 (3–7) | 5 (3–7) |
Sociability | 7 (5–8) | 7 (5–8) | 7 (5–8) | 7 (6–8) |
Almost half (48.9%) of our subjects reported having experienced SAC within the previous 12 months (42.7% among those within the LL profile group, 39.3% among NH, 63.1% among PH, and 66.2% among those within the HH profile group). The most frequently mentioned consequences within the LL profile group were accident or injury (16.8%), unintended sexual intercourse (15.5%), and arguments or fights (14.6%). Corresponding figures were 14.0, 17.8, and 16.3% within the NH profile group, 28.9, 27.5, and 28.2% within the PH profile group, and 31.5, 33.3, and 32.4% within the HH profile group.
Analysis revealed differential impacts of drinking location profiles on SAC among young adults who engaged in RSO within the preceding 12 months. Relative to those classified as low or intermediate for both dimensions, no significant difference in SAC was detected among those classified as high for the
In order to rule out the possibility that the higher odds of experiencing SAC was solely explained by more drinks being consumed at specific locations, we further adjusted for drinking volume in a regression model. The median (IQR) drinking volumes for the LL, NH, PH, and HH profile groups were 8 (
In a second model, the association between the drinking location profiles and SAC was further adjusted for the sociodemographic factors – age, linguistic region, residence, highest achieved education of the participant, family affluence, and highest achieved education of the father (Table
OR (95% CI) | ||
---|---|---|
Drinking profiles |
||
LL | 1.00 | |
NH | 0.83 (0.68–1.01) | 0.062 |
PH | 1.79 (1.45–2.21) | <0.001 |
HH | 1.94 (1.40–2.69) | <0.001 |
RSOD |
||
Monthly | 1.00 | |
Weekly | 1.61 (1.35–1.93) | <0.001 |
Daily | 2.74 (1.54–4.87) | 0.001 |
Drinking volume (standard drinks per week) | 1.01 (1.00–1.02) | 0.043 |
Age | 1.03 (0.87–1.23) | 0.702 |
Residence | ||
Rural | 1.00 | |
Urban | 1.16 (0.98–1.38) | 0.087 |
Linguistic region | ||
German | 1.00 | |
French | 1.06 (0.88–1.26) | 0.543 |
Family affluence | ||
Above average income | 1.00 | |
Average income | 1.03 (0.80–1.34) | 0.810 |
Below average income | 0.93 (0.72–1.20) | 0.564 |
Education participant | ||
Primary school | 1.00 | |
Higher vocational school | 1.11 (0.89–1.39) | 0.346 |
High school/university | 1.20 (0.94–1.51) | 0.140 |
Education father | ||
No secondary education | 1.00 | |
Secondary education | 1.13 (0.81–1.57) | 0.472 |
Tertiary education | 0.97 (0.81–1.16) | 0.712 |
In contrast, when personality factors (sensation seeking, anxiety/neuroticism, aggression/hostility, and sociability) were included in regression analysis, the odds of SAC declined among those within the PH or HH profile group [OR = 1.63 (1.32–2.02),
In a fourth and final model, the influence of peers with significant drinking problems was added to personality variables. This adjustment exerted a small influence on the association between drinking profiles and SAC. Again, the OR of the NH profile group did not change [OR = 0.83 (0.68–1.02),
OR (95% CI) | ||
---|---|---|
Drinking profiles |
||
LL | 1.00 | |
NH | 0.83 (0.68–1.02) | 0.075 |
PH | 1.57 (1.27–1.96) | <0.001 |
HH | 1.72 (1.23–2.41) | 0.001 |
RSOD |
||
Monthly | 1.00 | |
Weekly | 1.42 (1.18–1.70) | <0.001 |
Daily | 2.36 (1.31–4.26) | 0.004 |
Drinking volume (standard drinks per week) | 1.01 (1.00–1.01) | 0.212 |
Personality traits | ||
Sensation seeking | 1.51 (1.36–1.69) | <0.001 |
Anxiety/neuroticism | 1.08 (1.04–1.13) | <0.001 |
Aggression/hostility | 1.16 (1.12–1.21) | <0.001 |
Sociability | 1.03 (0.99–1.08) | 0.140 |
Peers with significant drinking problems | ||
None | 1.00 | |
One or two | 1.59 (0.91–2.78) | 0.104 |
Some of them | 2.16 (1.52–3.05) | <0.001 |
Most of them | 1.37 (1.12–1.66) | 0.002 |
Principle component analysis on the frequency and amount of drinking at 11 different types of drinking location revealed two dimensions: a
Subsequent regression analysis revealed differential impacts of drinking location profiles on SAC among adolescents and young adults who had engaged in RSO drinking over the previous 12 months. Relative to those classified as low or intermediate for both dimensions, no significant difference in reported SAC was found among those classified as high for the
Differences in the associations between the drinking profiles and alcohol-related problems were attenuated, but persisted after controlling for alcohol consumption (volume and RSOD), personality traits, and peer-influence, indicating independent effects of these variables on SAC. In contrast, including sociodemographic factors did not alter the associations. The self-selection hypothesis of choosing drinking locations based on personality traits (
Location-specific differences related to RSOD and drinking problems have been examined in several previously published studies. In a study by Single and Wortley (
Our results are also partly consistent with those of Stockwell et al. (
Sociodemographic variables could not explain the difference in the association between drinking profile by drinking location and SAC. One reason could be that only young males were included in our study. Furthermore, our results are consistent with previous studies that failed to identify socio-economic status or age as independent predictors of alcohol-related consequences once drinking patterns were controlled for (
Our study has several potential limitations. First, all data were self-reported. Results may therefore be distorted because of assessment bias, recall bias, or social desirability. For example, students willing to experience consequences might have been more likely to report such consequences (
Numerous studies have demonstrated that problematic alcohol consumption, like RSO drinking, is widespread in young people (
On one hand, compliance with alcohol policies and existing legislation at sensible locations with on-premise alcohol purchases – like pubs/bars, discos/nightclubs, and special events – should be monitored more closely. For instance, continuing to serve obviously intoxicated customers has been shown to be the best predictor of negative alcohol-related consequences (
The fact that negative alcohol-related consequences from RSOD may yet appear in the early adulthood emphasizes the need for timely preventive measures. Successful interventions in early life may not only reduce negative alcohol-related consequences during adolescence, but also prevent adult drinking problems (
In summary, it can be stated that a structure of timely preventive measures is needed to reduce alcohol-related harm. However, as the availability of alcoholic beverages is high and there is a long tradition of including alcohol consumption in daily life in Switzerland, the implementation of structural changes poses several challenges (
The present findings suggest that problematic alcohol consumption and negative alcohol-related consequences are strongly associated with a cluster of specific drinking locations. This may, together with further, recently collected data, serve as a basis for specific targeting of public health interventions that aim to lower risky drinking patterns and their negative consequences.
Since RSO drinkers with different cultural backgrounds or different demographics, like higher age or female gender, may select different drinking locations, research also is needed to positively influence regulations drafted by policy makers (
Gerhard Gmel and Meichun Mohler-Kuo are the study’s main investigators. Caroline Bähler and Meichun Mohler-Kuo conceptualized the manuscript. Caroline Bähler analyzed the data and wrote the first draft of this manuscript. Michelle Dey and Simon Foster assisted in data analyses. Gerhard Gmel made major contributions to the content of the manuscript. Petra Dermota was involved in data collection and questionnaire development. All authors contributed to manuscript writing and helped to improve it.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The authors would like to thank Charlotte Eidenbenz and Joseph Studer for their administrative assistance with the project. The project was funded by the Swiss National Science Foundation (33CS30_139467).