Current issues of ACP Journal Club are published in Annals of Internal Medicine


21 of 27 variables predicted mortality in older adults

ACP J Club. 1998 Jul-Aug;129:23. doi:10.7326/ACPJC-1998-129-1-023

Source Citation

Fried LP, Kronmal RA, Newman AB, et al. Risk factors for 5-year mortality in older adults. The Cardiovascular Health Study. JAMA. 1998 Feb 25;279:585-92. [PubMed ID: 9486752]



What are the predictors of mortality for community-living adults who are ≥ 65 years of age?


5-year population-based study of participants in the Cardiovascular Health Study (CHS) with a mean follow-up of 4.8 years.


4 communities in the United States.


5201 participants (mean age 73 y, 57% women) who were ≥ 65 years of age and who planned to live in the community for ≥ 3 years were identified from a random sample of the Health Care Financing Administration Medicare enrollment lists and stratified by age group (65 to 74 y, 75 to 84 y, and > 85 y). 61% had ≥ 2 chronic diseases. Exclusion criteria included restriction to a wheelchair in the home, inability to attend the clinical examination, or active treatment for cancer.

Assessment of risk factors

Patients had clinical examinations and were interviewed to obtain data on demographics, health status, health habits, physical activity, function, and clinical history.

Main outcome measure

5-year mortality rate.

Main results

By 5 years, 646 deaths (12%) occurred. 21 of 27 variables were associated with mortality in a multivariate regression analysis. Greater mortality was associated with moderate or severe aortic stenosis (P < 0.001); increasing age (P < 0.001); 100% stenosis of the internal carotid artery (P < 0.001); being male (P < 0.001); abnormal ejection fraction (P < 0.001); poor health (self-assessed, P < 0.001); fasting blood glucose level > 7.2 mmol/L (130 mg/dL, P < 0.001); serum creatinine level 80 to 97 µmol/L (0.9 to 1.1 mg/dL) or > 106 µmol/L (1.2 mg/dL, P < 0.005); use of diuretics (P < 0.001); congestive heart failure (P < 0.005); difficulty with ≥ 2 instrumental activities of daily living (IADLs) (P < 0.001); brachial systolic blood pressure > 169 mm Hg (P < 0.001); > 50 pack-years of smoking (P < 0.005); and major electrocardiographic abnormality (P < 0.001). Lower mortality was associated with albumin level > 37 g/L (3.7 dg/L, P < 0.001); forced vital capacity > 3 mL (P < 0.005); high cognitive function (P < 0.005); weight > 70.2 kg for men and > 59.0 kg for women (P < 0.001); > 4100 kJ per week of physical activity (P < 0.005); tibial artery blood pressure > 168 mm Hg (P < 0.05); and annual income ≥ US $50 000 (P < 0.05).


At 5 years, mortality in older adults was associated with 21 variables, including age, sex, income, weight, some serum measures, some noninvasive physiologic measures, congestive heart failure, cognitive function, functional status, and self-assessed health status.

Source of funding: National Heart, Lung, and Blood Institute.

For correspondence: Dr. L.P. Fried, Welch Center for Prevention, Epidemiology and Clinical Research, 2024 East Monument Street, Suite 2-600, Baltimore, MD 21205, USA. FAX 410-614-9625.


The study by Fried and colleagues shows that several independent risk factors—many of which represent subclinical disease or lifestyle factors—act singly and in concert to increase the overall risk for mortality in older persons. Of the 21 risk factors identified (after adjustment), 6 are traditional targets in primary or secondary prevention: weight (i.e., obesity), physical activity, smoking, alcohol consumption, and control of blood pressure and serum lipid levels. The study by Fried and colleagues substantiates the protective effect of physical activity, which may delay or prevent functional dependence (another risk factor for mortality). When middle-aged and older men increase their level of physical fitness, their risk for cardiovascular and all-cause mortality decreases (1). The paradoxic protective effect of greater weight and the lack of association of low-density lipoprotein cholesterol levels with mortality should be interpreted cautiously because they occurred after researchers controlled for diabetes and when the patients had difficulty with IADLs and carotid stenosis (a likely proxy for atherosclerosis). Excessive weight is clinically unacceptable when it contributes to other comorbid conditions. Treatment of hypercholesterolemia is indicated at all ages for patients with known coronary artery disease but remains controversial in healthy elderly patients, especially women. However, in a cohort study of persons who were ≥ 70 years of age, low levels of high-density lipoprotein cholesterol predicted coronary heart disease events and mortality in women as well as men (2).

Even after researchers adjusted for subclinical disease and physiologic measures, self-rated health remained a robust independent predictor of mortality in the Cardiovascular Health Study cohort. Poor self-rated health has been associated with up to a 94-fold greater risk for mortality in older persons with ≤ 1 chronic disease (3). Clinicians should routinely assess the self-rated health of their older patients, and a rating of fair or poor in apparently healthy patients should prompt an extensive review of systems.

Calvin H. Hirsch, MD
University of California, DavisSacramento, California, USA

Calvin H. Hirsch, MD
University of California, Davis
Sacramento, California, USA


1. Blair SN, Kohl HW 3d, Barlow CE, et al. Benzodiazepines and zolpidem for chronic insomnia: a meta-analysis of treatment efficacy. JAMA. 1995;273:1093-8. [PubMed ID: 7707596]

2. Corti MC, Guralnik JM, Salive ME, et al. HDL cholesterol predicts coronary heart disease mortality in older persons. JAMA. 1995;274:539-44. [PubMed ID: 7629981]

3. Schoenfeld DE, Malmrose LC, Blazer DG, Gold DT, Seeman TE. Self-rated health and mortality in the high-functioning elderly—-a closer look at healthy individuals: MacArthur field study of successful aging. J Gerontol. 1994;49:M109-M15. [PubMed ID: 8169332]