The result of the analysis is not entirely surprising, since several earlier studies had shown that obese H1N1 patients were more likely to need inpatient care and intensive care, Louie and colleagues noted, but it had not been possible to show that obesity itself was the risk.
To clarify the issue, Louie and colleagues analyzed outcomes and risk factors for 1,076 California patients who were in the hospital for at least 24 hours. Of those, 375 were younger than 20, and 82 were pregnant and were excluded from the analysis. Of the remaining 619, body mass index data was available for 534 patients.
The researchers found that 51% of those patients were obese, with a body mass index greater than 30, and 19% had a body mass index of 40 or higher.
In total, 92 patients died, and of those 56 (or 61%) were obese, including 28 (or 30%) with a body mass index of 40 or higher. In the latter group, 21 had a body mass index of 45 or greater, Louis and colleagues found.
Multivariate analysis, adjusting for such things as age, obesity category, and comorbid diseases, showed:
- A body mass index of at least 40 was associated with an odds ratio for death of 2.8 (with a 95% confidence interval from 1.4 to 5.9), compared with those of normal weight.
- Being 50 or older was associated with an odds ratio for death of 2.1, with a 95% confidence interval from 1.2 to 3.7.
- A group of miscellaneous immunosuppressive conditions — including congenital immunodeficiency, asplenia, and adrenal disorders but excluding cancer, transplantation, receipt of immunosuppressive drugs, and HIV — was associated with an odds ratio of 3.9, with a 95% confidence interval from 1.6 to 9.5.
- Asthma had a negative association, with an odds ratio of 0.5 and a 95% confidence interval from 0.3 to 0.9.
Noting that most of the extremely obese patients actually had a body mass index of 45 or higher, the researchers redid the analysis with that level as a cut-off. They found that while the other associations did not change greatly, the odds ratio associated with body mass index rose to 4.2, with a 95% confidence interval from 1.9 to 9.4.
Louie and colleagues cautioned that data were extracted from nonstandard medical records, case ascertainment was based on passive reporting by clinicians, and underreporting may have occurred. As well, height and weight data were more likely to be available for patients who died.
There were also significant differences between patients for whom body mass index was available those for whom it was missing, they noted, although the differences were not significant in bivariate analyses and were therefore unlikely to bias the results.