From the authors:
We thank our colleagues for their interest in our recent study in the European Respiratory Journal on the risk of pneumonia hospitalisation in obese individuals 1.
I. Schreter and colleagues present interesting data on body mass index (BMI) as a predictor of pneumonia in a cohort of patients hospitalised with influenza. Their findings suggest that increased BMI (reference value not stated) is related to higher odds for the presence of pneumonia among hospitalised influenza patients, after adjustment for age, sex and comorbidity. Pneumonia was present at the time of admission in most influenza patients with pneumonia, suggesting that obesity may have worsened the pre-hospital course of influenza. This association may be biological, i.e. due to impaired immune response, risk of aspiration, or an altered ventilation pattern in obesity. However, it may also be that obese persons are, on average, less likely to seek timely medical advice and antiviral medication treatment. The challenge in such observational epidemiological studies (including our own) is the risk of uncontrolled confounders associated with being obese, such as a less healthy lifestyle, a possible increase in substance abuse, fewer immunisations, and altered healthcare-seeking behaviour.
C.D. Hingston and colleagues suggest two pathophysiological mechanisms that may have increased the risk of pneumonia in our obese study participants: hyperglycaemia leading to bacterial growth in airway surface fluid, and acid-suppressing drug use leading to bacterial growth in the gastric contents that may be aspirated. Unfortunately, there were no available medication data in our dataset to examine the latter mechanism. However, we observed a higher occurrence of gastro-oesophageal reflux diagnoses in severely obese individuals (3% among those with a BMI ≥35 kg·m2) than in normal-weight participants (1% of those with a BMI 22.5–24.9). Concerning hyperglycaemia, we observed that 17% of severely obese versus only 1% of normal-weight participants in our study developed new uncomplicated diabetes during follow-up, and 6% versus 0.3%, respectively, developed complicated diabetes. Development of any comorbidity was a strong predictor of subsequent pneumonia in our cohort (hazard ratio (HR) 4.4 (95% CI 4.0–4.8) for a Charlson comorbidity index score of 1–2, and HR 11.5 (95% CI 9.4–14.2) for a Charlson index score of ≥3), and associations between baseline obesity and pneumonia risk vanished when adjusting for subsequent comorbidity 1. Diabetes is known to increase pneumonia risk by 25–75%, particularly when long-term glycaemic control is poor 2. We therefore find it likely that diabetes mediated some of the effect of obesity in our study, perhaps by the mechanisms suggested by C.D. Hingston and colleagues. With regard to the use of statins, in a previous study we reported on their potential role in decreasing pneumonia severity 3, and we agree that differences in statin and other drug use may have caused some unmeasured confounding in our study.
The issue of obesity and infection in general is complicated. A certain degree of mild obesity or being slightly overweight might even be beneficial to withstanding infection 4 or when admitted with critical disease to the intensive care unit 5. Based on our own results and those of others, including those noted during the recent influenza A (H1N1) pandemic, we believe that severe obesity (perhaps in particular abdominal obesity) increases the risk of severe influenza and pneumonia, as also reflected in the current immunisation guidelines 6. We believe there is less evidence that mild obesity or being slightly overweight predicts severe respiratory tract infections, particularly where it is not associated with obesity-related comorbidities such as cardiovascular disease or diabetes 1.
We wholeheartedly agree with the remarks by I. Schreter and colleagues that, given the globally rising prevalence of obesity, we need to improve our understanding of the biological association between obesity and infections. This task requires joint efforts from epidemiologists and clinical and laboratory researchers within endocrinology, infectious diseases, respiratory medicine and related specialties.
Footnotes
Statement of Interest
None declared.
- ©2011 ERS