Health, Housing, and the Heart

See p 2629 Mortality rates of homeless adults in the United States and Canada have been shown to be 3 to 5 times higher than those of the general population.2,3 Homeless women in Toronto have a 10-fold risk of death when compared with housed women in that city.4 In our ongoing observational study of a cohort of 119 chronically homeless people living on the streets of Boston, almost one third died during a 5-year period from 2000 through 2004. Heart disease is a leading cause of death in older homeless people 45 to 64 years old, and despite other common causes of death in younger homeless people 25 to 44 years old, heart disease is 3 times more common in this group than in the age-matched general population.5 Risk factors for cardiovascular (CV) disease are potentially treatable targets in the prevention of morbidity and premature death in this high-risk subpopulation. The study by Lee et al shows, however, that CV risk factors and CV disease itself in homeless adults in Toronto are grossly undertreated. The authors report the prevalence and treatment of major CV risk factors in adults living in emergency shelters and compare these to the general population. Most striking are the high prevalence of heavy smoking and the high rates of poorly controlled diabetes and undiagnosed and undertreated hyper- tension and hypercholesterolemia. These health disparities occur in the setting of universal health insurance, suggesting that such coverage is necessary but not sufficient to overcome the barriers to high-quality health care for this vulnerable population. In an effort to estimate the risk of myocardial infarction or coronary death in homeless patients without known CV disease, the authors use Framingham multiple risk factors equations. These equations are limited and use only tradi- tional, major, independent risk factors for CV disease. The authors' estimate of an absolute 10-year risk of myocardial infarction or coronary death in the homeless cohort of 5% should be considered with caution. Relatively small numbers of homeless individuals in each age group with resultant wide confidence intervals render this estimate not significantly different from that observed in men from the Framingham population. This conclusion is inconsistent with earlier evi- dence that heart disease is more common in homeless populations. The risk of CV disease in homeless people may not be entirely the result of traditional major risk factors. Recent cocaine and alcohol use are highly prevalent in the homeless cohort. The use of these substances is a known risk factor for CV disease,6,7 although it is not taken into account in the Framingham equations. Psychosocial issues, including stress, anger, and depression, are other possible novel CV risk factors not considered in the Framingham equations. The risk of CV outcomes in the homeless cohort may well be underestimated. Research is needed to establish CV risk