The present study reports an association of the reduction in depressive manifestations with the decrease in CRP and leptin after a dietary treatment for weight loss in subjects with MetS. Moreover, the reduction in fat mass was also involved in the decrease of depressive symptoms, but it was not implicated in the association of this variable with CRP and leptin. Previous investigations have reported a decrease in depressive manifestations after a weight loss treatment [13, 28], however, this study specifically demonstrates a relationship of the decrease in CRP and leptin with the reduction in self-perceived depression in subjects suffering MetS.
The domains of self-perceived depression were divided in cognitive and somatic questions in order to better interpret the primary cause of patient depression . At baseline, there was a ranked relationship between increasing number of MetS components and a higher rate of somatic questions but not of cognitive components. These findings were similar to those observed in a previous study also based on self-reported BDI questionnaire .
Question number 19 of the BDI test, related to weight loss, was apparently distorting the total BDI score. This observation was confirmed because the weight loss was associated with the decrease in body weight at the end of the dietary treatment, showing that those subjects who lost most weight, had a higher score on the weight loss question of the test. Therefore, the weight loss question was removed from the total BDI score in some analyses. Moreover, the improvement in the fatigability and loss of libido questions (somatic-related questions) might be considered as benefits of the weight loss.
Noteworthy, a positive association between decreases in CRP values and reductions in depressive symptoms was observed as has been reported in previous studies . Inflammation seems to be, in part, responsible for the link between depressive symptoms and MetS. However, the precise nature of this relationship is not clear . Theoretically, the association of leptin with depression may be related both to its metabolic properties and neurobiological activities . Leptin affects cognition and mood in the hippocampus, the cortex and other brain areas associated with cognition [9, 31]. Obesity is related to higher levels of circulating leptin reflecting, in part, an increased formation of adipocytes, and causing leptin resistance in some obese subjects. A positive association between the decline in BDI score and the drop in leptin values was found, in line with results from previous studies [9, 32]. Furthermore, in an unadjusted model, a positive association was noted between insulin and depressive symptoms, being this link well-established [33, 34]. However, this association was no longer statistically significant when change in body fat was introduced in the model.
A previous weight loss intervention study showed that depression scores at baseline predicted adherence to a dietary treatment . In accordance with this outcome, a higher BDI score in drop-outs at baseline was observed, although this value did not reach statistical significance. A positive association between lower BDI score and decreases in body weight was found, which is in agreement with previous studies . In addition, there is some evidence suggesting that adiposity is directly related with depression . In this study, the higher decrease in depressive symptoms was noticeable in those participants with a greater decline in total fat mass. Since adipose tissue is known to secrete inflammatory cytokines and leptin , it might be hypothesized that the decrease in body fat mass may have contributed to reduce CRP and leptin and subsequently decrease depressive symptoms.
Furthermore, the Mediterranean diet has been associated with reduced prevalence and incidence of MetS and depressive symptoms [11, 12]. A healthy dietary pattern has been associated with better mood  and diets rich in ω3 PUFA may decrease serum cytokine production and depressive signs . Also, an inverse association between both vitamin D and vitamin C levels with depressive symptoms has been observed [38, 39]. In this context, we have previously shown that folate intake during this dietary intervention was well correlated with the decrease in depressive symptoms .
Because both MetS and depressive manifestations seem to share several common mechanisms, it has been suggested that the dietary recommendations for MetS might be helpful for depression treatment . The present study strengthens this hypothesis, as the dietary treatment for MetS manifestations also reduced depressive symptoms. Moreover, several unhealthy lifestyle habits have been related with depressive symptoms, such as fast food consumption, high fat intake or low physical activity [5, 40]. In this study, physical activity was not directly associated with depressive symptoms, being in agreement with previous investigations .
The control diet based on the AHA guidelines is a well-design strategy for weight loss , what may explain that no differences between dietary groups were observed in the variables analyzed in this study. As both dietary groups proved to be effective in reducing depressive symptoms and no differences between them were found, both groups were merged and analyzed together as a unique experimental group. A longitudinal observational analysis comparing the BDI scores before and after intervention was conducted, serving the volunteers as their own control. The clear differences between the beginning and the end of the study lend support to the soundness of the analysis. One of the strengths of within-subjects (paired) analyses is the reduction in error variance associated with individual differences, which increases statistical power. Moreover, in order to control the possible confounding role of the intervention group, this variable was included in the multiple-adjusted models investigating the association of anthropometric and biochemical variables with depressive manifestations in the complete sample .
The study has some limitations. Firstly, depressive symptoms were evaluated using a self-report questionnaire, the Beck Depression Inventory, which is not designed as a diagnostic tool but as a screening method . However, this test was chosen as it is widely recognized, it has been shown to be valid for clinical assessment  and it has previously been used to record depressive symptoms in weight loss studies [13, 28]. Secondly, this study only aimed to evaluate the association between the variables included and it cannot be determined any conclusions on causality between changes in body weight, fat mass, leptin and CRP and changes in depressive symptoms. In addition, the prevalence of depressive symptoms in our sample was low, which can be explained with the fact that subjects presenting psychiatric disorders at enrollment were not allowed to participate in the study. Also, the number of participants in this study is not very high, but it may be proposed that type-II errors were overcome since important statistical differences were found.