Hypoglycemia is reported to be strongly associated with depression in diabetic patients [4, 5]; however, the contrary observation in the present study can be because the present study deals with low blood glucose levels in people without diabetes. This calls for the need to take up population-based studies to understand the association of depression with low blood glucose in the non-diabetics/general populations.
Overall high blood glucose level was also found to be not associated with depression in the present study. This observation is contrary to reports suggesting a higher prevalence of depression among individuals with diabetes [2, 6]. Possible explanations of this contrary observation are the low prevalence of high blood glucose levels and diabetes in the studied population (Table 2), and unawareness of participants about their high blood glucose level (being a population-based study, most of the participants were naive about their blood glucose level). Most of the studies reporting a strong association of depression with hyperglycaemia were conducted among individuals who were aware of their diabetic condition which is likely to influence their depression status . According to Fisher et al., depression in diabetic individuals can be due to the worries, concerns and fears collectively called “diabetes-related emotional distress” among individuals struggling with progressive chronic disease . In fact, a meta-analysis revealed that the risk for depression in individuals with undiagnosed diabetes was similar to those with normal glucose metabolism and higher for those who were aware of their diabetic status . Moreover, in a recent population-based study conducted among Venezuelan adults, depression was not found to be associated with diabetes .
Interestingly, high blood glucose level was observed to pose a 3.618-folds increased risk for depression in females (p = 0.058) in the present study, however, no such relationship was found among males. Several studies have highlighted that diabetic females are at a greater risk of depression than their male counterparts . However, this apparent association between high blood sugar levels and depression in those females who were unaware of their impaired glucose metabolism/ diabetes needs more investigation.
Further, in the present study, TC and LDL were found to be inversely associated with depression. These observations are in concordance with earlier reports [8, 9, 13]. However, several studies have reported contrary findings as well [10, 21]. Other lipid variables did not show any significant association with depression. Of all the lipid variables, the inverse correlation between high TC and depression has been the most consistent and strongest .
Various explanations are put forward to explain the association of low serum cholesterol with depression. Lipid plays an important role in neuronal function . Cholesterol is an essential component in the functioning of neurotransmitters, particularly serotonin, disruption in which may lead to depression [7, 22]. According to a hypothesis, cholesterol is loosely bound with the phospholipid layers of biological membranes and is freely exchanged with serum cholesterol . Any reduction in serum cholesterol can decrease the membrane-bound cholesterol of neurons leading to a lower lipid microviscosity and altered functioning of serotonin transporters and receptors . These processes can effectively reduce serotonin in the brain and cause depression. However, the relationship between cholesterol and depression appears to be bidirectional. Poor appetite, a very common symptom of depression, may lead to lower fat and cholesterol intake . Cholesterol synthesis can also be impaired due to depression associated cytokines activation .
Further, after gender stratification, the inverse relationship of high TC and high LDL with depression remained significant in males but not in females. Previous studies have also highlighted the gender difference in the association of depression with lipid variables [9, 24]. In a study conducted on suicide attempters, a significantly lower value of serum cholesterol levels was found in males but not in females . Another study reported an association of depression with lower LDL levels in men but not in women (this report is partially in concordance with the observation of the present study) . In the present study, high TG and high VLDL appeared to increase the risk for depression in females but not in males. This observation has not been widely reported and requires further investigation.
Gender plays an important role in lipid and lipoprotein metabolism . Gender-specific association of lipid variables with depression can be due to gender-specific genetic architecture. Studies have indicated that both depression  and dyslipidaemia  may have sex-specific genetic architecture. Other quantitative traits like serum cortisol and whole blood serotonin as well have been reported to show sex differences . In fact, sex can influence both penetrance and expressivity of a variety of traits including lipid traits . Steroid-related genes and hormones have also been implicated in gender-specific effects on lipid metabolism and depression [26, 29, 30]. Hence, the differential association of lipid variables with depression in men and women can be due to complex genetic, hormonal, and environmental interactions. Nevertheless, since the present study is cross-sectional, the association of lipid variables with depression should not be inferred as causality.
The present study has several limitations that should be mentioned. Firstly, the depression status of the participants was ascertained using BDI-II. Although BDI-II is a widely used, cross-culturally validated and cost-effective tool, it is not the gold standard for determining depression status. Clinical interview, which is considered the gold standard method, should ideally be employed for confirmation of depression. However, time and cost considerations did not permit us to conduct clinical interviews in the present study. Also, since clinical interviews were not conducted, participants could not be given diagnosis or therapeutic advice. Further, the cross-sectional nature of the present study can, at best, examine the association of independent variables with depression, which should not be inferred as causality. Longitudinal studies should be conducted to explicate cause-effect relationships between the studied biochemical variables and depression. Lastly, the number of female participants was (nearly twice) more than the number of male participants. As the data collection was done using the household survey method (in the north Indian rural setting, where male members are more likely to be not present at their homes during the day hours), field investigators encountered (and hence recruited) eligible female individuals at a much higher frequency than male individuals (which is also reflected in the sample composition). Since the sample was not gender-balanced, (apart from the overall analysis) all the relevant analyses have been performed separately for both males and females.