Supervised vs. self-selected physical activity for individuals with diabetes and obesity: the Lifestyle Gym program.

The effectiveness of different programs of physical activity outside randomized studies is difficult to determine. We carried out an audit in two different units where either a supervised physical activity (PA) program or a self-selected PA program was in use in individuals with type 2 diabetes or obesity. The supervised PA cohort (n = 101) received progressive gym training (120 min, twice a week for 13 weeks) by a dedicated team, with nutritional counseling during resting periods. The self-selected PA cohort (n = 69) was enrolled in a 13-week cognitive-behavioral program (120 min/week, in groups of 12-15 individuals), chaired by an expert team. Body weight and physical fitness (6-min walk test) were measured at baseline and after 6 and 12 months. Outcome measures were attrition, weight loss ≥10 % initial body weight, 10 % increase in 6-min walk test; their association with a PA program was tested by logistic regression analysis. Attrition rate was lower in the supervised PA group (28 vs. 45 % than in the self-selected cohort, P = 0.023). After adjustment for confounders, the supervised PA program was associated with a lower risk of attrition at 1 year (odds ratio 0.45; 95 % confidence interval, 0.21-0.98) at logistic regression analysis. Body weight similarly decreased in both groups (more rapidly in the supervised PA cohort); also physical fitness improved in a similar way, and no differences in achieved targets of body weight (supervised, 31 %; self-selected, 18 %; P = 0.118) or fitness (supervised, 62 %; self-selected, 49 %; P = 0.312) were demonstrated. Different PA programs produce very similar health benefits, but an initially supervised program has lower attrition rates, which might translate into better outcomes in the long term.