Differential associations of somatic and cognitive-affective symptoms of depression with inflammation and insulin resistance: cross-sectional and longitudinal results from the Emotional Distress Sub-Study of the GRADE study.

aims/hypothesisInsulin resistance and inflammation are components of a biological framework that is hypothesised to be shared by type 2 diabetes and depression. However, depressive symptoms include a large heterogeneity of somatic and cognitive-affective symptoms, and this may obscure the associations within this biological framework. Cross-sectional and longitudinal data were used to disentangle the contributions of insulin resistance and inflammation to somatic and cognitive-affective symptoms of depression.

methodsThis secondary analysis used data from the Emotional Distress Sub-Study of the GRADE trial. Insulin resistance and inflammation were assessed using the HOMA-IR estimation and high-sensitivity C-reactive protein (hsCRP) levels, respectively, at baseline and at the study visits at year 1 and year 3 (HOMA-IR) and every 6 months (hsCRP) for up to 3 years of follow-up. Depressive symptoms were assessed at baseline using the Patient Health Questionnaire (PHQ-8), and a total score as well as symptom cluster scores for cognitive-affective and somatic symptoms were calculated. For the cross-sectional analyses, linear regression analyses were performed, with inflammation and insulin resistance at baseline as dependent variables. For the longitudinal analyses, linear mixed-effect regression analyses were performed, with inflammation and insulin resistance at the various time points as dependent variables. In all analyses, depressive symptoms (total score and symptom cluster scores) were the independent variables, controlled for important demographic, anthropometric and metabolic confounders. For the analysis of insulin resistance (HOMA-IR), data from 1321 participants were analysed. For the analysis of inflammation (hsCRP), data from 1739 participants were analysed.

resultsIn cross-sectional analysis and after adjustment for potential confounders, a one-unit increase in PHQ-8 total score was significantly associated with a 0.8% increase in HOMA-IR (p=0.007), but not with hsCRP (0.6% increase, p=0.283). The somatic symptom score was associated with a 5.8% increase in HOMA-IR (p=0.004). Single-item analyses of depressive symptoms showed that fatigue (3.6% increase, p=0.002) and increased/decreased appetite (3.5% increase, p=0.009) were significantly associated with HOMA-IR cross-sectionally. The cognitive-affective symptom score was not significantly associated with HOMA-IR at baseline. In longitudinal analyses, a one-unit increase in PHQ-8 total score was significantly associated with a 0.8% increase in hsCRP over time (p=0.014), but not with HOMA-IR over time (0.1% decrease, p=0.564). Again, only the somatic symptom cluster was significantly associated with hsCRP over time (5.2% increase, p=0.017), while the cognitive-affective symptom score was not. CONCLUSION/

interpretationThe results highlight the associations of depressive symptoms with markers of inflammation and insulin resistance, both cross-sectionally and longitudinally, in individuals with type 2 diabetes. In particular, somatic symptoms of depression appear to be the driver of these associations, even after controlling for concomitant conditions, with a potential role for fatigue and issues with appetite.

trial registrationClinicalTrials.gov NCT01794143.