There is an increased prevalence of depression in people with chronic illness, but this may be particularly so in older patients with diabetes. In this article, S Grimmer and R Mittra summarize the main findings in the literature regarding the relationship between depression and diabetes in older adults, and stress the importance of effective diagnosis and treatment of the two conditions in primary care.

There is an increased prevalence of depression in people with any chronic illness but this may be particularly so in older patients with diabetes¹.

In a study by Amatop and colleagues, 1139 subjects over the age of 65 years were randomly selected from the electoral register and were assessed by a standard interview at home. The tools used were the Geriatric Depression Scale (GDS), the Mini-Mental State Examination (MMSE) and the Activity of Daily Living (ADL) Index. Body mass index was also measured. Individuals with diabetes were identified as those using diet control or drug treatment.

Fasting plasma glucose values from the preceding four weeks were available, and subjects with two values greater than 7.7mmol/litre were also considered to be diabetic.

Information about other chronic illnesses, e.g. pulmonary disease, osteoarthritis, heart disease, cancer, peptic ulcer, renal disease, liver cirrhosis, and gall stones was also documented.

The prevalence of diabetes was determined by gender and age. For age determinations the sample was divided into two groups – less and greater than 75 years. Differences in prevalence between the groups were evaluated by Chi square tests. One-way analysis of variance was used to test differences in mean score for subjects with and without diabetes with respect to age, MMSE score, ADL and BMI. Multiple linear regression analysis of variance was used to control the effect of potential confounders on the GDS score. Non insulin dependent diabetes affected 14.7 per cent of the sample. Depressive symptoms were more prevalent in individuals with diabetes (13.6 per cent) than in those without (8.7 per cent).

On multiple linear regression analysis, diabetes was found to be significantly associated with depression, independent of age, gender, loneliness or chronic disease.

The prevalence of depression in patients with diabetes

A literature review performed in order to determine the prevalence of depression in patients with diabetes², analysed controlled and uncontrolled studies and the specific criteria used to diagnose depression. Some of the controlled studies evaluated depression using structured interviews and some used depression symptom scales. In the studies using structured interviews, there was evidence of an increased prevalence of major depression in subjects with diabetes compared with their respective controls; the prevalence of current depression in the structured interview group of patients with diabetes was 8.5 to 27.3 per cent compared with 11 to 19.9 per cent in uncontrolled studies. These findings were corroborated by studies using depression symptom scales and the range of significant depression was 21.8 to 60 per cent in controlled studies compared with 10 to 28 per cent in uncontrolled studies.

Women with diabetes had a higher rate of depression than men. However, some of the studies can be criticised as control for bias was found to be inadequate. For example, factors such as socio-economic status and obesity may independently cause depression. Assortative mating (using spouses in diabetic and control groups), and concomitant illnesses could also confound results due to other environmental factors influencing psychological wellbeing, for example, if a patient had a spouse with chronic illness, they may be more likely to suffer from depression.

Other biases could be related to the severity of diabetes complications, self-reported diabetes by some patients, life-time or current depression, sample size, and recruiting patients from a selected population (e.g. one hospital).

Lustman and colleagues³ followed a group of patients with diabetes and depression over a five-year-period. They concluded that such patients experience a chronic course of psychiatric illness that is possibly more malevolent than that reported in the medically well. The clinical syndrome of depression is not directly related to advancing diabetes and the mean age of onset of depression is generally well removed from the mean age of onset of diabetes.

Data were inadequate to support the hypothesis that depression in such patients was a reaction to their diagnosis of diabetes or to its complications. The rate of relapse of depression was very high in the diabetes sample and inadequate treatment (none of the patients with depression were offered treatment as it was a naturalistic follow-up study) may be a contributory factor to the apparent tenacity of the psychiatric illness.

Depression and diabetes: interaction at a biological level?

