Abstract
Anorexia nervosa (AN) is a condition typically associated with adolescence. Increasingly, more cases of AN in the elderly are being reported [1–3]. The elderly population in developed countries is expanding and society may still be adjusting to this phenomenon. Consequently, while the elderly may feel physically active and want to continue their sexuality, society may expect the reverse [4, 5]. It is possible that AN in the elderly may develop as a maladaptive reaction to this, because conflict between biological changes and societal pressures are known to be associated with the onset of AN in adolescents [6].
A distinct diagnostic identity, named anorexia tardive, has been proposed with onset after age 25 [7]. This is more likely to be precipitated by a loss, and to have more pronounced depressive symptoms. However, the literature also points out that this may simply be incidental to the greater age of this group [8]. The exact nature of the precipitants may vary with age: for the adolescent it is more likely to be menarche, while for older age groups a loss is more common.
Making the diagnosis of AN in the elderly could be challenging. This is because it is difficult to detect, remains uncommon, and a key feature for diagnosis, which is disturbance of the hypothalamic-pituitary-gonadal axis (characterized by amenorrhoea in women and decreased sexual interest and libido in men), is difficult to establish in this age group. Current diagnostic criteria (both DSMIV [9] and ICD-10 [10]) require amenorrhoea as a key feature, thereby preventing diagnosis in postmenopausal women. ICD-10 [10] does allow for the diagnosis of ‘atypical anorexia nervosa’ to be made where one of the key features (such as amenorrhoea) is missing. However, as amenorrhoea in the elderly is secondary to physiological changes, we argue that the diagnostic criteria for AN should make allowances for this feature. We describe a case that highlights these issues.
History
Mrs VC, a 77-year-old widow, was admitted to the acute psychiatric ward in March 2000 with a diagnosis of refractory depression.
She gave a history of a happy and uneventful childhood. She was relatively shy and did well academically. Although conscious of her appearance during adolescence, she denied involvement in dieting or weight loss measures. Her body weight during adulthood was reported to be stable at around 65 kg. She was first married at 19 years of age to a man who she described as verbally abusive and alcoholic. They had a daughter together and the relationship ended after two years. Her second marriage was to a man 16 years her senior when she was 26 years of age. They had a son and, despite finding her husband difficult to live with at times, they remained together for 47 years until his death from colon cancer in 1993. Collateral history from her son and daughter-in-law indicated that Mrs VC demonstrated no signs of grieving following the death. She rapidly cleared all of his possessions from the house and set about redecorating it. At this time she weighed 65 kg and she decided to loose 1/2 kg by walking 2–3 km a day and reducing her calorie intake. These small measures soon increased, as she cut all sweet foods, dairy products, red meat and fat from her diet.
About 18 months after her husband's death, she developed generalized abdominal pain as well as pain in the right iliac fossa, lumbar spine and both shoulders. At this time she became quite depressed and her weight dropped to 40 kg. She had multiple medical admissions over the next 2 years but no explanation was found for her physical symptoms. During the peak of her weight loss, her daughter-in-law noticed that the groceries were not being used and prepared meals were often found in the rubbish bin. When observed, she would cut her food into small pieces and chew for an extended period, while fiddling with her knife and fork. After consuming a small amount she would compliment the meal and say she was full.
At this time she stated to her daughter-in-law that she was preparing herself for death. She placed dust-sheets over her furniture, gave away many possessions and retired to her bedroom. She drew the curtains and refused to respond when friends visited her. If family came to visit she would get up from bed, but dress inappropriately in three or four layers of clothing. Although she weighed only 40 kg at the time, she stated that she ‘felt heavier than Marilyn Monroe’. At times, she was discovered in her room exercising heavily and covered in sweat.
Initially, a diagnosis of depression in the context of bereavement was made and she was prescribed courses of moclobemide (600 mg for 12 months), trimipramine (50 mg for 3 days), paroxetine (40 mg for 17 months) with concurrent amitriptyline (20 mg for 11 weeks), citalopram (40 mg for 6 months), doxepin (10 mg for 5 months) and sodium valproate (200 mg for 17 days). Drugs were stopped either due to intolerable side-effects or because of lack of benefit. Early in 1999, she was admitted to hospital with a diagnosis of depression; however, she was discharged with little improvement. Again in March of 2000, she was admitted to the mental health inpatient unit and it was during this admission that this history of a severe eating disorder was elicited. Electroconvulsive therapy (ECT) was commenced in an attempt to treat the depressive aspect of her condition.
