Prevalence of symptoms of depression and anxiety in a diabetes clinic population C. E. Lloyd*, P. H. Dyer² and A. H. Barnett²
Abstract *School of Health & Social Welfare, The Open University, Milton Keynes, UK ²Department of Medicine, University of Birmingham and Birmingham Heartlands Hospital, Birmingham, UK Received 15 February 1999; revised 7 December 1999; accepted 9 January 2000
Aims To investigate the use of a short questionnaire to measure psychological symptoms in a busy clinic setting, and to examine the prevalence of these symptoms in adults with diabetes. The perceived need for psychological treatment services was also measured. Methods Adults (> 18 years) with either Type 1 or Type 2 diabetes were
invited to complete a short demographic form and the Hospital Anxiety and Depression Scale (HADS) whilst waiting for their routine diabetes outpatients appointment. Complication status was measured via patients' medical records. Glycaemic control (HbA1c) was also recorded.
Results A high response rate (96%) was achieved. Prevalence rates of
psychological symptoms were high (overall 28% of study participants reported moderate±severe levels of depression or anxiety or both). Men were somewhat more likely to report moderate±severe depressive symptoms, whereas women reported more moderate±severe anxiety. A signi®cant association between depression and poor glycaemic control was observed in the men, but not in the women. Regression analysis demonstrated that the interaction between sex and glycaemic control, HbA1c and sex were all signi®cantly associated with depression and anxiety (R2 = 0.16 and 0.19, respectively). One-third of subjects reported that at the moment they would be interested in receiving counselling or psychotherapy if it was currently available at the diabetes clinic.
Conclusions This study has shown that the HADS is an appropriate
questionnaire to use in a clinic setting in adults with diabetes. There may be a stronger association between glycaemic control and psychological symptomatology in men than in women. There remains a signi®cant proportion of individuals with diabetes who require psychological support, which, if available, might help improve glycaemic control and thus overall wellbeing. Diabet. Med. 17, 198±202 (2000)
Keywords anxiety, depression, glycaemic control, Type 1 diabetes mellitus,
Type 2 diabetes mellitus
Abbreviations CAD, coronary heart disease; HADS, Hospital Anxiety and
Depression scale
Introduction Correspondence to: Dr Cathy E. Lloyd, The Open University, School of Health & Social Welfare, Walton Hall, Milton Keynes MK7 6AA, UK. E-mail:
[email protected] 198
Psychological morbidity may be more prevalent in diabetic individuals than in those without diabetes, although the reasons for this remain unclear [1,2]. Suffering from a chronic condition such as diabetes may lead to increased ã 2000 British Diabetic Association. Diabetic Medicine, 17, 198±202
Original articles
feelings of anxiety or depression, particularly if the individual experiences dif®culties with glycaemic control or develops complications [3]. Psychological problems may have an adverse effect on diabetic control, and improvements in depression can lead to clinically signi®cant improvements in glycaemic control [4]. There have been some concerns that prevalence rates of psychological symptomatology may be overestimated in diabetic populations, because some of the questionnaires used to measure depression contain items that measure somatic symptoms of this condition, some of which could be confounded with symptoms of uncontrolled diabetes [5,6]. However Lustman and others have shown that instruments which do include somatic symptom items can accurately measure depression in individuals with diabetes [2±4]. Despite these concerns, there has been some small growth in the provision of psychological treatment for patients with diabetes. These treatments have shown some success, and it has been demonstrated that psychological treatment such as psychotherapy can improve psychological wellbeing as well as diabetic control [2,4,7]. Prior to any attempt to develop a service to provide psychological care, however, it is necessary to determine the extent of the need for psychotherapy (or other psychological treatment) locally. In accordance with World Health Organization recommendations [8], the Diabetes Directorate at Birmingham Heartlands Hospital proposed the development of psychological services for individuals attending diabetes outpatients. It was decided, however, to ®rst investigate the prevalence of psychological morbidity in a pilot study, and to examine the perceived need for psychological treatment services (including counselling and psychotherapy) amongst patients attending the diabetes outpatient clinics.
