Effect of Intervention on Glycaemic Control among Adult Population with Type 2 Diabetes Mellitus and Comorbid Depression in Central Karnataka, India

Introduction

Both Type 1 and Type 2 Diabetes mellitus (DM) are complex and demanding conditions that can lead to various psychological problems.¹ The existing research suggests that there is a two-way relationship between psychological conditions and diabetes, with some psychological problems increasing the risk of developing diabetes, while others may arise as a result of diabetes diagnosis or complications, which can further exacerbate the disease outcomes. However, the treatment literature indicates that addressing psychological disorders can have a positive impact on both psychological and medical outcomes, including glycemic control. In other words, intervening in psychological conditions may improve the management of diabetes and its associated complications.^1,2

Research has shown that gender, marital status, and body mass index (BMI) are the main factors associated with significant depressive symptoms.^2,3 Additionally, the prevalence of major depressive disorder is high among patients with Type 2 Diabetes mellitus (T2DM), indicating a need for routine depression screening in primary care settings.³ Another study found that depression is a common comorbid health problem in patients with diabetes, with a prevalence rate of 44.7%. The presence of diabetic complications, low income, diabetic nephropathy, negative life events, and poor social support were identified as significant risk factors associated with depression in this study.⁴ Thus concluding that specific attention is needed to diagnose early and treat promptly. Studies report prevalence of clinical depression among diabetics as 28% in the world and 32% in Asia.⁵ In India, various authors have reported widely variable (2%-84%) prevalence of depression in DM.⁶ One of the studies investigated the shared genetic etiology underlying T2DM and major depressive disorder (MDD) where the authors used large-scale genome-wide association studies data to identify 496 Single Nucleotide Polymorphisms (SNPs) that are associated with both Type 2 Diabetes mellitus and Major Depressive Disorder.⁷ Diabetes Related Distress (DRD) and depression were common and correlated in Asian adults with T2DM at primary care level.^8-11 Socio-demographic, more than clinical characteristics, were related to DRD and depression.7 One study concluded that further research should prospectively examine whether a targeted depression treatment goal (PHQ-9 score <15) in patients with diabetes results in improved control of some important disease parameters.^12,13 Several studies have looked at the role of Cognitive-Behavioural Therapy (CBT) & antidepressant medication (ADM) in patients with clinical depression and diabetes.^14-18 With this background, we decided to study the effect of intervention on Glycemic control among patients with diabetes and comorbid depression.

Objective

To assess the effectiveness of intervention in people with T2DM and co morbid depression, for improved glycemic control & depression.

Materials and Methods

Study area: Davangere Study period: Three years (April 2019 to May 2022)

Study population: Adult population aged 30–70 years with Type 2 Diabetes Mellitus (T2DM) attending tertiary care teaching hospital.

Study design: Randomized control trial, concurrent parallel design.

Allocation ratio: 1:1

Eligibility criteria

People in the age group of 30 – 70 years living with type 2 diabetes mellitus, on oral medications/insulin/both for at least 12 months with HbA1c >7.5%, with vital signs within normal limits, willing for psychiatric evaluation and to take anti-depressant medications/cognitive behavior therapy as needed, were recruited for the study after informed written consent.

Exclusion criteria

People with known psychiatric illness or history of suicide attempt in the past 12 months, history of schizophrenia or psychotic symptoms, bipolar disorder, organic brain syndrome or dementia, alcohol, or substance abuse or those on opioid medications due to potential interaction with ADM, were excluded from the study.

Intervention

Once-a-day scheduled administration of ADM Escitalopram starting at 5 mg was done to one group and Cognitive-Behavioural Therapy, a form of psychotherapy, was administered to the other randomized group.

Study design

Single blinded randomised controlled trial. This trial was conducted among type II diabetes patients with comorbid depression attending teaching hospitals in Davangere. Type II diabetes patients in the age group of 30 - 70 years on oral hypoglycaemics or insulin or both for past one year, with current mild-to-moderate depression according to Patient Health Questionnaire (PHQ-9) scale19 (score 1-14) and with HbA1c >7.5% within at least three months at the time of screening and those willing to be part of the trial were included in the study.

