International Journal of Collaborative Research on Internal Medicine & Public Health

editor@imedpub.com Submit a Manuscript

Depression and associated factors in diabetic patients attending an urban hospital of Bangladesh

Mahbubur Rahman , Md. Anisur Rahman , Meerjady Sabrina Flora , Muhammad Rakibuz- Zaman

Mahbubur Rahman1*, Md. Anisur Rahman2, Meerjady Sabrina Flora2, Muhammad Rakibuz-Zaman3
  1. Department of Community Medicine, Faridpur Medical College, Faridpur, Bangladesh
  2. Department of Epidemiology, National Institute of Preventive and Social Medicine (NIPSOM), Bangladesh
  3. Columbia University Arsenic and Health Research in Bangladesh, Dhaka, Bangladesh
Corresponding Author: Email: dr_mahbub@yahoo.com
Related article at Pubmed, Scholar Google

Abstract

Introduction: Diabetes mellitus is frequently associated with co-morbid depression, contributing double burden to the individual and society.

Aims & Objectives: To find out the proportion of depression among patients with type 2 diabetes mellitus and to determine factors associated with it.

Methods: A cross sectional study was conducted among 178 patients with type 2 diabetes mellitus attending Bangladesh Institute of Research and Rehabilitation in Diabetes, Endocrine and Metabolic Disorders (BIRDEM), Dhaka. Data were collected through face-to-face interview and reviewing documents. Depressive symptoms were measured using Centre for Epidemiological Studies Depression Scale; considering score 16-21 as mild to moderate depression and score F 22 as severe depression.

Results: The proportion of depression among the study population was 34.8% which included 14.6% with mild to moderate depression and 20.2% with severe depression. Both mild to moderate and severe depression were more common in females and among singles. Insulin users were six times more likely to develop severe depression (OR, 6.33 with 95% CI, 1.75 to 22.89) than users of oral anti-diabetic agents. Glycemic status measured by HbA1c was the best predictor. Both poor and fair glycaemic control were associated with any level of depression. Patients with poor glycemic control had odds ratio of 4.75 for mild to moderate depression (95% CI=1.37-16.41) and 10.39 for severe depression (95% CI,=3.66-29.43) in reference to good glycemic control. Patients with fair glycemic control were four times more likely to have mild to moderate depression (OR, 4.31 with 95% CI, 1.57 to 11.85) and severe depression (OR, 3.77 with 95% CI, 1.42 to 10.02) than patients with good glycemic control.

Conclusion: Depression was identified as a significant health problem among patients with type 2 diabetes mellitus. Both diabetes and depression should be considered simultaneously during treatment plan.

Keywords
Bangladesh, depression, HbA1c, type 2 diabetes mellitus

Introduction
Diabetes mellitus is being increasingly recognized as a serious global health problem. In 2000, Bangladesh had 3.2 million people with diabetes and was listed globally at 10th, which would occupy the 7th position with 11.1 million in 2030.1 The prevalence of type 2 diabetes in Bangladesh was 5.2% (rural 4.3%, urban 6.9%) in 1994-952 and increased to 11.2% (urban)3 and 6.8% (rural) in 2003-04.4 Diabetes mellitus is frequently associated with co-morbid depression, contributing double burden to the individual and society.5 These associations may be related to increased risk of depressive symptoms in individuals with diabetes, increased risk of type 2 diabetes in individuals with depressive symptoms, or both.6 At any single point in time, approximately one-third of diabetic patients have symptoms of depression severe enough to warrant treatment.7 However, depressed adults have 37% increased risk of developing type 2 diabetes mellitus.8
Depression plays an important role in nonadherence to medical treatment. Therapeutic compliance, especially diligence in maintaining dietary restrictions and exercise, often declines in depressed patients for their irregular life styles and loss of interest in health.9 This further leads to poor glycaemic control, increased diabetes-related complications, increased diabetic symptoms burden10-12 and ultimately increased health care use and costs.13 Both minor and major depression are strongly associated with increased mortality in patients with type 2 diabetes.14,15 On the other hand, support to encourage psychological well-being would be expected to reduce not only the occurrence of psychological problems but also metabolic problems and complications.12
Despite the specific relevance of depression to diabetes and its serious impact on the disease and health care system, it is estimated that only one-third of people with both diabetes and major depression are recognized and appropriately treated for both disorders.16 Clearly identifying diabetic patients with comorbid depression, knowledge about effective psychological support and improving access to effective treatments should be public health and research priorities. This important association between type 2 diabetes mellitus and depression and its consequences on the outcome of diabetes mellitus were observed by studies, conducted in western populations5,7,9. A few studies were done in Asia.17 Data on depression in the general population in South Asia, including Bangladesh are inadequate and so in type 2 diabetic patients. So far reviewed, only one such study was found which was done in rural Bangladeshi.18 Diabetes is more prevalent in urban than in rural population2 and they are frequently exposed to a number of unfavorable conditions which make them vulnerable to depression. Therefore, this study was designed to explore the situation in urban population of Bangladesh.