Patients and methods: Consecutive adult (> 18 years) clinic attenders with diabetes were approached by one of the diabetes nurses during their routine diabetes clinic visit, and invited to participate in this study. In order to ensure that it was possible to analyse the data collected according to type of diabetes, individuals were approached until a target number of patients with Type 1 or Type 2 diabetes was attained. Those who agreed to take part completed a short demographic questionnaire and the Hospital Anxiety and Depression Scale (HADS) [9]. The HADS consists of 14 items, with two sub-scales measuring symptoms of depression and symptoms of anxiety [9]. Although few data have been published with regard to the use of the HADS in diabetic patients, this well-validated questionnaire has been used extensively in many other chronic disease populations [10,11]. Moreover, since it does not contain any items that measure somatic symptomatology, symptoms of depression/ anxiety cannot be confounded with the physical symptoms of uncontrolled diabetes. Scores of 11 or higher on either or both of the sub-scales are used to indicate moderate±severe symptoma-
ã 2000 British Diabetic Association. Diabetic Medicine, 17, 198±202
199
tology or clinically signi®cant disorder [9]. Scores of 8±10 on either sub-scale indicate mild levels of anxiety or depression. HbA1c was recorded from patients' medical records as a measure of glycaemic control (non-diabetic range < 5%). Satisfactory glycaemic control was de®ned as HbA1c < 7%, moderate control 7±9%, and poor control > 9%. Both the UK Prospective Study of Diabetes [12] and the Diabetes Control and Complications Trial [13] have suggested targets for satisfactory control as HbA1c values of less than 7%. The present cut-off of 9% for poor control mirrors the clinic cut-off, which is an arbitrary one. Whilst there is no evidence base for this, poorer control is associated with greater risk of microvascular complications. Type 1 diabetes was de®ned as acute onset prior to the age of 30 with immediate commencement of insulin therapy, the remainder being classi®ed as Type 2 diabetes. Complications, ascertained via the screening of medical records, were recorded as retinopathy (background or proliferative), neuropathy (absence of ankle jerks and reduced vibratory sensation, or impairment on nerve conduction studies or autonomic function tests), and coronary artery disease (CAD) (angina, myocardial infarction, or history of coronary artery bypass graft/coronary angioplasty). The presence of diabetic nephropathy (Albustix positive on three consecutive occasions or renal failure) was also recorded. Two questions were included in the demographic questionnaire asking speci®cally about patients' need for and use of a counselling/psychotherapy service. Participants were asked to return their completed questionnaires in a sealed envelope before leaving the clinic. Differences in prevalence rates of anxiety and depression according to demographic and clinical factors were calculated using chi-square analysis and Student's t-test. Differences in age and duration of diabetes were calculated using the median as a cut-off for each type of diabetes.
Results A high response rate (96%) was achieved, with no differences in responders/non-responders according to age, sex or type of diabetes. The demographic details of the study participants are reported in Table 1, and the study comprises a wide range of ages and durations of diabetes. Overall, 28% of study participants reported moderate± severe levels of depression or anxiety, or both. A greater proportion of patients reported these levels of anxiety (25%) than depression (8%). The proportion of patients who reported mild levels of depression or anxiety was similar; 18% reported mild levels of anxiety and 17% reported mild symptoms of depression. Although not statistically signi®cant, women tended to be more likely to report moderate±severe levels of anxiety than men (30% vs. 22% NS), whereas men tended to be more likely to report moderate±severe symptoms of depression compared to women (11% vs. 5% NS). There were no signi®cant differences in the prevalence of psychological symptomatology according to type of diabetes. Older Type 1 diabetic patients (i.e. age above the median 33 years) were signi®cantly more likely to report moderate±
200
Depression and anxiety in a diabetes clinic · C. E. Lloyd, P. H. Dyer & A. H. Barnett
Table 1 Demographic details of the study population
Age (years)* Duration (years)* Male (%) Employed (%) Smokers (%) Poor glycaemic control (HbA1c > 9%) (%) Interested in counselling or psychotherapy (%)
Type 1 diabetes (n = 48)
Type 2 diabetes (n = 61)
34 (18±80) 16 (< 1±50) 52 77 29 16 35
61 (32±90) 8 (< 1±36) 66 34 10 8 31
*Mean (range).
