Practice
In the Literature
Table 1: Increase in odds ratio* for myocardial infarct for 1 standard deviation increase
in body mass index, waist circumference and waist-to-hip ratio by ethnic group
Is waist-to-hip ratio a better
Odds ratio (95% confidence interval)
marker of cardiovascular
Ethnic group Body mass index† Waist circumference† Waist-to-hip ratio†
risk than body mass index? Overall 1.10 (1.07–1.13) 1.19 (1.16–1.22) 1.37 (1.34–1.41)
European 1.14 (1.09–1.20) 1.25 (1.19–1.31) 1.44 (1.36–1.51)
Yusuf S, Hawken S, Ôunpuu S, et al.
Chinese 1.19 (1.11–1.27) 1.24 (1.16–1.33) 1.08 (1.03–1.14)
Obesity and the risk of myocardial in-
South Asian 0.99 (0.93–1.05) 1.03 (0.97–1.10) 1.52 (1.41–1.64)
farction in 27 000 participants from 52
Other Asian 1.29 (1.17–1.43) 1.58 (1.41–1.78) 2.60 (2.25–3.01)
countries: a case–control study. Lancet
2005;366(9497):1640-9. Arab 1.00 (0.93–1.07) 1.07 (0.99–1.16) 1.43 (1.31–1.57)
Latin American 1.12 (1.04–1.21) 1.20 (1.11–1.29) 1.43 (1.32–1.56)
Background: It has been unclear which Black African 1.29 (1.10–1.52) 1.57 (1.31–1.88) 1.36 (1.09–1.69)
measure of obesity — body mass index Mixed-race African‡ 1.07 (0.94–1.22) 1.16 (0.99–1.34) 2.25 (1.79–2.84)
(BMI), waist or hip circumference, or Reprinted, with permission, from Yusuf S, et al. Obesity and the risk of myocardial infarction in
waist-to-hip ratio — shows the strongest 27 000 participants from 52 countries: a case–control study. Lancet 2005;366:1640-9.
*Adjusted for age, sex and region.
relation to risk of cardiovascular disease †The standard deviation for body mass index is 4.15, for waist circumference 12.08, and for waist-to-hip
(CVD). This study compared the relation ratio 0.085.
between these measures of obesity and ‡Black and white mixed-race in South Africa.
risk of myocardial infarction.
trast, increasing hip circumference ap- researchers estimate that a waist-to-hip
Methods: The researchers performed a peared to be protective (OR 0.73, 95% ratio cut-off of 0.83 for women and 0.9
case–control study. Consecutive patients CI 0.66–0.80 for lowest v. highest quin- for men would result in a 3-fold increase
presenting with their first myocardial in- tiles). Both waist and hip measurement in population attributable risk for my-
farction within 24 hours of symptom relations with myocardial infarction re- ocardial infarction. This is particularly
onset were enrolled from 52 countries. mained after adjustment for BMI, important in regions such as Asia, which
These patients were matched by sex and height and other risk factors. have not had significant problems with
age to at least one patient with no previ- obesity as measured by BMI but would
ous history of CVD. Discussion: The results of this study sup- have considerably greater cardiovascular
port other research that abdominal obe- risk if waist-to-hip ratio was used.
Results: A total of 27 098 participants sity is a better risk indicator of CVD than
were enrolled. The researchers found that BMI,1,2 although the cross-sectional de- Practice implications: Physicians should
although BMI values were directly related sign does not offer biologic explanations measure waist-to-hip ratio rather than
to risk of myocardial infarction, the rela- for the association.3 BMI to help determine the risk of CVD
tion disappeared after adjustment for Directly measured data for obesity in their patients. Waist circumference is
waist-to-hip ratio and other risk factors. measurements in varied population also useful for predicting risk, especially
Increasing waist-to-hip ratio was groups were previously unavailable, and in certain ethnic populations. The find-
also found to be associated with in- this study provides sound evidence for a ings of this study support reducing ab-
creasing risk of myocardial infarction stronger relation of waist-to-hip ratio dominal obesity, and the authors sug-
(odds ratio [OR] 2.52, 95% confidence with myocardial infarction risk than BMI gest benefits may be derived from
interval [CI] 2.31–2.74 in highest v. across all ethnic groups. It also holds for increasing hip circumference, perhaps
lowest quintile), but, in contrast to men and women, across all ages and in through increased muscle mass.
BMI, the relation remained significant the presence of other metabolic risk fac-
after adjustment for BMI and other risk tors (diabetes, lipoprotein abnormali- Sally Murray
factors and even among those regarded ties), smoking and hypertension, and Editorial Fellow
as being very lean or of normal weight across all levels of BMI. In contrast, BMI CMAJ
(BMI < 25 kg/m2). Furthermore, unlike had a variable relation with myocardial
that for BMI, this association was evi- infarction in different ethnic subgroups
dent across all world regions (Table 1). and no association in patients with hy- REFERENCES
Waist circumference was also more pertension or a raised apolipoprotein 1. Kragelund C, Hassager C, Hildebrandt P, et al. Im-
DOI:10.1503/cmaj.051561
pact of obesity on long-term prognosis following
strongly related to myocardial infarc- B/apolipoprotein A1 ratio. acute myocardial infarct. Int J Cardiol 2005;98:123-31.
tion risk than BMI (OR 1.77, 95% 1.59– Using waist-to-hip ratio rather than 2. Dalton M, Cameron AJ, Zimmet PZ, et al. Waist cir-
cumference, waist-hip ratio and body mass index
1.97 for highest v. lowest quintiles). BMI as a measure of obesity and hence and their correlation with cardiovascular disease
Among Chinese and black African peo- risk for CVD makes a considerable differ- risk factors in Australian adults. J Intern Med
2003;254(6):555-63.
ple it was the strongest predictor of ence to the proportion of people consid- 3. Kragelund C. A farewell to body-mass index.
myocardial infarction (Table 1). In con- ered at risk of myocardial infarction. The Lancet 2005;366:1589-91.
CMAJ • January 31, 2006 • 174(3) | 308
© 2006 CMA Media Inc. or its licensors
