A PPG Sensor-Based Feedback Intervention for Heavy Drinking Young Adults

Total drinks consumed will be measured using the Time Line Followback Interview at baseline, Week 6, and Week 10. This standardized interview asks subjects to self-report how many drinks they consume each day over the past 4 weeks at baseline and then since the last assessment point at Weeks 6 and 10. Higher scores indicate a greater number of drinks consumed. Total drinks will be summed over the past 4 weeks at intake, Week 6, and Week 10. Totals were transformed using a square root transformation since these values were not normally distributed. Mixed effects models were then conducted to evaluate the effect of condition on total drinks over time with condition, time, and their interaction in the model and sex and baseline total drinks as covariates.

Total drinks per drinking day will be measured using the Time Line Followback Interview at baseline, Week 6, and Week 10. Total drinks/drinking day will be summed over the past 4 weeks at intake, Week 6, and Week 10. Totals were transformed using a square root transformation since these values were not normally distributed. Mixed effects models were then conducted to evaluate the effect of condition on total drinks/drinking day over time with condition, time, and their interaction in the model and sex and baseline drinks per drinking day as covariates.

This tools asks subjects to self-report how many drinks they consume during a one month period. The score of this measure will be determined by the amount of self-reported alcohol consumption that occurred each day. A heavy drinking day for a man would be ≥5 drinks per sitting and for a women it would be ≥4 drinks per sitting.

Self-reported percent heavy drinking days will be measured using the Time Line Followback Interview at baseline, Week 6, and Week 10. This standardized interview asks subjects to self-report heavy drinking occasions over the past 4 weeks at baseline and then since the last assessment point at Weeks 6 and 10, defined as ≥5 drinks per sitting and for a women it would be ≥4 drinks per sitting. Higher scores indicate a greater percentage of heavy drinking days. The percentage of heavy drinking days will be summed over the past 4 weeks at intake, Week 6, and Week 10. Mixed effects models were then conducted to evaluate the effect of condition on percent heavy drinking days over time with condition, time, and their interaction in the model and sex and baseline percent heavy drinking days as covariates.

Self-reported percent abstinent days will be measured using the Time Line Followback Interview at baseline, Week 6, and Week 10. This standardized interview asks subjects to self-report how many days they did not consume any alcohol each day over the past 4 weeks at baseline and then since the last assessment point at Weeks 6 and 10. Higher scores indicate a greater percentage of abstinent days. The percentage of abstinent days will be summed over the past 4 weeks at intake, Week 6, and Week 10. Mixed effects models were then conducted to evaluate the effect of condition on percent abstinent days over time with condition, time, and their interaction in the model and sex and baseline percent abstinent days as covariates.

Mean alcohol related consequences were measured using the Brief Young Adult Alcohol Consequences Questionnaire at baseline, Week 6, and Week 10. Each consequence is scored 1 point and a total score reflects the total number of consequences. Higher scores indicated more consequences.Total score range 0-24. The three timepoints are summed then averaged. Mixed effects models were then conducted to evaluate the effect of condition on consequences over time with condition, time, and their interaction in the model and sex and baseline consequences as covariates.

Mean sleep quality will be measured using the PROMIS - Sleep Disturbance Form 8 assessment. The sleep disturbance assessment has 8 questions that yield a total score (summed scores). This raw score is then converted to a standardized T score from 0-100 with a mean score of 50. A score above the mean would indicate that the subject experiences worse sleep quality. Mixed effects models were then conducted to evaluate the effect of condition on sleep quality over time with condition, time, and their interaction in the model and sex and baseline sleep quality as covariates.

Mean sleep quality will be measured using the PROMIS - Sleep-Related Impairment Form 8 assessment. The sleep impairment assessment has 8 questions that yield a total score (summed score). This raw score is then converted to a standardized T score from 0-100 with a mean score of 50. A score above the mean would indicate that the subject experiences more sleep-related impairment. Mixed effects models were then conducted to evaluate the effect of condition on sleep-related impairment over time with condition, time, and their interaction in the model and sex and baseline sleep-related impairment as covariates.

Mean sleep duration will be measured daily for 6 weeks by the PPG device. Sleep duration will then be averaged in 2-week intervals at Weeks 2, 4, and 6 and evaluated over time using mixed effects models with condition, time, and their interaction in the model and sex and an indicator variable of weekday vs. weekend as covariates. Sleep duration was transformed using a square-root transformation.

Heart rate variability (HRV) will be measured daily for 6 weeks by the PPG device. HRV will then be averaged in 2-week intervals at Weeks 2, 4, and 6 and evaluated over time using mixed effects models with condition, time, and their interaction in the model and sex and an indicator variable of weekday vs. weekend as covariates. Sleep duration was transformed using a log transformation.

Lowest Resting Heart Rate (RHR) will be measured daily for 6 weeks by the PPG device. The lowest value will then be averaged in 2-week intervals at Weeks 2, 4, and 6 and evaluated over time using mixed effects models with condition, time, and their interaction in the model and sex and an indicator variable of weekday vs. weekend as covariates. Sleep duration was transformed using a log transformation. RHR can vary anywhere between 40-100 beats per minute. Lower RHR would indicate better cardiovascular health.