Subjects and Methods
A cross-sectional study was undertaken from January to June 2009 in the Outpatient Department (OPD) of Bangladesh Institute of Research and Rehabilitation in Diabetes, Endocrine and Metabolic Disorders (BIRDEM), which is the largest tertiary diabetic care hospital in Dhaka and provides services to large number of patients with diabetes mellitus. Adult patients (age F 18 years) with type 2 diabetes mellitus selected by convenient sampling were invited to participate in the study and 93% agreed. Patients were diagnosed as diabetic by the attending physicians of BIRDEM as per American Diabetic Association guidelines. Those who had psychiatric problems before diagnosis of diabetes, family history of depression, been taking anti-depressant drugs and seriously ill were excluded from the study. Estimated sample size taking predicted prevalence 29.7%18at 95% confidence level with 7% absolute precision and considering 10% non response rate was 181. Three samples were excluded during data analysis for having missing data. Finally, data of 178 patients with type 2 diabetes mellitus were analyzed.
The protocol was approved by the Ethical Committees of National Institute of Preventive and Social Medicine (NIPSOM) and BIRDEM. Informed written consent was obtained from each individual prior to data collection. Data were collected by interview, record review and anthropometry. A semistructured questionnaire was developed based on Centre for Epidemiological Studies Depression (CES-D) scale. The CES-D contains 20-items using a four-point rating scale. Scores range from 0 to 60, with higher score indicating more severe depressive symptoms.19 CES-D has 61.4% sensitivity and 81.0% specificity. Positive and negative predictive values of the scale were 57.5% and 83.3%, respectively.20 A score of 16 or greater differentiated depressed from non-depressed adults19 and cut-off 22 was used to distinguish severity of depression; a score 16-21 for mild to moderate depression and F 22 for severe depression.15,21 Cronbach’s K of Bangla version of CES-D was calculated 0.89, which indicated sufficient internal consistency. Each questionnaire took approximately 35 to 40 minutes to fill up.
Depression score was constructed by summation of all 20 items of depression score. As depression score did not follow normal distribution, for statistical analysis log transformation of the depression score was done and geometric mean was used for comparison. To determine glycemic status, HbA1c level was categorized as HbA1c level < 7% as good glycemic control, 7 to 8 fair glycemic control and > 8% considered as poor glycemic control. Statistical comparisons between different groups were made using ttest, ANOVA for mean scores and chi-square tests for level of depression. The odds ratio (OR) with 95% confidence interval (CI) for risk factors was calculated. All the tests were two tailed and p<0.05 was considered to be statistically significant. Multiple logistic regressions were performed to adjust for potential confounding factors.