Sample size

The sample size was calculated based on the minimum difference expected between the groups (intervention and control) and the amount of variation within the groups i.e., Standard Deviation/Standard Error of Proportion. With level of significance set at 5% and power of study 80%, sample was calculated to be 26 in each group (ADM and CBT). With dropout rate of 20%, sample size of 32 was taken in each arm, bringing the total sample size to 64. Stratified block randomization technique was used to assign the subjects into experimental and control groups. The block allocation was concealed from both the investigator and the participant by using Sequentially Numbered Opaque Sealed Envelopes (SNOSE).

Study tools

PHQ – 9 questionnaire: The Patient Health Questionnaire (PHQ) is a tool for making criteria-based diagnoses of depressive and other mental disorders commonly encountered in primary care. The PHQ-9 can establish depressive disorder diagnoses as well as grade depressive symptom severity.¹⁹

MADRS Scale: The Montgomery and Asberg Depression Rating Scale (MADRS) is a diagnostic tool used to assess the severity of depressive episodes in people with mood disorders. It consists of ten items. Each item on the MADRS generates a score between 0 and 6, with a higher score indicating more severe depression. The total score is on a scale from 0 to 60. The following symptoms are covered by the questionnaire's questions: 1. Outward signs of sorrow; 2. Reports of sadness; 3. Internal tension; 4. Restlessness; 5. Decline in appetite; 6. Problems concentrating; 7. Laziness 8. Lack of emotion; 9. Pessimistic thoughts; 10. Ideas of suicide. Normal/symptom-free ranges for cutoffs are 0 to 6, mild depression is between 7 and 19, moderate depression is between 20 and 34, and severe depression is between 34 and above.²⁰

Cognitive Behaviour Therapy (CBT): This is a form of psychotherapy that teaches patients how to recognise and alter unhelpful or unsettling thought patterns that have an adverse impact on their emotions and behaviours.²¹

Method

After obtaining clearance from Institutional Ethics Committee as well as registration under CTRI (CTRI/2017/08/009250), consecutive T2DM patients visiting medicine OPD who met the inclusion criteria were recruited for the study. The patients were informed of the study aims and protocol, and once they voluntarily agreed to participate, written informed consent was obtained. PHQ–9 questionnaire was administered and patients with a score 5-14 were invited to be part of the study. Those who were selected for the trial were allocated into experimental (CBT) and control groups (ADM) based on block randomization technique with their allocation carefully concealed. ADM Escitalopram at the dose of 5 mg Once-Daily (OD) was administered to control group and Cognitive-Behavioural Therapy (CBT), a form of psychotherapy was administered to the intervention group. A pre-tested semi-structured proforma was used to collect information regarding socio-demographic details, clinical and laboratory findings including baseline HbA1c during in-person interviews. The intervention group participated in CBT delivered across ten sessions. The modules included introduction to CBT, mood monitoring, thought monitoring, aid in decision-making related to self-care, and relaxation training.²¹ Both groups were followed up and reassessed at 3-month, 9-month intervals and at the end of follow-up at 12-months for improvement in symptoms of depression as rated on the MADRS. HbA1c was assessed at baseline and at 3-month, 9-month and at 12-month intervals for a total of four times till the end of the study.

Assessment of adherence to intervention

• Adherent – to CBT (Participation in ≥ 8 sessions of CBT) & to ADM (If the patient has taken medication for more than six months)

• Partially non-adherent – to CBT (Participation in 1-7 sessions of CBT) & to antidepressive medication (ADM) (If the patient has missed medication for maximum of 5 days in a month)

• Non-adherent – to CBT (No participation in any sessions of CBT) & to ADM (If the patient has missed medication for more than 5 days in a month)

Data analysis

Data was analyzed with IBM SPSS v26.0, and followed per protocol analysis in this study to compare the effect of intervention on glycaemic control after adjusting for age and gender. Data were presented in percentages and proportions. Statistical significance was considered at an alpha level of 5%. The main secondary analysis was based on the numbers and proportions of patients who demonstrated improvement in glycaemic control (≥ 1% improvement in HbA1c values from baseline to the end of the long-term phase). Differences between groups was tested using logistic regression analysis controlling for baseline HbA1c value and baseline MADRS-score.

