No ‘Obesity Paradox’: BMI and Infarct Size Not Linked in STEMI

Clinicians should focus on hypertension, hyperlipidemia, and diabetes rather than on obesity itself, Gregg Stone advises.

No ‘Obesity Paradox’: BMI and Infarct Size Not Linked in STEMI

Contradicting the idea of an “obesity paradox,” where obese patients are thought to have better outcomes after STEMI despite a higher incidence of MI overall, a new pooled analysis of patients undergoing primary PCI shows no correlation between body mass index (BMI) and infarct size, microvascular obstruction, or LVEF.

Over the years, several studies have highlighted how obese patients have lower death and cardiovascular event rates after PCI and even TAVR, but others more recently have cast doubt on the hypothesis that excess weight is cardioprotective.

“It’s time to retire the ‘obesity paradox,’” senior author Gregg W. Stone, MD (Cardiovascular Research Foundation and Icahn School of Medicine at Mount Sinai, New York), told TCTMD. “The obesity paradox is explained by the company that it keeps, and we should not look at obesity per se. We should look at the prevalence of hypertension, hyperlipidemia, diabetes, cigarette smoking, and other parameters that predict a poor prognosis after STEMI.”

For the study, published in the April 27, 2020, issue of JACC: Cardiovascular Interventions, Bahira Shahim, MD, PhD (Cardiovascular Research Foundation, New York, NY), Stone, and colleagues included 2,238 STEMI patients undergoing primary PCI from six randomized controlled trials—29% were considered normal weight (BMI < 25 kg/m2), 45% were overweight (BMI 25 kg/m2 to < 30 kg/m2), and 26% were obese (BMI ≥ 30 kg/m2). All trials used either cardiac magnetic resonance (CMR) or 99mTc sestamibi single-photon emission computed tomography (SPECT) to assess infarct size.

After a median time to infarct assessment of 4 days, total LV mass and, therefore, total infarct mass were greater with increasing BMI. However, there were no associations found between BMI—either as a continuous variable or by weight category—and percent infarct size in either adjusted or unadjusted analyses. BMI also did not affect microvascular obstruction or LVEF.

CMR and SPECT Findings by BMI


Normal Weight



P Value


(n = 584)

(n = 941)

(n = 526)


Total LV Mass, g




< 0.0001

Total Infarct Mass, g





Infarct Size, % LV Mass





Microvascular Obstruction, % LV Mass










Over a median follow-up period of  350 days, BMI also was not associated with the 1-year composite risk of death or heart failure hospitalization (adjusted HR 1.21; 95% CI 0.74-1.71 for overweight vs normal weight patients; P = 0.59; adjusted HR 1.21; 95% CI 0.74-1.97 for obese vs normal weight patients; P = 0.45) or the risks of death or heart failure hospitalization separately.

Other Mechanistic Explanations?

In an accompanying editorial, Sebastian Johannes Reinstadler, MD, PhD, Bernhard Metzler, MD (both Medical University of Innsbruck, Austria), write that the study provides “strong evidence to eliminate any doubts regarding the lack of an association between obesity and infarct size in patients with STEMI treated with primary PCI.

“Another reassuring contribution of this investigation,” they continue, “is the observation that obesity is not related to microvascular obstruction, thus providing further evidence to dispel any association between obesity and the severity of myocardial damage after STEMI. This is of particular relevance because of the recent notion that microvascular obstruction, a marker of severe reperfusion injury following PCI, might be even more important for the individual prognosis than infarct size itself.”

They note that the timing of assessment of infarct size following STEMI is important given the “dynamic nature of infarct shrinkage” over time. While the study had a median time to assessment of 4 days, some infarcts were analyzed over the course of 1 month, and this “should be taken into consideration when interpreting the results,” they say.

The editorialists also note that BMI, though routinely assessed in the clinic, is not a true measure of adiposity compared to something like waist circumference, body roundness index, and body shape index. “Investigations directing their attention at these measures of adiposity would provide even deeper insights into the potential association between obesity and myocardial damage,” they write.

To this point, Stone said, “it really did not seem like there was a relationship across any level of BMI with infarct size or clinical outcomes. We could’ve looked at it, I suppose, by different parameters, but I don't think it would have made a difference. And BMI is widely accepted as an important prognostic metric for the development of heart disease or prognosis after the development of cardiovascular disorders.” Other metrics of adiposity, he added, “are useful when you're looking at metabolic syndrome and other endocrinologic derangements, but I don’t know that they’re necessarily any better when looking at the effect of weight or, if you will, obesity in terms of infarct size.”

Reinstadler and Metzler conclude that the study “clearly shows that overweight and obesity are not associated with infarct size [or] myocardial tissue reperfusion injury in the setting of acute STEMI treated with primary PCI. Future studies should therefore focus on other mechanistic explanations for a better understanding of the relationship between obesity and short- and long-term clinical outcomes in this setting.”

Stone agreed that more studies should follow, but ideally not looking directly at proving the existence of an obesity paradox. “It would be interesting to look at some of these cardioprotective proteins that are produced from adipose tissue that are increased in obese patients,” he said. “Maybe they are having a cardioprotective effect, but that effect is offset by other adverse consequences of obesity. So, I think there’s more science that we can understand, but in terms of ‘Is it a good thing to be obese if you’re going to have a heart attack?’ I think we've answered that question and the answer to that is no.

Note: Stone and several other co-authors are faculty members or employees of the Cardiovascular Research Foundation, the publisher of TCTMD.

  • Shahim, Reinstadler, and Metzler report no relevant conflicts of interest.
  • Stone reports serving as a consultant to Miracor, TherOx, and Abiomed.

We Recommend