Results
Among 178 respondents, 51% were male. Their mean age was 54.96 years (SD 9.76 years). Most of them were from urban area (72%) and married (82%). Average duration since detection of diabetes was 10.10 + 6.15 years. Mean BMI of the patients was 25.22 (SD 3.25). Majority of them were treated with oral anti-diabetic agents (60.5%). Average HbA1c level (%) was 7.2 ± 0.97 and 36.5% of the respondents presented with complications of diabetes (Table 1).
Mean depression score was calculated 10.50 with SD 9.08. The proportion of depression among the study population was 34.8% (CESD score F 16) which included 20.2% with severe depression (CES-D score F 22) and 14.6% with mild to moderate depression (CES-D score 16 to 21).
Mean depression score found higher among females (p<0.001), singles (p<0.001), less educated (p<0.005) and housewives (p<0.005) (Table 1). Duration since detection of diabetes mellitus showed positive correlation with depression score (r = 0.171, p < 0.05). Depression score was also higher among insulin users than those orally treated (p<0.005) and in those used syringe for insulin administration than pen users (p<0.05). Presence of diabetic complications was also associated with higher depression score (p<0.05). Moderate positive correlation was found between glycemic status of the respondents and depression score (r = 0.331, p < 0.001). For each unit increment of HbA1c level, depression score was increased by 3.345 unit (b = 3.345). Respondents having good glycemic control had least mean depression score, whereas those with poor glycemic control showed highest depression score (Table 1). BMI was not correlated with depression score. Multiple linear regression analysis model was constructed to identify predictors of depression score. Glycemic status (HbA1c %) was the best predictor of depression score, followed by marital status, gender and use of insulin device (Table 2).
Higher percentage of both mild to moderate and severe depression were found among female respondents (p<0.001), singles (p<0.001) and respondents with up to secondary level of education (p<0.05). Housewives suffered more from severe depression and retired persons from mild to moderate depression (p<0.05) (Table 3).
Treatment with oral anti-diabetic drugs and insulin together was associated with mild to moderate depression, while only insulin use was associated with severe depression (p<0.01). Both mild to moderate and severe depression were found prevalent among patients with poor glycemic control (p<0.001) (Table 4).
Logistic regression model was constructed which identified HbA1c level, marital status and gender as the best three predictors of depression. Diabetic individuals with fair glycemic control were six times and poor glycemic control were seven and half times more likely to develop depression than individuals with good glycemic control. Single respondents were five times and female were three times more likely to develop depression than their counterparts (Table 5).

