Testing inflammation as a mechanism for depressive symptoms in type 2 diabetes

Student thesis: Doctoral ThesisDoctor of Philosophy

Abstract

Background
Depressive symptoms are reported by 20% of people with type 2 diabetes and are twice as prevalent in females than males with type 2 diabetes. However, the reasons for the link are incompletely explained by psychological and behavioural factors alone, and there is crosssectional evidence that activation of the innate inflammatory response (elevated inflammation) is associated with depressive symptoms in type 2 diabetes. This thesis tests i) whether depressive symptoms are associated with worsening glycaemic control and incident diabetes complications over the first 2 years of diabetes; ii) whether elevated inflammation is associated with worsening depressive symptoms in type 2 diabetes and whether female sex moderates this association; iii) whether specific classes of pharmacological diabetes treatments – many of which reduce inflammation – lead to reduction in depressive symptoms; and iv) whether repositioning of sitagliptin – a diabetes treatments with anti-inflammatory effects – is a feasible strategy for improving depressive symptoms in people with type 2 diabetes.
Methods
Study 1 is set in a multi-ethnic, primary care cohort (n=1735) of adults with newly diagnosed type 2 diabetes and uses adjusted regression models to test whether depressive symptoms are associated with adverse diabetes outcomes over 2 years. In the same cohort, Study 2 uses adjusted general linear models to test whether elevated inflammation is associated with worsening depressive symptoms over 2 years and whether female sex moderates this association. Study 3 is a systematic review and random-effects meta-analysis of randomised controlled trials (RCTs), which tests whether licensed pharmacological diabetes treatments lead to improvement in depressive symptoms over time, including potential biological moderators of such effects. Finally, Study 4 tests the feasibility of conducting a placebo-controlled RCT of sitagliptin – a diabetes treatment with anti-inflammatory effects – for the treatment for depressive symptoms in 44 people with type 2 diabetes recruited from primary care.
Results
In Study 1, depressive symptoms were not associated with worsening glycaemic control (β = 0.02 [95% CI −0.05, 0.09], p = 0.63) over 2 years. By contrast, depressive symptoms were associated with increased risk of macrovascular complications (OR 2.78 [95% CI 1.19, 6.49], p = 0.018), which was attenuated (p = 0.09) after adjustment for inflammation. In Study 2, inflammation burden was not associated with worsening depressive symptoms within the unstratified sample (β = 0.02 [95% CI -0.07, 0.11], p = 0.65). However, sex interacted with degree of inflammation (p=0.005) in this association: whereas females had a worse course of depressive symptoms than males in those with elevated inflammation (β = 0.23 [95% CI 0.08, 0.39], p = 0.003), no sex differences in depressive symptoms were seen in those with low inflammation burden (β = -0.08 [95% CI -0.25, 0.09], p = 0.34). In Study 3, pioglitazone improved depressive symptoms compared to controls (pooled effect size = −0.68 [95% CI −1.12, −0.24], p = 0.003, Nstudies = 8, I2 = 83.2%). Conversely, metformin was comparable to controls overall (pooled effect size = 0.32 [95% CI −0.23, 0.88], p = 0.25, Nstudies = 6). Whereas neither insulin resistance nor glycaemic control moderated antidepressant effects of pioglitazone, female sex (p = 0.016) was associated with greater reduction in depressive symptoms. In Study 4, there was no difference in attrition or adverse events compared to placebo over 12 weeks of treatment with sitagliptin. Despite significantly lower 12-week HbA1c in the sitagliptin group (Cohen’s d = -1.19 [95% CI -1.90, - 0.48]), depressive symptoms at 12 weeks were lower in the placebo arm compared to sitagliptin (d = 0.71 [95% CI 0.13, 1.30]). Further, sitagliptin had only small effects on reducing C-reactive protein concentration (d = -0.32 [95% CI -0.81, 0.17]) and other measures of inflammation.
Conclusions
Depressive symptoms are not associated with worsening glycaemic control in newly diagnosed type 2 diabetes, suggesting that the adverse impacts of depressive symptoms are incompletely explained by this mechanism. Inflammation provides a potentially modifiable target for improving depressive symptoms in type 2 diabetes – at least in females – as well as the adverse association between depressive symptoms and biomedical outcomes. Whereas repositioning of diabetes treatments is a promising and feasible antidepressant strategy in general, sitagliptin is unlikely to improve depression in any powered trial. Future RCTs for depressive symptoms should test medications with stronger anti-inflammatory properties, such as pioglitazone, as well as interventions to target pathways upstream to inflammation, such as obesity and social stress.
Date of Award1 May 2020
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorKhalida Ismail (Supervisor) & Daniel Stahl (Supervisor)

Cite this

'