Figure 1 Relationship between psychological symptomatology and glycaemic control in men and women.
severe levels of depression, but not anxiety, compared to younger Type 1 patients (17% vs. 0%; P < 0.05). Age was not signi®cantly associated with the prevalence of anxiety or depression in Type 2 diabetic patients. There was no relationship between the duration of diabetes and the prevalence of symptoms in either Type 1 or Type 2 diabetic patients. Subjects with background or proliferative diabetic retinopathy were signi®cantly more likely to report moderate±severe levels of anxiety compared to subjects without retinopathy (30%, 56% vs. 21%, respectively; P = 0.02). Subjects with CAD were only slightly more likely to report moderate±severe symptoms of depression compared to those without, and the effect did not achieve signi®cance (14% vs. 8% NS). Subjects with neuropathy tended to be more likely to report depression than those without this complication (25% vs. 7%; P = 0.07). There were an insuf®cient number of subjects with nephropathy to perform any meaningful statistical comparisons.
The relationship between psychological symptomatology and glycaemic control in men and women is shown in Fig. 1. Men with moderate±severe anxiety were sightly more likely to be in poor glycaemic control compared to men with mild or no anxiety but this did not achieve signi®cance (33% vs. 12% NS). All (100%) the men with moderate±severe depression had poor glycaemic control, compared with only 15% of males with mild or no symptoms of depression (P = 0.02). In contrast, none of the women with moderate±severe symptoms of either anxiety or depression had poor glycaemic control. All (100%) the Type 1 diabetic patients with moderate±severe depression had poor glycaemic control compared to only 13% of Type 1 diabetic patients with mild or no depression (P = 0.02). This difference was not observed in subjects with Type 2 diabetes. Other possible factors associated with depression and anxiety were examined. Current smokers were signi®ã 2000 British Diabetic Association. Diabetic Medicine, 17, 198±202
Original articles
cantly more likely to report moderate±severe levels of anxiety than non-smokers (45 vs. 20%; P = 0.02). There was no difference in the prevalence of depression according to smoking status. There was no signi®cant relationship between psychological symptomatology and alcohol intake, marital status, number of children in the household, social class or employment status. Linear regression analysis demonstrated that sex, HbA1c, and the interaction between sex and HbA1c, were all signi®cantly associated with depression and anxiety (R2 = 0.16 and.19, respectively). In the model with anxiety as the dependent variable, age was also a signi®cant factor. Participants were asked whether they had ever felt that they would like to receive counselling or psychotherapy to help with their diabetes management, and 25% answered in the af®rmative. One-third (33%) reported that they would be interested in receiving counselling or psychotherapy if it were currently available at the diabetes clinic. Those who reported that they would like to receive psychological support were signi®cantly more likely to report moderate±severe levels of depression and moderate± severe levels of anxiety compared to those not interested in receiving such support (depression: 21% vs. 3%; P = 0.003, anxiety: 53% vs. 13%; P = 0.0001). There were no differences in the proportion of patients who were interested in receiving psychological support according to complication status, glycaemic control level, sex, age or duration of diabetes.
Discussion This study shows that it is possible to measure psychological symptomatology in a busy clinic setting by using a short, easy to complete questionnaire. The acceptability of the questionnaire was demonstrated by the very high response rate and the reported ease with which the clinic staff were able to administer it. The utility of the HADS is further supported by the ®nding that those who reported moderate±severe symptoms were also more likely to be those who asked for psychological help. Although the HADS is not a diagnostic tool, it can be a useful screening device for identifying those individuals in need of psychological care or treatment prior to their consultation [9]. Those in need of care may then be referred appropriately and quickly. The present study suggests that there may be a high proportion of diabetic patients with psychological problems, with more than a quarter of those surveyed reporting moderate±severe symptoms of either anxiety or depression or both. This high prevalence rate has been shown in other diabetic patient groups [1,2,14]. Both men and women reported psychological dif®culties, which is somewhat contrary to earlier studies in this area. This may be caused by the exclusion of somatic items from the HADS, which women may be more likely to report than ã 2000 British Diabetic Association. Diabetic Medicine, 17, 198±202
201
men. However recent research has shown that men are more likely to experience psychological problems than previously thought [15]. Those with psychological problems were also those who had a history of diabetes complications, a ®nding supported by the only other study of the HADS in diabetes that could be found [16]. A signi®cant relationship between psychological symptoms and glycaemic control was observed in the men but not in the women who participated in the study. This ®nding, re¯ected in both univariate and regression analyses, may be associated with differences in coping patterns, both in general and with regard to diabetes self-care, and requires further investigation [17]. The study also included two questions on the perceived need for psychological treatment services. These questions were asked in order to determine whether the patients themselves would like to use such a service, regardless of the perceptions of those in the medical professions. Whilst 25% reported that they had ever felt that they would like to receive help, more patients (33%) reported that they would be interested in receiving psychotherapy or counselling at the time of the study. This latter (higher) proportion may re¯ect the patients starting to consider their psychological status for the ®rst time. The need for psychological support was not reported by any one particular group of patients; they were equally likely to be Type 1 or Type 2 diabetic patients, male or female, or to have good or poor diabetic control. They were more likely, however, to report symptoms of anxiety or depression. The development of our service has been carried out in light of these ®ndings. For example, we have not restricted our referrals to those in poor diabetic control. Whilst this latter group of patients do bene®t from psychological treatment [7], other patients can also bene®t enormously. In the short time that our psychological service has been operating, preliminary audits suggests that such patients commonly report an improved overall quality of life.