Results

Among 208 patients approached, 138 were excluded who did not give the consent and did not meet the inclusion criteria. Remaining 67 patients were allocated into two groups, 35 and 32 respectively through block randomization. Among 35 participants who underwent CBT, 32 completed three months follow-up and among 32 participants who received ADM, 28 completed the follow-up. At nine months, 4 (28.5%) participants were lost to follow-up in CBT group and 6 (43.75%) were lost to follow-up in ADM group. Only 25 were left for analysis in CBT group and 18 in ADM group at the end of the follow-up. None of the participants were dropped because of worsening of the depression and there were no study related adverse events among the participants (Figure 1).

Profile of study participants

Majority of the participants in either of the groups were females (40, 59.7%). Among the study participants, majority (12, 34%) of the patients in CBT group were in the age group 41- 50 years, with the mean age being 44±5.12 years, while in CBT group, majority (11, 34%) of them were in the age group of 31-40 years, with a mean age of 42±3.36 years. The majority of the study participants were Hindus (35, 52.3%) and most of them belonged to Class III socio economic class (29, 43.3%) as per modified BG Prasad classification (updated for January 2019) and the majority of the participants (28, 41.8%) were currently working or employed in occupation (Table 1).

Adherence of study participants to interventions

Among the study participants allocated to CBT arm, by the end of 12-month follow-up, 10 (28.6%) dropouts were observed. Among those who completed one year follow-up, majority (19, 76%) attended more than eight sessions. In the ADM arm, 14 (43.8%) dropouts were observed. Among those who completed the one year follow-up, majority (11, 61%) were adherent to medication. Loss of interest, and death were the most common causes for participants to leave the study.

Post-intervention outcome

The mean glycosylated hemoglobin was reduced to 7.7% from 9.231% in CBT group and in the ADM group, it reduced to 9.012% from 9.556%. The MADRS score was reduced to 9.17 from 27.68 in CBT group and in the ADM group, it reduced to 20.63 from 29.303 (Table 2 and 3).

Post-intervention (Acute/3-Month) outcomes: Depression - Controlling for baseline, the CBT arm had 6.23 lower depression scores on the MADRS than the ADM arm; the lower scores indicate less severity of depression in both the groups. Diabetes Control - After adjusting for baseline, the CBT arm had better glycemic control, as seen by a decrease in HbA1c compared to the ADM group (Table 2 and 3).

Post-intervention (1-year follow up) outcomes: At posttreatment, controlling for baseline, the CBT arm had 15% higher adherence as assessed by follow-up visits (95% CI 14.11 to 16.12, p = 0.0001). Depression: Controlling for baseline, the CBT arm had lower mean depression scores on the MADRS than the ADM arm (95% CI -10.068 to –9.606, p <0.0001). Diabetes control: After adjusting for baseline, the CBT arm had better glycemic control (Table 2 and 3). 

Discussion

The study was designed to test whether treating depression through an evidence-based approach like CBT would be superior to pharmacological intervention in improving glycemic control among diabetes patients with comorbid depression. Several previous studies have showed CBT as a widely studied treatment for depression, and well-documented depression treatment trials among people with diabetes have already demonstrated that treatment for depression will not consistently improve glycemic control and/or diabetes self-management practices.

