Le Xiao1, Jia Zhou1, Britta Galling2, Run-Sen Chen1, Gang Wang3. 1. The National Clinical Research Center for Mental Disorders & Beijing Key Laboratory of Mental Disorders, Beijing Anding Hospital & the Advanced Innovation Center for Human Brain Protection, Capital Medical University, Beijing, China. 2. Department of Child and Adolescent Psychiatry and Psychotherapy, Centre for Integrative Psychiatry, School of Medicine, Kiel, Germany; Department of Child and Adolescent Psychiatry, Psychosomatic Medicine and Psychotherapy, Charité-Universitätsmedizin Berlin, Berlin, Germany; Department of Child and Adolescent Psychosomatic Medicine and Psychotherapy, Altona Children's Hospital, Hamburg, Germany. 3. The National Clinical Research Center for Mental Disorders & Beijing Key Laboratory of Mental Disorders, Beijing Anding Hospital & the Advanced Innovation Center for Human Brain Protection, Capital Medical University, Beijing, China. Electronic address: gangwangdoc@ccmu.edu.cn.
Abstract
BACKGROUND: In patients with major depressive disorder (MDD), poor antidepressant treatment response might be associated with an excessive body mass index (BMI). However, the impact of underweight on treatment response is unclear. Moreover, it has not been studied whether a continuous or categorical BMI measure should be used to predict of treatment response. METHODS: Post-hoc analysis of data collected in a clinical trial including adults with MDD (n=202) reporting outcomes of antidepressant medication, i.e. paroxetine, mirtazapine or paroxetine+mirtazapine. Measures included baseline BMI (underweight=BMI <18.5, normal weight:=BMI:18.5-23.9, overweight=BMI≥24) and symptom severity (17-item-Hamilton Depression scale; HAMD-17) assessed at weeks 0, 2, 3, 4, 6 and 8. Univariate analyses were used to explore the effect of baseline BMI on HAMD-17 reduction, response (defined as ≥50% HAMD-17 reduction) and remission (defined as HAMD-17 ≤7) at endpoint. Pearson correlation were used to explore the relationship between body weight, BMI as continuous measure and HAMD-17 reduction. Logistic regression was used to determine the predictors for remission. Multiple linear regression was used to establish the correlation of BMI with change of HAMD-17. RESULTS: 111 (55.0%) patients were normal weight, 20 (9.9%) were underweight, 71 (35.1%) were overweight. Underweight patients showed the best improvement to antidepressant treatment. Non-remitters had greater body weight and BMI than remitters (P<0.05). The reduction of HAMD-17 was correlated with baseline body weight (r=-0.16, P=0.032) and BMI (r=-0.19, P=0.012). Logistic regression found patients with BMI<24 to be 2 times (OR=1.958, 95%CI: 1.015, 3.774) remitters (P=0.045) than overweight patients. The multiple linear regression showed that the change of HAMD-17 total score decreased with increasing BMI (β=-0.32, P = 0.016). CONCLUSION: We confirmed that BMI can predict treatment outcomes in MDD. For the first time we found that underweight patients benefit most from antidepressant treatment. The findings may be useful to physicians in their decision regarding the choice of antidepressants according to BMI.
BACKGROUND: In patients with major depressive disorder (MDD), poor antidepressant treatment response might be associated with an excessive body mass index (BMI). However, the impact of underweight on treatment response is unclear. Moreover, it has not been studied whether a continuous or categorical BMI measure should be used to predict of treatment response. METHODS: Post-hoc analysis of data collected in a clinical trial including adults with MDD (n=202) reporting outcomes of antidepressant medication, i.e. paroxetine, mirtazapine or paroxetine+mirtazapine. Measures included baseline BMI (underweight=BMI <18.5, normal weight:=BMI:18.5-23.9, overweight=BMI≥24) and symptom severity (17-item-Hamilton Depression scale; HAMD-17) assessed at weeks 0, 2, 3, 4, 6 and 8. Univariate analyses were used to explore the effect of baseline BMI on HAMD-17 reduction, response (defined as ≥50% HAMD-17 reduction) and remission (defined as HAMD-17 ≤7) at endpoint. Pearson correlation were used to explore the relationship between body weight, BMI as continuous measure and HAMD-17 reduction. Logistic regression was used to determine the predictors for remission. Multiple linear regression was used to establish the correlation of BMI with change of HAMD-17. RESULTS: 111 (55.0%) patients were normal weight, 20 (9.9%) were underweight, 71 (35.1%) were overweight. Underweight patients showed the best improvement to antidepressant treatment. Non-remitters had greater body weight and BMI than remitters (P<0.05). The reduction of HAMD-17 was correlated with baseline body weight (r=-0.16, P=0.032) and BMI (r=-0.19, P=0.012). Logistic regression found patients with BMI<24 to be 2 times (OR=1.958, 95%CI: 1.015, 3.774) remitters (P=0.045) than overweight patients. The multiple linear regression showed that the change of HAMD-17 total score decreased with increasing BMI (β=-0.32, P = 0.016). CONCLUSION: We confirmed that BMI can predict treatment outcomes in MDD. For the first time we found that underweight patients benefit most from antidepressant treatment. The findings may be useful to physicians in their decision regarding the choice of antidepressants according to BMI.