There was no past or family history of psychiatric illness. Her medical illnesses included temporal arteritis, angina, peptic ulcer disease and chronic atrial fibrillation. Her daily medications included temazepam 20 mg nocte, doxepin 10 mg, citalopram 40 mg, isosorbide mononitrate 60 mg, prednisone 10 mg, asprin 150 mg, omeprazole 20 mg, calcitriol 0.5 micrograms and fludrocortisone 0.1 mg. On physical examination, she weighed 43.8 kg, and had a body mass index (BMI) of 16.5. Her potassium level was 3.4 mmol L–1, and thyroid function tests were normal. A computed tomograpy (CT) head scan demonstrated mild cerebral atrophy, particularly in the frontal and temporal regions, and mild dilatation of both frontal horns.
On interview, the patient was immaculately presented, with a downcast gaze and psychomotor retardation. Little rapport was established, but she readily admitted to low mood. She denied any suggestions of an eating disorder or an abnormal body image. When the Eating Attitude Test (EAT-26) [11] was administrated, her guarded responses resulted in a normal outcome, contrasting directly with the collateral history. She refused to discuss issues of grief surrounding her husband's death; however, she did describe guilt associated with the burden that she had become for her family. Cognitive testing with the minimental state exam [12] gave a score of 30/30.
The patient underwent a course of nine treatments of ECT. At the time of discharge she had gained 2 kg and was eating full and regular meals. On interview, she was euthymic, with a reactive and cheerful affect.
Discussion
Mrs VC developed a depressive disorder with melancholic features in the context of unresolved bereavement. She also had features suggestive of AN onset at the age of 72. This was demonstrated by her active avoidance of fattening food, accompanied by excessive exercise, resulting in a BMI of 16.5 when she presented to us at the age of 75.
The literature on the aetiological factors for AN in the elderly is scarce. The majority of cases of AN present in adolescence, where predisposing personality traits of perfectionism and low self esteem, in combination with social pressures to retain a thin figure are thought to be important [13]. A maladaptive adjustment process to puberty may then precipitate AN [14]. Crisp suggested that the amenorrhoea and changes in body shape caused by AN are indicative of a regression back into childhood, thus enabling avoidance of the emotional problems that adolescence brings with it [6]. It is thought that obsessive dieting behaviour brings a sense of autonomy at a stage of life when independence is being sought [15].
For Mrs VC, the demise of her husband was a significant threat to many aspects of her life. This included her sexuality, her role as a wife and homemaker, and her individual autonomy. We see AN as a maladaptive adjustment to these changes, which allowed her to regress in much the same way to that seen in the adolescent. It is worth noting in this regard that previous reports of anorexia in the elderly have followed bereavement [1, 16], the illness of a spouse or separation from their children by marriage or other causes [1].
The diagnosis of AN in the elderly is made difficult by an increased reluctance to discuss ‘psychological issues, eating habits or sexual issues’ in this age group, and in particular for those with AN [2]. We found that Mrs VC denied abnormal dieting behaviour and even obtained a normal score on the EAT-26 subscale [11]. From her history alone, the diagnosis of AN would have been missed. It was only with a strong collateral history that the primary diagnosis of AN could be made. It has been reported that physical disease may masquerade as AN in elderly patients [17]. This is important to exclude, but no physical cause for these symptoms was found in the case of Mrs VC.
The association between AN and depression merits further discussion. Nicholson and Ballance state that ‘the consensus is that, in eating disorders, depressive mood is secondary to the psychological and physical symptoms of the primary disorder’ [16]. In our patient, the depressive features appeared significantly later than the AN and responded successfully to ECT, although there is some suggestion that ECT may be an effective treatment for AN as well [3, 17].
This case report highlights that anorexia nervosa can occur in the elderly and that there may be similarities in this group to presentation in the adolescent. We propose that the current diagnostic criteria should be broadened to allow clinicians to make the diagnosis of AN in the elderly. Comorbid depression is common, and should be treated aggressively. It is crucial to obtain collateral history from the family in order to establish the diagnosis of AN in the elderly.