Our study went with an approach to evaluate if an intensive intervention will have greater reduction in depression levels (as measured by MADRS score) and improve glycemic control (as measured by HbA1c levels). Post three months intervention, in CBT arm, participants had lower mean HbA1c (0.188% reduction), better adherence as assessed by the follow-up visits and better reduction in MADRS score (6.23-unit reduction) as compared to the group of participants on ADM. In CBT group, gains attained by reduction in mean HbA1c was maintained over the follow-up period at 12-months. The difference in the mean HbA1c levels between the arms was evident as observed by the ANCOVA model.

For depression, the participants in CBT group showed lower mean MADRS scores all throughout the follow-up period, at 3-months (6.23 points reduction), 9-months (15.4 points reduction) and at 12-months (18.51 points reduction), as compared to the participants in ADM group. Although mean scores of both HbA1c and MADRS showed continued improvement over the follow-up period in both the arms, it was more robust and evident in CBT arm compared to ADM group. Among the participants in ADM group, the immediate effect (post 3- month) was not encouraging when we look at mean HbA1c values (+0.238%) and mean MADRS scores (+1.147 points), but at 9-months and 12-months, we were able to see better outcomes. However, depression scores during the follow-up did not worsen in either of the groups, and data showed a trend for continued improvement in scores on the MADRS.

Whenever clinically indicated, participants were referred for further care for depression. The maintenance of clinically significant effects on glycemic control over the follow-up, and depression suggests some degree of independence between these outcomes. This is in contrast with some of the previous trials that were successful in treating depression in diabetes but failed to demonstrate corresponding benefits for self- management or glycemic control.^22,23

In our study, we integrated adherence counseling as part of CBT sessions. The rationale for this was based on previous studies where association between depression and poor diabetes treatment adherence was evident and the failure of previous trials consistently impacted the glycemic control or diabetes self-management.^13,23 Another similar study was successful in improving depression and glycemic control by improving adherence among patients on ADM.²⁴ There were also trials where interventions failed to improve selfmanagement practices or medication adherence but improved depression and glycemic control.²⁴ In another trial, medications for depression, blood pressure, lipid and glycemic control were also provided to participants where improvements in glycemic control was seen in the absence of improvements in self-management.²⁵ Thus, the present study adds to the evidence that treating depression is necessary to improve the diabetes outcomes in patients living with diabetes mellitus and comorbid depression. Treating depression will help patients to improve their self-management practices and it was visible in both of our intervention and comparison arms of the present study. Some of the studies also showed better glycemic control in patients with ADM compared to the those not on ADM. Overall, the results of our study helped us to recommend that this approach could bring about improvements in glycemic control.

Limitations

The study also has few limitations to note. Our study design does not use any scales for medication adherence in either of the groups, and because of COVID, we had more loss-to-follow up in both the study arms. Finally, use of ADM among study participants also affects the inter-participant outcomes because of variability in dose.

Conclusion

Time and again, it is evidently proved that prevalence of clinical and/or sub-clinical depression is high among diabetics and is associated with poor self-management practices, outcomes, and mortality. Thus, there is a need for evidence-based treatment options for improving glycemic control among people living with diabetes mellitus and comorbid depression.

Key Message

Our study concludes non-pharmacological interventions could effectively improve severity of depression and glycemic control among people with type 2 diabetes with comorbid depression.

Ethical policy and Institutional Review board statement

The research study was approved by the Institutional Ethical Committee (JJMMC/IEC/29-2017) and registered with the Clinical Trial Registry – India (CTRI/2017/08/009250). The authors certify that they have obtained all appropriate patient consent forms. In the form, the patient (s) has/have given his/her/their consent for the clinical information to be reported in the journal. The patients understand that their names and initials will not be published, and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.

Conflict of Interest

There are no conflicts of interest

Financial support & sponsorship

The study was supported by Rajiv Gandhi University of Health Sciences (RGUHS), Bangalore under Faculty Research Grants for Advance Research. Grant No.: M016/2019

Acknowledgement

The authors wish to acknowledge all the research staff in department of Community Medicine for their support & help in data collection. We also thank our study participants for their cooperation during the study. The authors also thank the principal of the institute for kind support throughout the study.