36. Krahn, A.D., Manfreda, J., Tate, R.B., Mathewson, F.A.L., Cuddy, T.E.: Evidence that height is an independent risk factor for coronary artery disease (the Manitoba Follow-up Study). Amer J Cardiol 74: 398-399, 1994.
Details of the Manitoba Follow-up Study have been reported previously (2). In summary, 3983 men fit for pilot training in the Royal Canadian Air Force during World War II formed the study population. The mean age at enrollment (July 1, 1948) was 30.8 years (range 18 to 62), with 90% of study members between age 20 and 39 years. Height, rounded to the nearest inch (2.54 cm), body weight, a physical examination, and an electrocardiogram were recorded at entry to the study. There were no height restrictions for entry into the Royal Canadian Air Force; the cohort included pilots and ground crew. Follow-up included annual mail contact, medical examinations including body weight, cardiac examination, an electrocardiogram every 3 to 4 years, and review of hospital and clinic records of clinical events. Height was recorded at baseline, but not during follow-up. Only 7 study members (0.2%) were <20 years of age when their height was measured.
End points included total CAD, premature CAD (age <60 years), and cardiovascular and total mortality. Risk factors for end points were examined in subjects aged 35 and 50 years. These ages were selected because they represent early and middle age prior to the peak incidence of CAD, and a large sample was available for analysis. Of the 3,983 study members, 3,097 (77%) had their 35th birthday (834 enrolled after age 35, 52 dead) and 3,631 (91%) had their 50th birthday (106 enrolled after age 50, 246 dead) while in the study.
Systemic hypertension was defined as >2 blood pressure readings >160/90 mmHg or presence of antihypertensive treatment. CAD was defined as the presence of stable or unstable angina, myocardial infarction, or sudden death. A family history for CAD was considered present when a first-degree relative developed or died of CAD before 60 years.
For each age, the rate of each end point per 1,000 person-years of follow-up was calculated within each height category. The relation between height and other risk factors was examined comparing average height of smokers and nonsmokers, and those with and without a family of premature CAD. Correlation coefficients were calculated to examine the relation of height to other CAD risk factors. The independent effect of height on the rate of each end point was examined using the Cox proportional-hazards model, controlling for the effect of smoking, diabetes, body mass index, hypertension, systolic and diastolic blood pressure, and family history of premature CAD.
Height at entry was normally distributed. Because exact quartiles of the height distribution placed study members of the same height into different groups, we divided the cohort into 4 height categories that approximated quartiles. Height categories were <67 inches, referred to as short (18% of cohort), 68 to 69 inches (27%), 70 to 71 inches (32%), and >72 inches, referred to as tall (23%). Figure 1 illustrates the incidence of CAD, and cardiovascular and total mortality from age 35 and 50 within each height category. Height category was inversely related to CAD incidence (P <0.001). The inverse relation of height category to total and cardiovascular mortality was not significant. When the short and tall height categories at age 35 years were compared, there was a 45% increase in CAD, 18% increase in total mortality. At age 50, these differences were more marked: 54% of rCAD, 21% for cardiovascular mortality, and 23% for total mortality. There was no significant correlation of height with systolic (r=0.03) or diastolic (r=0.01) blood pressure, or body mass index (r=-0.05). Height was similar between smokers and nonsmokers. Men with a family history of premature CAD were 0.3 inches shorter than those without (p=0.001).
Height, as a continuous variable, was an independent risk factor for CAD from age 35 (p=0.003) and 50 (p = 0.001) years, in addition to smoking, blood pressure, family history (all p <0.001), body mass index at age 35 (p=0.01), and diabetes at age 50 (p = <0.001). A 1 inch decrease in height was associated with a 1.05 times increase in risk of subsequent CAD (95% confidence interval, 1.02 to 1.08; p<0.001) at both ages. In multivariate analysis, height was not significantly associated with cardiovascular or total mortality. The rate of CAD events before age 60 years across height categories was similar to that for all subjects with CAD (Figure 1). In multivariate analysis a 1-inch decrease in height increased risk for premature CAD by 1.06 times (p<0.01) from age 35, and 1.05 times (p=0.01) from age 50.