Results of this study clearly showed that depression was associated with poor treatment outcomes in included patients. For instance, depressed patients had significantly poorer glycemic control than non-depressed patients. Results further indicated that this poor outcome could be explained by lower compliance to the recommended self-care activities such as taking dose on time, dietary restrictions, and foot care.
Before data collection of this study, we found only two studies from Pakistan that assessed mental health of diabetic patients. Both studies compared diabetic patients and with non-diabetic controls, and included 50 or fewer patients in each group [25
]. Depression was found in greater proportion in the diabetic groups than non diabetics in both studies. A recent study by Khuwaja et al (2010) showed that 43.5% of patients attending diabetic clinics in Karachi had depression [27
]. The study further indicated that blood pressure, fasting blood sugar, triglycerides, and BMI were associated with depression. However, measures of glycemic control and compliance to self-care activities were not included in that study. Our results clearly pointed out that information of treatment outcomes in depressed diabetic patients might help clinicians to better monitor such patients.
Our sample contained mostly patients with long duration of illness, diabetes-related complications, and less than half following self-care activities recommended by their endocrinologist. Results showed that majority of them had poor glycemic control which was consistent with previous findings in Pakistan showing that 78% of the diabetic patients had poor glycemic control [28
]. The higher likelihood of poor glycemic control in the depressed patients than non-depressed patients as indicated in this study was consistent with findings from other international studies [21
]. Moreover, our study indicated that depression was higher in patients with diabetes-related complications such as renal, cerebrovascular, cardiac and neuropathic and ophthalmic [30
]. The possible explanation for above findings might be that the depression and psychological stress increase sympathetic nervous system activity, inflammatory, and platelet aggregation, and decrease insulin sensitivity, thereby contributing to poor glycemic control and increasing the risk of complications [31
]. This suggested that regular evaluations of depression and glycemic control might be needed in diabetic clinics.
In our study, female gender was significantly associated with depression. It was previously shown that female gender was associated with depression in Type 2 Diabetes [33
]. In a meta-analysis by De Groot et al, the combined prevalence of depression was significantly higher in Type 2 diabetic women than Type 2 diabetic men [17
]. Our results showed that females were also less likely to follow recommendations of exercise and more likely to have greater BMI values. This suggested that women patients with Type 2 Diabetes Mellitus deserve special clinical attention in terms of ensuring that lifestyle changes were adopted by them.
Our study showed that depressed patients were less likely to follow necessary treatment recommendations such as taking proper doses, dietary restrictions, regular exercise, and foot care. Regression removed exercise from the final model however, other associations remained significant. This was consistent with previous findings of Lin et al showing that depression was associated with less physical activity, unhealthy diet, and low adherence to oral hypoglycemic, anti-hypertensive, and lipid-lowering medicines [12
]. Another prospective cohort study showed that depression could predict lower medication adherence as well as adherence to dietary recommendations and exercise [35
]. This suggested that more robust monitoring of depressed diabetic patients might be needed, perhaps by involving their family members in the process of care [36
This study has several limitations. Firstly, the cross-sectional design is not without limitation since causality cannot be established. Although it may be argued that the possibility of selective loss of follow-up or non-random treatment changes is eliminated in such a design. Secondly, educational status and household income were not measured in this study although they have associations with both depression and diabetic patients with depression [35
]. However participants can be considered to belong to middle or upper-middle socio-economic strata as they are coming to a specialty clinic and medical expenses are mostly borne out of pocket in Pakistan. Thirdly, self-report was used as a measure for self-care activities as no scale is validated in our population. Nevertheless, this method is highly sensitive to detection of non-adherence to medicines [38
]. Lastly, generalization of study findings would require similar studies in other settings in Pakistan.
In conclusion, this study suggested that patients with type 2 diabetes should be screened for depressive symptoms, poor glycemic control, and non-compliance to self-care activities. Further, women and those who had a family history of diabetes should be closely watched for depressive symptoms. Finally, developing a valid instrument to measure compliance to self-care activities might be required to improve care in Pakistani diabetic clinics.