Table 5.

Details of reduction studies.

Study	Design	Focus	Population	Physical Activity Intervention	Comparison(s)/Control(s)	Outcomes and follow ups	Results	PA Data (including adherence)	Additional findings
Murphy et al. (1986)	RCT (3 arm)	Alcohol	Social drinkers recruited from undergraduate classes (N = 60).	Running group. Met as a group 3/week for 70 min and asked to run ‘some other time per week’ on their own for 8 weeks.	Meditation group. Instructed to meditate for 20 min twice a day. Met as a group 3/week for a group meditation session.	Daily journal recording alcohol intake.	Running condition reduced consumption by 60% at longest follow up from baseline, significantly lower than control, and a non-significant reduction compared to meditation.	Those in the running group showed significantly increased VO2 max levels post intervention compared to baseline.	60 recruited, only 31 analysed
USA			Age range: 21–30 years		No treatment control. Asked to keep journals of their behaviour.	Weeks 1–2 (pre-intervention) mean, weeks 3–6 mean, and weeks 7–10 mean
			Male (100%)					All subjects in the running group completed the 8 week intervention, running on average 3.4 times per week.	Those who ran less than 3.5 times per week reduced alcohol intake and increased VO2 max measure by the same amount as those who ran more than 3.5 times per week.
Correia et al. (2005)	RCT (3arm)	Alcohol and other drugs	Undergraduates at a large private university (N = 133, n = 105 randomised).	Activity Increase (AI): Oral and written instructions to increase days engaged in physical activity and creative/artistic activity by 50% in next 28 days, including signing a contract with behavioural targets and asked to self-monitor behaviour	Substance use Reduction (SR): Oral and written instructions to reduce frequency of consumption during next 28 days including signing a contract with behavioural targets and asked to self-monitor behaviour.	Daily Drinking Questionnaire indicating frequency and quantity of alcohol in previous 28 days.	Both SR and AI significantly reduced number of substance use days (includes alcohol); and number of total standard drinks from baseline to follow up. AI did not significantly differ from control.	AI group reported significantly more exercise days at follow up compared to SR and control group. SR reported significant decrease in PA days, AI reported significant increase in PA days
USA			Age, mean (SD) 19.76 (3.76)		No change control.	28 days
			Male (31%)
Georgakouli et al. (2017))	Pre-Post	Alcohol	Heavy drinking male volunteers (N = 11).	8-week supervised exercise intervention of moderate intensity (50–60% of Heart Rate Reserve). No more details.		Alcohol use questionnaire (not specified). How many alcohol units do you drink per day? How many AU did you drink last night? How many times did you consume alcohol over the last month? How many AU did you drink per occasion over the last month? How many days do you usually drink alcohol? How many AU do you usually drink per week? Time until first drink after an exercise session.	Significant decreases in alcohol use across all questions other than desire to stop or reduce drinking and amount drank last night.	Significant decreases found for weight and waist circumference and increases in flexibility and situps. Self-reported met minutes per week (IPAQ) significantly increased form baseline to follow up.	No change in desire to drink, but changes in amount drunk.
Greece			Age, mean(SD): 30.3 (3.5)			Mid and post intervention
			Male (100%)
Oaten and Cheng (2006)	Pre-Post (within and between)	Alcohol	Sedentary undergraduates (N = 24); cohort 1 (n = 9), cohort 2 (n = 6), cohort 3 (n = 9).	Exercise programme: 3–4 times per week over 8 weeks of cardio vascular exercise, tailored to suit the individual by gym staff, including aerobic classes, free-weights, and resistance training. **All participants were subjected to a thought suppression task not reported as part of the intervention	Waiting list control phase (no exercise)	7-day recall procedure with quantity the measure of interest (standard units of alcohol).	Average decrease of 5 drinks per week during exercise phase compared to waiting list control phase	No data in relation to adherence rates, other than 100% retention, and that participants attended the gymnasium, and attendance increased over time.	Improvements in a wide range of regulatory behaviours observed.
Australia			Age, mean (SD); 24 (6)			Baseline, 1 month, and 2 months
			Male (25%)
Reddy et al. (2014)	Pilot RCT	Alcohol and other drugs	Females aged 18–65 with at least sub-threshold PTSD (N = 38).	12 Kripalu-based Hatha yoga session of 75 min incorporating trauma-sensitive yoga. The intervention also incorporated elements of “mindfulness ad dialectical behavioural therapy, a specialized form of cognitive behavioural therapy”. (n = 20)	An assessment control (n = 18).	AUDIT and DUDIT.	AUDIT and DUDIT scores improved in the yoga group and worsened in the control group over time. Changes were not significantly different between groups over time.		69% of yoga group reported noticing PTSD symptoms less at follow up, compared to 80% of the control group reporting no difference or an increase in symptom perception.
