Weight loss in people with Alzheimer's disease: a prospective population based analysis

BMJ 1997; 314 doi: (Published 18 January 1997) Cite this as: BMJ 1997;314:178
  1. Diane Cronin-Stubbs, professor of nursinga,
  2. Laurel A Beckett, associate professor of medicine (statistics)a,
  3. Paul A Scherr, chief of branchb,
  4. Terry S Field, associate epidemiologistc,
  5. Marilyn J Chown, associate epidemiologistc,
  6. David M Pilgrim, neurologistd,
  7. David A Bennett, associate professor of neurologya,
  8. Denis A Evans, professor of medicinea
  1. a Rush Institute on Aging, 1645 West Jackson Boulevard, Suite 675, Chicago, Illinois 60612, USA
  2. b Health Care and Aging Studies Branch, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, Georgia
  3. c Division of Medicine, Harvard Medical School, Boston, Massachusetts
  4. d Harvard Community Health Plan, Harvard Medical School
  1. Correspondence to: Dr Cronin-Stubbs
  • Accepted 2 October 1996


Weight loss among older people, often due to conditions such as cancer and heart disease, has been associated with increased risks of disability and mortality.1 Alzheimer's disease may be a more important source of weight loss than previously recognised, and we examined this issue in a community sample of older adults from a population based, longitudinal study.

Subjects, methods, and results

We collected data during five annual structured clinical evaluations of subjects from a stratified random sample of the population aged 65 years and older of East Boston, Massachusetts, a geographically defined community. We measured weight and height, with participants wearing light clothing, without shoes. Body mass index (weight (kg)/height (m)2) was used to standardise weight for height. Of the 467 clinical evaluation subjects, 338 survived and underwent follow up assessments and 280 (83%) had two or more weight measurements. We diagnosed Alzheimer's disease on the basis of structured neurological examination, neuropsychological performance testing, brief psychiatric evaluation, medical history, and review of drug treatment. We used computerised Medicare hospital discharge records to identify cancer, heart disease, and congestive heart failure, using ICD-9-CM codes.2 We used repeated measures and random effects regression models to examine the relation between Alzheimer's disease and change in weight.

Figure 1 (top) presents a plot of the mean body mass indices of the subjects with Alzheimer's disease and those without, at the five annual assessment points. The figure overestimates the effects of the disease because people with the disease tend to be older. After adjustment for both age and sex, the body mass index in subjects without Alzheimer's disease decreased by an average of 0.14 a year, compared with 0.52 a year in similar subjects with probable Alzheimer's disease, a difference of 0.38 (P<0.01). A trend in the data showed that severity of Alzheimer's disease did not modify weight loss: the body mass index decreased by 0.59 a year in subjects with mild disease and by 0.47 in those with advanced disease. The reduction in body mass index associated with Alzheimer's disease was substantially greater than that associated with heart disease (by 0.19, P=0.05), congestive heart failure (by 0.23, P=0.07); and cancer (by 0.14, P=0.40). Figure 1 (bottom) illustrates the predicted trajectory of weight loss for an unaffected woman of average age (78 years) and body mass index (28.8) at baseline, compared with a similar woman with Alzheimer's disease.


Top: Estimated weight loss (reduction in body mass index) due to Alzheimer's disease in stratified random community sample of residents (unadjusted for age and sex). Bottom: Illustration of predicted trajectory of weight loss for average 78 year old woman with body mass index 28.8 at baseline (adjusted data, from a random effects model)


Alzheimer's disease is recognised as responsible for weight loss both in clinical teaching and in several studies of special groups,3 but the contribution of specific conditions to weight loss has not been examined carefully in population data. Several features of this study strengthen confidence in its results, including the longitudinal population based design, adequate sample size, and uniform diagnostic criteria for Alzheimer's disease. Missing data may have led to an underestimate of the relation between Alzheimer's disease and weight loss because the 187 of 467 participants with missing weights were generally older, thinner, and more likely to have Alzheimer's disease.

The mechanisms linking Alzheimer's disease and weight loss are uncertain,3 but in this study many people with Alzheimer's disease had mild disease, and the magnitude of weight loss in this group did not support the idea that weight loss in Alzheimer's disease is confined to those with severe disease. This differs from the clinical perspective that weight loss occurs with advanced disease and suggests that behavioural problems–such as agitation and pacing–typically associated with more advanced disease,4 do not fully explain weight loss. Pathological changes in the hypothalamus have been described in Alzheimer's disease, and this pathology may affect centres regulating food intake and metabolic processes.5


We thank the residents of East Boston and the staff of the East Boston Neighborhood Health Center for their cooperation and support. This project was presented in part at the annual meeting of the Society for Epidemiologic Research, June 1995.

Funding: This work was supported in part by National Institute on Aging (grant AG-10161 and contracts N01-AG-12106 and N01-AG-02107) and the National Institute of Mental Health (award K07-MH-00953).

Conflict of interest: None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract