However, the rates of relapse and physical and mental comorbidities are rather high, pointing out the need for adjuvant therapies and long-term life-style modifications.Compared small molecule to smoking, evidence is more limited in alcohol abuse/dependence, and RCTs are extremely rare. Nine studies were identified that investigated the effects of EX programs on abstinence, relapse rates, and/or different associated somatic, emotional, and psychological outcomes (see Table 2).Table 2Studies investigating EX in the therapy of alcohol abuse/dependence.In these studies, treatment duration ranged from four weeks [28, 33] to four months [31], with training frequencies ranging from three to five times a week. EX interventions were mostly aerobic [28�C34, 36], but one study [35] used a holistic ��Body-Mind�� intervention which prohibits conclusions about EX alone.

Six studies reported drinking episodes, craving, or days of abstinence as substance-related outcomes [28, 30, 32, 34�C36], and four of these studies found significantly stronger improvements in the EX group [30, 32, 35, 36], whereas two studies did not find any group differences [28, 34]. Secondary psychological outcomes like depression, anxiety, stress, self-concept, locus of control, and sleep quality, which increased at least in one of the EX conditions, were reported in four studies [28, 29, 31, 33]. In contrast, two studies did not find group differences concerning the reduction of depression [34, 35] and anxiety [34].Significant increases in fitness were reported in eight studies [28�C32, 34�C36], which were preserved at 5-month followup in one study [34], whereas one other study [33] did not find significant changes in fitness.

Only one study fulfilled criteria for a RCT [34], whereas the other studies had several methodological limitations. Seven studies included control groups [28, 30�C35], whereas one study was a one-group pre-post comparison [29], and one study [36] did not employ control conditions at all. Random assignment of study participants to a treatment condition was performed in four studies [28, 31, 32, 34], whereas one study used a time-staggered Entinostat control group [33], another study compared samples from different centers [30], and one study did not state their assignment strategy [35].Sample sizes were small in five studies [28, 31�C33, 36], and none of the studies performed intention-to-treat analyses to correct for the high number of dropouts. Four studies did not specify the patients’ diagnoses [30, 33, 34] or included subjects without a clinical diagnosis of alcohol abuse or dependence [32].