Depression and diabetes may arise due to an interaction at a basic biological level. This may be related to dysregulation of the hypothalamic pituitary-adrenal axis. This could account for a cross-sectional relationship of depression to diabetes. This interesting hypothesis has been explored in studies where depression and stressful life events preceded the onset of type 2 diabetes. Eaton⁴ interviewed a cohort of patients in the Baltimore area to assess the prevalence of psychopathological disorders (the Diagnostic Interview Schedule of the National Institute of Mental Health⁵ was used to diagnose depression). The patients who entered the study were all suffering from varying degrees of depression. They also included questions on diabetes. Any interviewee who gave positive answers to questions about a past or current history of diabetes or being under current treatment were excluded, thus 3481 patients were identified for interviewing. A total of 1897 were re-interviewed 13 years later using a diabetes-specific questionnaire and 89new cases of diabetes were identified. The study concluded that depression may thus relate to future diabetic onset.

The influence of confounding variables, e.g. obesity and that of ascertainment bias could not be totally excluded. A limitation of this work was the problem of undetected diabetes, as data were collected through self-reporting, which would lead to the diagnosis being under-estimated.

Similarly, Kawakami et al⁶ followed up a group of male employees in Japan for 8 years. The incidence rates of type 2 diabetes were higher in those who had moderate to severe levels of depression at baseline: Mooy et al⁷ studied a population of 2262 Caucasian patients without diabetes. Those who reported a major stressful life event in the previous five years (using a validated questionnaire) were given an oral glucose tolerance test⁸. Diabetic curves were identified in 5 per cent of the subjects, but only 3.9 per cent of controls.

Could depression have an aetiological role in diabetes?

Studies suggest that depressive disorders are accompanied by increased activity of the sympathoadrenal system with increased levels of catecholamines in cerebrospinal fluid, plasma, or urine. This is associated with increased blood glucose and impaired glucose tolerance⁹. Depression has been associated with dysregulation of the hypothalamic pituitary-adrenal axis with increased release of glucocorticoids, decreased glucose uptake, and elevated glucose levels. Depression may impair the ability to handle a carbohydrate load and thus increase the risk of developing Type 2 diabetes through a greater release of these counter-regulatory hormones.

It may be that depression has an adverse effect on glycaemic control, but it needs to be established what effect depression has on the long-term course of diabetes.

Lustman¹⁰ completed a meta-analysis of studies involving adults, in which diabetic control was assessed by glycated haemoglobins and depression severity measured by standardized instruments. They found that depression was associated with higher glycated haemoglobins. Thus, depression may be a cause or consequence of hyperglycaemia and there may be a reciprocal interaction between depression and glycaemic control, where depression produces hyperglycaemia and hyperglycaemia provokes depression.

There is also a relationship between depression, hospitalization, and mortality in elderly patients with diabetes. Rosenthal¹¹ carried out a three-year prospective study comparing 135 older patients with diabetes and a cohort of non-diabetic patients. He found that frequency of hospitalization was twice as high in the diabetic group and presence of depression (assessed by the Geriatric Depression Score) was the strongest predictor of death. Certain other factors were more common among patients who died, particularly proliferative retinopathy, symptomatic coronary artery disease, and increased body weight. Rosenthal suggested that depression is a marker for poor outcomes rather than a reflection of poor self-care in diabetes, which may lead to a poor outcome.

Hyperglycaemia-associated depression may have substantial medical consequences, and recognition and proper treatment of the condition is important. Treatment with nortriptyline¹² and fluoxetine¹³ was found to improve depressive symptoms and thus these agents produced a 'trend' towards better glycaemic control. However, more data involving larger numbers of patients and longer duration of treatment and follow-up are needed. A 10-week randomized controlled trial of cognitive behaviour therapy (CBT)¹⁴ of 51 Type2 patients with depression was carried out.

All patients also participated in a supportive diabetes education programme. Ten-week and six-month outcomes were assessed. Depression remission was better in the CBT group at 10 weeks and this was maintained at 6 months compared to the control group who had no active intervention.

No improvement in glycated haemoglobin was noted at 10 weeks, but became apparent at 6months in the CBT group (9.5 per cent compared with 10.9 per cent in controls). Thus CBT may prove an effective treatment modality.

