Jury duty auto enrollment. The trial was conducted in accordance with the protocol which was approved by the Institutional Review Board for Human Subjects Research (IRB) at the University of Arkansas for Medical Sciences (UAMS). Informed consent was obtained from all study participants. Subject recruitment and data collection {#s2c} --------------------------------------- From March 2010 to November 2013, subjects were recruited from the greater Little Rock metropolitan area and surrounding areas with assistance from the Pulaski County Circuit Clerk's Office. Eligible subjects were 18 years of age or older, registered to vote in Pulaski County (the only county in Arkansas where voters are required to participate in the jury service program), and were able to speak, read, and understand English. Potential jurors were randomly selected from the voter registration database. After a consenting interview, respondents completed a baseline paper-and-pencil survey consisting of questions related to demographics, income and education, alcohol consumption, and other behavioral data. Respondents were also asked to rate their current health using a 5-point scale (1 = excellent; 2 = very good; 3 = good; 4 = fair; 5 = poor). Covariates {#s2d} ---------- Baseline demographic variables collected included age, sex, race, Hispanic ethnicity, education, annual income, employment status, household size, smoking status, alcohol consumption, and self-reported health status. Alcohol consumption was measured using the Alcohol Use Disorders Identification Test -- Consumption (AUDIT-C) scale (a score of 2 or higher was used as the cut-off for hazardous alcohol consumption) [@pone.0113641-Bohn1]. Income was assessed based on the question: "What is your total combined household income from all sources before taxes, in dollars?" Income data were collected in six income ranges (\<\$19,999; \$20,000-\$39,999; \$40,000-\$59,999; \$60,000-\$79,999; \$80,000-\$99,999; \$100,000 or more; prefer not to answer) and in deciles (1 = lowest; 10 = highest). Health status was measured using the question "In general, would you say your health is?" with a 5-point scale ranging from 1 = excellent to 5 = poor. Self-reported health measures have been shown to be valid and reliable proxy measures of overall health status and functional status [@pone.0113641-Kosinski1]--[@pone.0113641-Bekkering1]. Exposure: Alcohol-related social environments {#s2e} --------------------------------------------- Subjects were asked to indicate whether they had friends who consumed alcohol using the question "Do you have any friends who drink beer or wine or liquor" and the question "In the past 12 months, how often have you been to a party where there was alcohol being consumed, and you did not drink?" If they answered affirmatively to either question, they were also asked if they had been intoxicated in the past 12 months. If they answered affirmatively to both of these questions, they were considered to have experienced an alcohol-related social environment. Outcomes {#s2f} -------- Respondents completed follow-up surveys 3 and 12 months after enrollment in the trial. Response rates were 80% and 69% at 3 and 12 months, respectively, and follow-up data were used in this analysis. The items from the 3-month and 12-month questionnaires were identical and included: "Think about the last 3 months -- Have you had a drink of any alcoholic beverage?" and "Have you consumed 5 or more drinks in a day within the last 3 months?" Respondents were classified as drinkers if they indicated that they were current drinkers. To be consistent with previous research [@pone.0113641-McGue1]--[@pone.0113641-Heath1], respondents who indicated in the 3-month survey that they had consumed alcohol within the last 30 days were excluded from the analysis. As with most previous alcohol studies, we did not ask about alcohol consumption prior to enrollment; the only other published study using similar data also defined a "drinker" as an individual who consumed 5 or more drinks in a day within the last 30 days (12 months) [@pone.0113641-McGue1]. Statistical analyses {#s2g} -------------------- Characteristics of respondents at the 3-month and 12-month follow-up visits were calculated and compared using the chi-square test for categorical variables and the Wilcoxon Rank Sum test for continuous variables. The primary outcomes in this study were: drinking frequency in the previous 3 months; reporting any 30-day abstinence in the previous 12 months; and AUDIT-C score at 12 months. To assess the effect of drinking patterns on drinking frequency and abstinence, data on respondent characteristics were pooled across the 3 and 12-month follow-up visits. Characteristics that were used for the pooled analysis included: respondent age, respondent sex, smoking status (current, former, never), and self-rated health status (good, fair, poor). The associations between alcohol-related social environment at baseline and the outcomes at the follow-up visits were assessed using logistic regression models. Odds ratios (ORs) were reported to assess the association between alcohol-related social environment at baseline and drinking frequency in the previous 3 months, any 30-day abstinence in the previous 12 months, and AUDIT-C score at 12 months. All models included controls for covariates described above, and respondents with missing data were included in the analytical dataset using the method of listwise deletion. Based on a post-hoc power analysis, 180 respondents (45 per group) were needed to detect a statistically significant difference between groups with an alpha of 0.05 and 80% power. All analyses were performed using SAS version 9.3 (SAS Institute Inc., Cary, NC). Results {#s3} ======= Baseline characteristics of respondents {#s3a} --------------------------------------- At baseline, there were 178 respondents enrolled from the jury service database. Four individuals were excluded because they did not complete the baseline questionnaire and the first follow-up questionnaire; 170 respondents were included in the analysis. Of the respondents, 76 (44%) were male and 100 (60%) were married, divorced, or widowed; mean age was 54 years (standard deviation \[SD\] = 14). The vast majority of respondents self-identified as white (91%); 8% identified as Hispanic and 1% as black. In terms of education and income, 56% had less than a high school degree and 32% earned less than \$35,000 annually; 12% were current smokers. Most respondents had less than a good perceived health status (60%), but approximately 80% were either employed (80%) or were homemakers or retired (12%). All characteristics were comparable between the 3-month and 12-month follow-up samples. The associations between alcohol-related social environments at baseline and the outcomes at the 3-month and 12-month follow-ups {#s3b} --------------------------------------------------------------------------------------------------------------------------------- [Table 1](#pone-0113641-t001){ref-type="table"} summarizes the baseline characteristics of the study population. There were approximately equal proportions of respondents who reported drinking a beer, a mixed drink, a glass of wine, or more than one alcoholic beverage in the previous 3 months. A total of 61% of respondents were male; most were married or living with a partner (59%) and were not employed (70%). Almost half of the respondents rated their health as good, and half as fair or poor. Among drinkers, the mean AUDIT-C score was 4.7±2.6, which corresponds to hazardous alcohol consumption in the general population [@pone.0113641-Bohn1]. Almost half of the respondents reported consuming 5 or more drinks in a day (47%), and almost 40% reported experiencing an alcohol-related social environment during the previous 12 months. 10.1371/journal.pone.0113641.t001 ###### Baseline characteristics of respondents who responded to both the baseline and 3-month follow-up surveys (*n* = 170). {#pone-0113641-t001-1} Characteristic Number (%) or mean (SD)[a](#nt102){ref-type="table-fn"} ------------------------------------------------------- --------------------------------------------------------- Sex  Male 82 (48.2%)  Female 88 (51.8%) Age  18--29 33 (19.4%)  30--39 24 (14.1%)  40--49 32 (18.8%)  50--59 46 (27.1%)  60+ 33 (19.4%) Ethnicity  White 154 (91.1%)  Black 8 (4.7%)  Hispanic 6 (3.5%) Marital status  Married, living with a partner, or widowed 100 (59.2%)  Single, separated, divorced or never married 70 (41