USA			Age, mean (SD); 44.4 (12.4)			Baseline, end of intervention, and 1 month after intervention
			Female (100%)
Scott and Myers (1988)	Non-RCT	Alcohol and other drugs	Native adolescents belonging to the River Desert Community of the Algonquin nation attending regular school gym classes (N = 74).	Treatment was administered during first 30 min of normally scheduled gym classes (every other day on a 4-day cycle for 70 min). Physical fitness training designed to enhance aerobic capacity, flexibility, and strength was delivered for 24 weeks, with students encouraged to set personal fitness goals. (n = 30)	Regular physical education classes emphasising sports-specific skills (n = 44).	The Native American Drug Use Survey.	Within and between subject tests showed no difference in experience with alcohol and drugs, recent (past 2 months) or long term (past 12 months) substance use. Alcohol and drug use in past 12 months showed greater increase in comparison group but was not significantly different.	Treatment groups showed significant greater improvement in cardiovascular fitness than the comparison group	Consistent pattern of increased use across grades, and from pre-test to post-test suggesting age and aging may influence use.
Canada			Age, mean (range); 14.6 (12–18)			Baseline, end of intervention
			Male (47.2%)
Weinstock et al. (2014))	Pilot randomised trial	Alcohol	Sedentary hazardous drinking college students (N = 31).	MET: 50-min one off session framed as a “wellness intervention” for increasing exercise, providing personalised feedback on exercise habits in comparison to population norms and exercise guidelines. Along with developing a change plan to start exercising. Alcohol was not discussed.	MET + CM: MET intervention plus 8 weeks of contingency management. Weekly meeting with interventionist provided reinforcement for verified completion of activities in previous week and develop a new contract for the upcoming week. Verified completion of activities was rewarded with randomly drawn prizes averaging $230 if they completed all activities. Activities included walking with a pedometer, jogging on a treadmill for 15 min, or attending an exercise class at the gym; all activities were explicitly defined in terms of duration and length.	TLFB for previous 60 days to assess alcohol use.	No significant effects were found on alcohol outcomes relating to time or condition.	Both groups increased PA, MET + CM exercised significantly more often.	MET + CM drinking significantly more days than MET at baseline.
USA			Age, mean years (SD); MET: 20.1 (1.2); MET + CM: 21.0(2.3) Male; MET (26.7%); MET + CM (43.8%)			Baseline, end of intervention		MET + CM attended an average of 6.94 (SD = 2.24) sessions out of a possible 8, completing 17.9 (SD = 8.8) verified completed activities.
								Both groups reported significant increases in exercise frequency, estimated weekly calorie expenditure, and VO2 max with no differences between groups.
Weinstock et al. (2016)	Randomised trial	Alcohol	Sedentary heavy drinking college students (N = 70).	MI + EC (Motivational Interviewing + Exercise contracting): Two 50-min motivational interviewing sessions (one following baseline and 4 weeks later), plus eight weekly individually delivered exercise contracting sessions. Alcohol was not discussed unless raised by participant. Participants were reinforced ($5 gift certificate) for each EC session they attended.	MI + CM (Motivational Interviewing + Contingency management): Two 50-min motivational interviewing sessions (one following baseline and 4 weeks later), plus eight weekly individually delivered exercise contracting sessions. Alcohol was not discussed unless raised by participant. Participants were reinforced (draw from a prize bowl) for completing verified exercise activities as agreed in the EC sessions.	TLFB previous 60 days.	Significant reduction in total weekly standard drinks and weekly binge drinking episodes at 2 and 6 months compared to baseline with no differences between arms.	89% of participants attended all eight exercise contracting sessions. An average of 30.4 activities were selected, of which 17.9 activities were verified completed. No change in cardiorespiratory fitness was observed.	Changes in exercise were not predictive of changes in drinking.
USA			Age, mean years (SD); MI + EC: 19.9 (1.3); MI + CM 20.1(1.6)Male; MI + EC (37.1); MI + CM (51.4%)			Baseline, 2 months (end of intervention), and 6 months post baseline	Both arms significantly increased exercise frequency during intervention period, which decreased at 6 months but still higher than baseline. MI + CM showed greater increases in exercise frequency than MI + EC.