The relationship between obesity and mood disorders (especially depression) has been extensively studied in recent years, and a bidirectional association seems unequivocal . However, most of the studies have only evaluated indirect measurements of adiposity, such as BMI and WC [5–8]. These anthropometric variables, although largely used in clinical practice, may neither indicate the total body fat nor differentiate lean from fat mass. In the present study, we used DXA to determine total lean mass and the percentage of body fat and correlated these parameters with depressive and anxiety symptoms in overweight individuals with MS. Our most relevant results were the following: i – a significant correlation was found between BMI and both lean and fat mass; ii – overall, no correlation was found among anthropometric variables and psychiatric symptoms; iii - there was a direct correlation between the percentage of body fat and the severity of depressive and anxiety symptoms; and iii – there was a strong and inverse correlation between lean mass and specific psychiatric symptoms.
The relationship between total fat and psychiatric symptoms in overweight individuals is of great relevance for clinical practice. Recently, a few studies have demonstrated that depressive symptoms may be an important predictor of abdominal obesity [10, 16]. The most acceptable hypothesis for this relationship involves the hypothalamus-pituitary-adrenal (HPA) axis. As a chronic stressor, mood disorders may change cortisol secretion by stimulation of the HPA axis. This elevation in serum cortisol can increase abdominal fat deposition, promoting glucose intolerance and hypertension [16, 17]. However, the lack of correlation between depression and visceral fat has been demonstrated in the elderly . In contrast, we could demonstrate that the total amount of body fat may be more related to psychiatric symptoms than central fat per se. Because only patients with MS were included in the present study, we may speculate that we have already included patients with increased visceral fat.
One unexpected finding was the inverse relationship found between anxiety symptoms and WC. We could not find a reasonable explanation for this novel finding and we may not exclude that this may be only a false positive result. Further studies are necessary to replicate and clarify this issue.
The existence of a strong and inverse correlation between lean mass and anxiety/depressive symptoms in overweight individuals with MS, even more significant than the correlation between body fat and these same symptoms, is a novel finding with major implications. Body Mass Index is the anthropometric variable frequently used in clinical research to investigate the relationship between weight excess and psychopathology. However, major limitations of BMI include that it cannot differentiate lean from fat mass excess  and also that there may exist important ethnic-specific differences with this parameter . These limitations may be of particular relevance in men, who may present increased BMI related to muscle hypertrophy and not fat. Therefore, the use of BMI may not reflect a simple increase in total fat, and this finding may be an important confounding bias in several studies that fail to find correlations between BMI and psychiatric disorders [21, 22].
Unfortunately, our study was not powered to determine a causal relationship between lean mass and anxiety/depressive symptoms in overweight individuals with MS. We may, however, raise the hypothesis that an increase in lean mass may indicate a healthier individual and therefore be a protective factor for depressive and anxiety symptoms. This hypothesis has already been partially supported by findings of Wagner et al.  and Wallymahmed et al.  in different populations, and recently by Gubata et al.  in a young population. In contrast, one may also speculate that patients with a better psychological profile (i.e., lower rates of psychiatric symptoms) may be more prone to physical activity, which would lead to a significant increase in muscle mass. Lastly, an inverse situation also seems plausible. It has already been demonstrated that older patients with reduced lean mass (sarcopenia) may present an increased prevalence of depressive symptoms . Similar results were obtained by Kress et al. , who have demonstrated that underweight men in the U.S. military active service had increased odds of depressive symptoms. It seems reasonable to speculate that in this population, underweight men would indicate men with diminished lean mass.
Our study has a few limitations. First of all, only a small number of overweight and obese patients with MS were evaluated. This is a very selective population and further studies are necessary to confirm whether our findings would also be applicable for different populations (e.g. lean individuals, obese individuals without MS, morbidly obese patients). Second, as only 10 individuals were male, we could not determine the impact of gender in the relationship between body composition and psychiatric symptoms. Third, Prodigy densitometer does not distinct between subcutaneous and visceral fat, which might have allowed for a more thorough analysis of the role of each type of fat in the present study. Finally, physical activity might be an important confounding variable in the relationship between body composition and psychiatric symptoms. Although we did not include a specific instrument to measure physical activity, this information was obtained in the initial evaluation. None of the participants was practicing regular physical activity in the 6 months prior to study enrollment. Therefore, we may speculate that physical activity may not be an important determinant of our findings.