乌鸦传媒

BARCELONA -- While the obesity paradox is apparently here to stay, underweight may be underappreciated for risk in percutaneous coronary intervention (PCI), researchers suggested here.

In a national inpatient database, mortality rates and readmissions after PCI were highest in people with a BMI less than 19, following a reverse J-curve when compared to other BMI groups (both P<0.001 for trend), reported Afnan Tariq, MD, of Lenox Hill Hospital in New York City, and colleagues at the (ESC) meeting.

Mortality rates in people who got left heart catheterization were 6.0% among the underweight individuals in the study, and dipped to a low of 1.2% in the obese before rising up to nearly 2% in the morbidly obese.

The comorbidity-adjusted odds of 30-day readmission were 18% elevated for the underweight compared with normal weight individuals, and reduced by 8% to 11% in the other groups.

Length of stay, too, peaked in underweight individuals after left heart catheterization, at an average of 10.5 days compared with about 5.1 days for normal weight individuals. Predictably, hospitalization costs followed the same trajectory of more than $30,000 for underweight versus less than $25,000 among normal weight people.

"All correlate exceedingly well that underweight patients do worst," Tariq told reporters at an ESC press conference. "We really, as clinicians, need to be cognizant that underweight patients can be very sick and they tend to do worse. We need to recognize that when they come into the office ... the typical patient is not always the sickest patient."

The only group with a predominance of females was the underweight group. The study looked at the National Readmission Database, nested within the Health Care Utilization Project that has records on 15 million hospitalizations through the Nationwide Inpatient Sample Database with ICD-9 codes.

The 1,074 underweight patients had a high comorbidity burden on the Elixhauser scale, but Tariq noted that it may not fully account for important factors like frailty.

Confounding was likely a problem, cautioned Josepa Mauri Ferre, MD, of the Institut Català de la Salut in Barcelona, an ESC spokesperson at the press conference.

The ICD-9 codes are probably a problem, without ability to collect information on the indication for angiography, stage of cardiomyopathy, etc., she said.

Tariq agreed that the study couldn't determine causality. But "I think there may be an inherent bias ... We are trained to see an obese, middle-aged male and say 'We have to intervene ... we have to put you on a statin, we have to put you on aspirin.' It's an 'ocular reflex' in the negative direction. We're not traditionally trained to act the same way for coronary disease in a BMI below 19 female," he noted.

"It's dangerous that the message is better to be obese," Ferre remarked.

"No one is saying that," responded Tariq. He opened by noting that "obesity is a risk factor for coronary disease, we know this. It's a risk factor for hypertension, it's a risk factor for diabetes, for metabolic syndrome." However, "once coronary artery disease has developed in an obese patient, they often have better clinical outcomes."

Comment