References 1 Peyrot M, Rubin R. Levels and risks of depression and anxiety symptomatology among diabetic adults. Diabetes Care 1997; 20: 585±590. 2 Lustman PJ, Clouse RE, Grif®th LS, Carney RM, Freedland KE. Screening for depression in diabetes using the Beck Depression Inventory. Psychosomatic Med 1997; 59: 24±31. 3 Lustman PJ, Grif®th LS, Clouse RE. Recognizing and managing depression in patients with diabetes. In Anderson BJ, Rubin RR eds. Practical Psychology for Diabetes Clinicians. Alexandria: American Diabetes Association, 1996. 4 Lustman PJ, Grif®th LS, Clouse RE. Improvement in depression is associated with improvement in glycemic control. Diabetes 1995; 4: 27A. 5 Beck AT, Garbin MG. Psychometric properties of the Beck Depression Inventory: 25 years of evaluation. Clin Psychol Rev 1988; 8: 77±100. 6 Bradley C. The well-being questionnaire. In Bradley C ed.
202
7
8
9 10
11
Depression and anxiety in a diabetes clinic · C. E. Lloyd, P. H. Dyer & A. H. Barnett
Handbook of Psychology and Diabetes. Chur, Switzerland: Harwood Academic, 1994. Fosbury JA, Boosley CM, Ryle A, Sonksen PH, Judd SL. A trial of cognitive analytic therapy in poorly controlled type 1 patients. Diabetes Care 1997; 20: 959±964. Bradley C, Gamsu DS, for the Psychological Well-being Working Group of the WHO/IDF St Vincent Declaration Action Programme for Diabetes. Guidelines for encouraging psychological wellbeing: report of a Working Group of the World Health Organisation Regional Of®ce for Europe and International Diabetes Federation European Region St Vincent Declaration Action Programme for Diabetes. Diabetic Med 1994; 11: 510±516. Zigmond AS, Snaith RP. The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand 1983; 67: 361±370. Carroll BT, Kathol RG, Noyes R, Wald TG, Clamon GH. Screening for depression and anxiety in cancer patients using the Hospital Anxiety and Depression Scale. General Hospital Psychiatry 1993; 15: 69±74. Tyrer S. Psychiatric assessment of chronic pain. Br J Psychiatry 1992; 160: 733±41.
12 UK Prospective Diabetes Study Group. Intensive blood glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 1998; 352: 837±853. 13 The Diabetes Control and Complications Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. New Engl J Med 1993; 329: 977±986. 14 Roy M, Collier B, Roy A. Excess of depressive symptoms and life events among diabetics. Comp Psychiatry 1994; 35: 129±31. 15 Shajahan PM, Cavanagh JTO. Admission for depression among men in Scotland, 1980±1995: retrospective study. Br Med Journal 1998; 316: 1496±1497. 16 Carrington AL, Mawdsley SK, Moreley M, Kincey J, Boulton AJ. Psychological status of diabetic people with or without lower limb disability. Diabetes Res Clin Prac 1996; 32: 19±25. 17 Spiess K, Sachs G, Moser G, Pietschmann P, Schernthaner G, Prager R. Psychological moderator variables and metabolic control in recent onset type 1 diabetic patients ± a two year longitudinal study. J Psychosomatic Research 1994; 38: 249±258.
ã 2000 British Diabetic Association. Diabetic Medicine, 17, 198±202
