RESUMO
OBJECTIVE: Cardiometabolic disorders contribute to morbidity and mortality in people with severe mental illnesses (SMI), yet lifestyle-intervention efficacy in patients with SMI is unclear. Israel's unique mental-health rehabilitation hostels (MHRHs) provide housing to subjects with SMI. We tested how multi-component lifestyle intervention affects cardiometabolic risk-factors in at-risk SMI populations residing in MHRHs. METHODS: In a prospective, cluster-randomized, controlled study, six MHRHs, paired by residents' functioning level, were randomized to lifestyle intervention (nutrition education, physical education), or usual care. Subjects recruited included those with ≥1 of: BMI > 25 kg/m2; plasma triglycerides ≥150 mg/dL; HbA1c ≥ 5.7%; fasting plasma glucose ≥ 100 mg/dL and plasma HDL < 40(men)/ 50(women) mg/dL. Primary outcome was BMI change after 15 months; other outcomes were plasma lipids levels and glycemic control. Low cooperation in one MHRH pair led to their exclusion, the others were assigned to intervention or control. RESULTS: Eighty residents were enrolled to intervention groups and 74 to control. Compared to baseline, intervention-arm participants experienced improvements in BMI (-0.83 kg/m2 [-1.36, -0.29] 95%CI), triglycerides (-30.60 mg/dL [-49.39, -11.82]95%CI) and LDL (-15.51 mg/dL [-24.53, -6.50]95%CI) (all P ≤ 0.003). BMI improvement correlated with number of dietitian consultations (r = -0.30; P = 0.001). No significant differences were found between treatment arms in BMI (-0.46 kg/m2 [-1.11, 0.18]95%CI;P = 0.189), triglycerides (-24.70 mg/dL [-57.66, 8.25]95%CI), LDL (-9.24 mg/dL [-20.50, 2.03]95%CI), HDL and glycemic control. CONCLUSIONS: Lifestyle intervention significantly improved BMI, LDL and triglycerides compared to baseline in at-risk MHRHs residents with SMI, yet compared to usual care the differences did not reach statistical significance. The association between the number of dietitian's consultations and BMI improvement suggests that programs should highlight participants' adherence.