Discussion
In this study, high proportion of depression (34.8%) was found, which was much higher than that in adult general population of Bangladesh (4.6%)22 and it supported the link between diabetes and depression. Most of the earlier studies supported the higher prevalence of depression among patients with type 2 diabetes mellitus5,23 except a few.24 The current study proportion was also higher than that in rural diabetic population (29.7%).18.This difference might accountable to urban - rural difference in sample characteristics. Moreover, rural prevalence data were community-based while the current study was done in a hospital. Earlier studies found that prevalence of co-morbid depression was higher in clinical than in community samples.5 A recent study done in United Arab Emirates found that 12.5% of diabetic patients obtained a score of 19 or above on the K6, indicating possible mental health concerns.17 This lower prevalence might be attributed to the differences in socio-demographic characteristics of the samples. Nationalities other than Emirati in the same study showed higher rates of scores of 19 or above. Further, although K6 is an efficient screening tool for “serious mental illness” (SMI) with a sensitivity of 0.36 and specificity of 0.96,25 but its efficiency as a screening tool for depressive symptoms is yet to be explored.
Mean depression score among females was found significantly higher than males. After controlling for other socio-demographic variables, gender still was significantly associated with presence of depressive symptoms [OR = 5.107, 95% CI (1.197- 21.792)]. This finding was supported by most of the other previous studies18,23 but not all.26 The mean age of the respondents was 54.96 years and 60% of them were in the 45 to 64 years age group and age distribution was comparable with that of type 2 diabetes in developing countries.27 No statistical significant relation was found between age and depression. It is consistent with other studies.26,28 One study found that major depressive disorder was more common in 31- 59 year old group.29 Current study found 40- 49 years age group suffered most from severe depression, athough 60 – 69 years group had highest percentages of overall depressive symptoms.
Single respondents had shown higher depression score than their married counterparts [OR = 4.183, 95% CI (1.589- 11.010); p < 0.01] which was consistent with other study findings.13,28
Educational status of this study sample was much better than the national figure.30 Selection of urban health centre and residence of the respondents might explain this. Although mean depression score was found higher in lower education group, after adjustment for other sociodemographic variables the association did not exist. Association of education with depression was revealed by most other studies24,31,32 except a few.28
Mean depression score found highest among housewives and lowest among businessmen and this was because depression associated with female gender and females were most commonly housewives in Bangladesh and so in this study. Miyaoka et al. detected correlation of unemployment with depression score in their study.28
Average monthly incomes of depressed and non-depressed group were almost equal in this study which is not consistent with other study findings of higher prevalence of depression in low income individuals.13,28,33
Mean BMI among the depressed respondents was 25.78 ± 3.59, which was close to the findings of study among rural population of Bangladesh.18 No significant association could be found between BMI and depression, which was supported by most of the other studies18,31, while some other studies found the association.14
On average, duration of diabetes in depressed group was two years longer than in nondepressed group. Depression score was correlated with duration of diabetes mellitus. Other studies did not detect this relationship.26,28
In this study more than 90% of the respondents received pharmacological treatment in addition to diet and discipline. Respondents with pharmacological treatment had higher depression score. Treatment with oral agent revealed lowest depression score and higher score was found when insulin was used. Patients viewed oral treatment as the least and insulin as the most burdensome treatment34 and insulin was associated with higher frequency of depression.35,36 Those who used conventional method (syringe) for insulin administration had higher mean depression score than pen device users. Pain of injection might be the reason for higher prevalence of depression among insulintreated patients.
Depression is significantly associated with a variety of diabetic complications (diabetic retinopathy, nephropathy, neuropathy, macrovascular complications, and sexual dysfunction).26,28,37 In this study mean CES-D scores in diabetic patients with complications were found higher than in patients without complications. But no relation was found.
HbA1c has been considered as the direct indicator of glycemic status of a diabetic individual and its estimation as the most accurate way of monitoring how well a diabetic patient is being controlled.38 Moderate correlation was revealed between depression score and HbA1c level, which was supported by other studies29,37,38, while Engum et al. found no significant association between them.24
Although the present study was not based upon a representative sample, it provided a broader basis for the estimation of proportion of depression and factors associated with it among patients with type 2 diabetes mellitus. Within limitations, the researchers tried to have the best possible sample by choosing BIRDEM, which is the largest service provider for diabetic patients in Bangladesh. Cross-sectional design of the current study prevented the researcher to infer about the causality between depression and diabetes. Further, assessment of depression was based on self-report, using a screening tool, rather than a gold-standard diagnostic tool. However, CES-D is a widely used, well validated measure for depressive symptoms, particularly suited for epidemiological studies, both in general population and in diabetes patients. To minimize recall bias, some data were validated by reviewing the documents.

Conclusion
This study has identified depression as a significant health problem among patients with type 2 diabetes mellitus and highlighted some of the factors associated with depression among them. This association need to be further studied in depth and finding of the current study should be replicated in order to formulate effective intervention program.

Acknowledgements
We acknowledge authorities of Bangladesh Institute of Research and Rehabilitation in Diabetes, Endocrine and Metabolic Disorders (BIRDEM) for their support. We are also grateful to all the respondents for their sincere cooperation.

Disclosure
None to declare any competing interest.

References






































image
Table 1: Depression (CES-D) score and socio-demographic and clinical characteristics of the respondents
image
Table 2: Factors associated with depression score
image
Table 3: Level of depression and socio-demographic characteristics of the respondents
image
Table 4: Level of depression and clinical characteristics of the respondents
image
Table 5: Risk factors associated with depression
open access journals, open access scientific research publisher, open access publisher
Select your language of interest to view the total content in your interested language

Viewing options

Flyer image

Open Access Journals

Share This Article

paper.io

agar io

wowcappadocia.com
cappadocia-hotels.com
caruscappadocia.com
brothersballoon.com
balloon-rides.net

wormax io