The difference in HbA1c at six months between the CBT and control groups had a P value of 0.03. The authors acknowledge that the numbers were small and the duration of the study was not long enough. The authors also acknowledge that the control group (those who had only education) had less intervention in the form of personal contact with a clinician and thus the beneficial effect of CBT could be a reflection of this.

The nortriptyline study showed that HbA1c reduction was no better in patients treated with nortriptyline than placebo (P=0.5) over a period of eight weeks. In the fluoxetine study, the reduction in Hb1AC over a similar eight-week period was slightly better in patients receiving fluoxetine compared with placebo (P=0.13), though not significant.

Conclusions

In conclusion, the prevalence and severity of depression is increased in patients with diabetes. There may be a biological link between depression and diabetes and the latter may precede the former. Patients with diabetes and depression have a poorer quality of life and worse glycaemic control. They also have a greater rate of hospitalization and increased mortality. It is therefore important to recognize this disorder and treat it effectively. This may be particularly significant in primary care. Whether treatment of depression has any long-term effect on the course of diabetes or its complications remain to be seen.

Dr S Grimmer is Consultant Physician

Dr R Mittra is Specialist Registrar, Department of Medicine for the Elderly, Ipswich Hospital NHS Trust, Ipswich

References

1. Amato L, Padisso G, Cacciatore F, for The Osservatorio Geriatrico of Campania Region Group. Non-insulin dependent diabetes mellitus is associated with a greater prevalence of depression in the elderly. Diabetes Metab 1996; 22(5): 314–8

2. Gavard JA, Lustman PJ. Prevalence of depression in adults with diabetes. Diabetes Care 1993; 16: 1167–78

3. Lustman PJ, Griffith LS, Clouse RE. Depression in adults with diabetes. Diabetes Care 1992; 15: 1631–9

4. Eaton WW, Armenian H, Gallo J, Pratt L, Ford DE. Depression and risk for onset of type II diabetes: a prospective population-based study, Diabetes Care 1996; 19: 1097–1102

5. Robins LN, Helzer JE, Croughan J, Ratcliff KS. National Institute of Mental Health Diagnostic Interview Schedule: its history, characteristics, and validity. Arch Gen Psychiatry 1981; 38: 381–90

6. Kawakani N, Takatsuka N, Shimuzu H, Ishibashi H. Depressive symptoms and occurrence of type 2 diabetes among Japanese men. Diabetes Care 1999; 22: 1071–6

7. Mooy J M, De Vries H, Grootenhius PA. Major stressful life events in relation to prevalence of undetected type 2 diabetes. Diabetes Care 2000; 23: 197–201

8. World Health Organization Diabetes Mellitus: Report of a WHO Study Group. 1985; World Health Organization, Geneva

9. Surwitt RS, Schneider MS, Feingloss MN. Stress and diabetes mellitus. Diabetes Care 1992; 15: 1413–22

10. Lustman PJ, Anderson RJ, Freedland KE, Groot MD, Carney RM, Clouse RE. Depression and poor glycaemic control. Diabetes Care 2000; 23: 934–42

11. Rosenthal MJ, Fajarde M, Gilmore S, Morley JE, Naliboff BD. Hospitalisation and mortality of diabetes in older adults. Diabetes Care 1998; 21: 231–5

12. Lustman PJ, Griffith LS, Clouse RE, Freedland KE, et al. Effects of nortriptyline on depression and glycaemic control in diabetes. Results of a double-blind, placebo-controlled trial. Psychosom Med 1997; 59: 241–50

13. Lustman PJ, Freedland KE, Griffith LS, Clouse RE. Fluoxetine for depression in diabetes. Diabetes Care 2000; 23: 618–23

14. Lustman PJ, Griffith LS, Feedland KE, Kissel SS, Clouse RE. Cognitive behaviour therapy for depression in type 2 diabetes: results of a randomised controlled clinical trial. Ann Intern Med 1998; 129: 613–21

return to main menu