BMI of 27-28 kg/m2 and little abdominal obesity best for patients with cerebrovascular disease

Background

A high BMI has been associated with higher survival rates in patients with cardiovascular disease (CVD), a finding that is known as the “obesity paradox” [1]. However, BMI cannot distinguish between sources or locations of body mass, whereas the amount of metabolically active abdominal adipose tissue is more accurately reflected by waist circumference and ultrasound measurements of intra-abdominal fat [2,3]. Moreover, BMI has been shown to have a non-linear (U- or J-shaped) relation with mortality, but previous studies assessing clinical outcomes in patients with cerebrovascular disease did not account for the possibility of non-linear trends [4].

In this study, the association between clinical measurements (BMI and waist circumference) and ultrasound visceral adipose tissue thickness (VAT) and subcutaneous adipose tissue thickness (SAT) [5] on risk of vascular events, vascular mortality, malignancy and all-cause mortality was evaluated in patients with clinically stable cerebrovascular disease (CeVD), while considering the possibility of non-linear relations. The source of patient data was the SMART cohort (n=1767 CeVD patients between 2000 and 2015), an ongoing, prospective study, evaluating the management of atherosclerotic disease and cardiovascular (CV) risk factors. CeVD was defined as either a primary referral diagnosis of ischemic stroke, TIA or retinal infarction, or as a medical history of stroke, TIA or carotid artery operation. Median follow-up was 6.8 years. All patients were clinically stable, not being admitted to a hospital at the time of baseline examination, not having a short life-expectancy and with a Rankin Score of ≤3.

The composite outcome of vascular events was fatal or non-fatal ischemic stroke, intracerebral hemorrhage, myocardial infarction (MI) or fatal abdominal aortic aneurysm rupture or fatal congestive heart failure (HF).

Main results

• Overweight individuals (BMI 25.0–29.9) had a lower risk of both incident malignancy (HR: 0.69; 95% CI: 0.49-0.98) and all-cause mortality (HR: 0.75; 95% CI: 0.57-0.98) compared with normal weight individuals (BM: 18.5-24.9). Obese and normal weight individuals did not differ for risk of any outcome.
• No evidence of non-linearity or a relation between BMI and outcomes were found in the unadjusted models. In the fully adjusted model, however, the relations between BMI and vascular events, vascular mortality, incident malignancy and all-cause mortality were non-linear.
• The nadir for vascular events was 27.5 kg/m2 (95% CI: 14.3-36.7), for vascular mortality 27.1 kg/m2 (95% CI: 21.9-29.3), for incident malignancy 28.1 kg/m2 (95% CI: 19.0-38.2) and for all-cause mortality 28.0 kg/m2 (95% CI: 26.3-29.2).
• No evidence of non-linearity in the relation between waist circumference and outcomes were found in the unadjusted models. In the fully adjusted models, no relations were found between waist-circumference and vascular events, vascular mortality or incident malignancy, in either males or females.
• The relation between waist circumference and all-cause mortality was non-linear (males: P = 0.001; females: P=0.02). The nadir for males was 94.8 cm (95% CI: 80.3–100.1) and for females 84.0 cm (95% CI: 18.7–134.8).
• No evidence of non-linearity was found in any relation between VAT% and risk of vascular events, malignancy or vascular- and all-cause mortality. In the fully adjusted models, VAT% had a positive relation with vascular mortality (HR: 1.23; 95% CI: 1.00-1.51) and all-cause mortality (HR: 1.22; 95% CI: 1.05-1.42).

Conclusion

References

Find this article online at International Journal of Obesity