– Overweight and obese military veterans who experienced an in-hospital stroke had a lower 30-day and 1-year all-cause mortality than did those who were normal weight in a large national study, Lauren Costa reported at the American Heart Association scientific sessions.

Underweight patients had a significantly increased mortality risk, added Ms. Costa of the VA Boston Healthcare System.

It’s yet another instance of what is known as the obesity paradox, which has also been described in patients with heart failure, acute coronary syndrome, MI, chronic obstructive pulmonary disease, and other conditions.

Ms. Costa presented a retrospective study of 26,267 patients in the Veterans Health Administration database who had a first stroke in-hospital during 2002-2012. There were subsequently 14,166 deaths, including 2,473 within the first 30 days and 5,854 in the first year post stroke.

Each patient’s body mass index was calculated based on the average of all BMI measurements obtained 1-24 months prior to the stroke. The analysis of the relationship between BMI and poststroke mortality included extensive statistical adjustment for potential confounders, including age, sex, smoking, cancer, dementia, peripheral artery disease, diabetes, coronary heart disease, atrial fibrillation, chronic kidney disease, use of statins, and antihypertensive therapy.

30-day adjusted all-cause mortality risk by body mass index

Breaking down the study population into eight BMI categories, Ms. Costa found that the adjusted risk of 30-day all-cause mortality post stroke was reduced by 22%-38% in patients in the overweight or obese groupings, compared with the reference population with a normal-weight BMI of 22.5 to less than 25 kg/m2.

One-year, all-cause mortality showed the same pattern of BMI-based significant differences.

Of deaths within 30 days post stroke, 34% were stroke-related. In an analysis restricted to that group, the evidence of an obesity paradox was attenuated. Indeed, the only BMI group with an adjusted 30-day stroke-related mortality significantly different from the normal-weight reference group were patients with Class III obesity, defined as a BMI of 40 or more. Their risk was reduced by 45%.

The obesity paradox remains a controversial issue among epidemiologists. The increased mortality associated with being underweight among patients with diseases where the obesity paradox has been documented is widely thought to be caused by frailty and/or an underlying illness not adjusted for in analyses. But the mechanism for the reduced mortality risk in overweight and obese patients seen in the VA stroke study and other studies remains unknown despite much speculation.

Ms. Costa reported having no financial conflicts regarding her study, which was supported by the Department of Veterans Affairs.

SOURCE: Costa L. Circulation. 2018;138(suppl 1): Abstract 14288.