The relationship between body mass index (BMI) and all-cause mortality in patients with atrial fibrillation (AFib) is U-shaped, with the risk highest in those who are underweight or severely obese and lowest in patients defined simply as obese, a registry analysis suggests. It also showed a similar relationship between BMI and risk for new or worsening heart failure (HF).
Mortality bottomed out at a BMI of about 30-35 kg/m2, which suggests that mild obesity was protective, compared even with “normal-weight” or “overweight” BMI. Still, mortality went up sharply from there with rising BMI.
But higher BMI, a surrogate for obesity, apparently didn’t worsen outcomes by itself. The risk for death from any cause at higher obesity levels was found to depend a lot on related risk factors and comorbidities when the analysis controlled for conditions such as diabetes and hypertension.
The findings suggest an inverse relationship between BMI and all-cause mortality in AFib only for patients with BMI less than about 30. They therefore argue against any “obesity paradox” in AFib that posits consistently better survival with increasing levels of obesity, say researchers, based on their analysis of patients with new-onset AFib in the GARFIELD-AF registry.
“It’s common practice now for clinicians to discuss weight within a clinic setting when they’re talking to their AFib patients,” observed Christian Fielder Camm, BM, BCh, University of Oxford (England), and Royal Berkshire NHS Foundation Trust, Reading, England. So studies suggesting an inverse association between BMI and AFib-related risk can be a concern.
Such studies “seem to suggest that once you’ve got AFib, maintaining a high or very high BMI may in some way be protective – which is contrary to what would seem to make sense and certainly contrary to what our results have shown,” Dr. Camm told this news organization.
“I think that having further evidence now to suggest, actually, that greater BMI is associated with a greater risk of all-cause mortality and heart failure helps reframe that discussion at the physician-patient interaction level more clearly, and ensures that we’re able to talk to our patients appropriately about risks associated with BMI and atrial fibrillation,” said Dr. Camm, who is lead author on the analysis published in Open Heart.
“Obesity is a cause of most cardiovascular diseases, but [these] data would support that being overweight or having mild obesity does not increase the risk,” observed Carl J. Lavie, MD, of the John Ochsner Heart and Vascular Institute, New Orleans, La., and the Ochsner Clinical School at the University of Queensland, Brisbane, Australia.
“At a BMI of 40, it’s very important for them to lose weight for their long-term prognosis,” Dr. Lavie noted, but “at a BMI of 30, the important thing would be to prevent further weight gain. And if they could keep their BMI of 30, they should have a good prognosis. Their prognosis would be particularly good if they didn’t gain weight and put themselves in a more extreme obesity class that is associated with worse risk.”
Dr. Carl J. Lavie
The current analysis, Dr. Lavie said, “is way better than the AFFIRM study,” which yielded an obesity-paradox report on its patients with AFib about a dozen years ago. “It’s got more data, more numbers, more statistical power,” and breaks BMI into more categories.
That previous analysis based on the influential AFFIRM randomized trial separated its 4,060 patients with AFib into normal (BMI, 18.5-25), overweight (BMI, 25-30), and obese (BMI, > 30) categories, per the convention at the time. It concluded that “obese patients with atrial fibrillation appear to have better long-term outcomes than nonobese patients.”
Bleeding risk on oral anticoagulants
Also noteworthy in the current analysis, variation in BMI didn’t seem to affect mortality or risk for major bleeding or nonhemorrhagic stroke according to choice of oral anticoagulant – whether a new oral anticoagulant (NOAC) or a vitamin K antagonist (VKA).
“We saw that even in the obese and extremely obese group, all-cause mortality was lower in the group taking NOACs, compared with taking warfarin,” Dr. Camm observed, “which goes against the idea that we would need any kind of dose adjustments for increased BMI.”
Indeed, the report notes, use of NOACs, compared with VKA, was associated with a 23% drop in risk for death among patients who were either normal weight or overweight and also in those who were obese or extremely obese.
Those findings “are basically saying that the NOACs look better than warfarin regardless of weight,” agreed Dr. Lavie. “The problem is that the study is not very powered.”
Whereas the benefits of NOACs, compared to VKA, seem similar for patients with a BMI of 30 or 34, compared with a BMI of 23, for example, “none of the studies has many people with 50 BMI.” Many clinicians “feel uncomfortable giving the same dose of NOAC to somebody who has a 60 BMI,” he said. At least with warfarin, “you can check the INR [international normalized ratio].”
The current analysis included 40,482 patients with recently diagnosed AFib and at least one other stroke risk factor from among the registry’s more than 50,000 patients from 35 countries, enrolled from 2010 to 2016. They were followed for 2 years.
The 703 patients with BMI under 18.5 at AFib diagnosis were classified per World Health Organization definitions as underweight; the 13,095 with BMI 18.5-25 as normal weight; the 15,043 with BMI 25-30 as overweight; the 7,560 with BMI 30-35 as obese; and the 4,081 with BMI above 35 as extremely obese. Their ages averaged 71 years, and 55.6% were men.
BMI effects on different outcomes
Relationships between BMI and all-cause mortality and between BMI and new or worsening HF emerged as U-shaped, the risk climbing with both increasing and decreasing BMI. The nadir BMI for risk was about 30 in the case of mortality and about 25 for new or worsening HF.
The all-cause mortality risk rose by 32% for every 5 BMI points lower than a BMI of 30, and by 16% for every 5 BMI points higher than 30, in a partially adjusted analysis. The risk for new or worsening HF rose significantly with increasing but not decreasing BMI, and the reverse was observed for the endpoint of major bleeding.
The effect of BMI on all-cause mortality was “substantially attenuated” when the analysis was further adjusted with “likely mediators of any association between BMI and outcomes,” including hypertension, diabetes, HF, cerebrovascular events, and history of bleeding, Dr. Camm said.
That blunted BMI-mortality relationship, he said, “suggests that a lot of the effect is mediated through relatively traditional risk factors like hypertension and diabetes.”
The 2010 AFFIRM analysis by BMI, Dr. Lavie noted, “didn’t even look at the underweight; they actually threw them out.” Yet, such patients with AFib, who tend to be extremely frail or have chronic diseases or conditions other than the arrhythmia, are common. A take-home of the current study is that “the underweight with atrial fibrillation have a really bad prognosis.”
That message isn’t heard as much, he observed, “but is as important as saying that BMI 30 has the best prognosis. The worst prognosis is with the underweight or the really extreme obese.”
Dr. Camm discloses research funding from the British Heart Foundation. Disclosures for the other authors are in the report. Dr. Lavie has previously disclosed serving as a speaker and consultant for PAI Health and DSM Nutritional Products and is the author of “The Obesity Paradox: When Thinner Means Sicker and Heavier Means Healthier” (Avery, 2014).
A version of this article first appeared on Medscape.com.