Age Modifies the Association Between Obesity and Mortality Among Hospitalizations for Severe Sepsis

To the Editor

While obesity is strongly associated with several chronic conditions(1), the role of obesity during acute illness is less clear, with some studies describing an “obesity paradox,” an apparent protective effect of obesity(2). Preliminary studies suggest that obesity may be associated with lower mortality among those with sepsis(3, 4), though some evidence is conflicting(5, 6). While the incidence and case fatality of sepsis is higher among older adults, it is unknown whether obesity-related immune dysregulation and clinical outcomes differs across the age spectrum. The objective was to determine whether the association between obesity and sepsis mortality is modified by age.

METHODS

This was a secondary analysis using data from the 2010 and 2011 California State Inpatient Database (SID), part of the Agency for Healthcare Research and Quality’s Healthcare Utilization Project (HCUP) (Rockville, MD, USA). We included hospital visits for adults aged ≥ 20 years who were hospitalized from the emergency department with an explicit ICD-9 discharge diagnosis of severe sepsis (995.92) or septic shock (785.52).

Obesity was defined by the validated ICD-9-based chronic disease indicator provided by the HCUP. This is a dichotomous (yes/no) variable based on the ICD-9 code for obesity, as actual body mass index (BMI) or other body composition measurements were not available. The primary outcome was in-hospital mortality. Demographic co-variates included age (analyzed in decades), sex, and race/ethnicity. Other factors, coded as dichotomous (yes/no) variables, included skilled nursing facility residence and co-morbidities defined by the validated ICD-9-based chronic disease indicator provided in HCUP for diabetes, hypertension, congestive heart failure, chronic pulmonary disease, cancer and renal failure.

Multivariable logistic regression models were constructed to calculate odds ratios (OR) and 95% confidence intervals (CI) for the adjusted association between obesity and in-hospital mortality. In order to evaluate effect modification by age, we constructed separate adjusted obesity-mortality models stratified by each age category. Analyses were performed using Stata 12.1 (StataCorp LP, College Station, TX).

RESULTS

There were 116,566 visits that met our inclusion criteria, of which 13,991 (12.0%) visits were coded as having obesity. Overall, 30,712 (26.3%) visits resulted in death during the hospital admission. Visits with obesity had substantially lower in-hospital mortality (18.4%), compared to the 27.4% for non-obese visits (absolute difference -9.0%; 95% CI, -9.7 to -8.3).

After adjusting for co-variates, obesity was inversely associated with in-hospital mortality (OR=0.74 [95%CI, 0.71–0.77]). In the adjusted models stratified by age category, the obesity-mortality association differed by age with younger age groups having minimal association (20–29 years (OR=0.96 [95%CI, 0.61–1.51]), 30–39 years (OR=1.09 [95%CI, 0.85–1.42]) and 40–49 years (OR=0.94 [95%CI, 0.80–1.10]) compared to older age groups (50–59 years (OR=0.76 [95%CI, 0.68–0.84], 60–69 years (OR=0.75 [95%CI, 0.68–0.82]), 70–79 years (OR=0.75 [95%CI, 0.68–0.82]), 80–89 years (OR=0.62 [95% CI, 0.55–0.70]), and ≥ 90 years (OR=0.73 [95%CI, 0.51–1.04]) (Figure 1). Similar results were observed when age categories were dichotomized to age <50 years (OR=0.99, [95%CI 0.87, 1.13]) and age ≥ 50 years (OR=0.65, [95%CI 0.62, 0.68], p for interaction <0.001).

Figure 1.

Adjusted associations between obesity and in-hospital mortality by age categories¹

¹Adjusted for sex, race/ethnicity, nursing home residence, and co-morbidities

DISCUSSION

Obesity was inversely associated with in-hospital mortality among severe sepsis hospitalizations. Furthermore, this association was modified by age, such that obesity was associated with lower mortality among older adults but not among younger adults. Our results are consistent with other studies that have reported an inverse association between obesity and inhospital mortality among critically ill patients(3, 4). Studies demonstrating no association between obesity and mortality included younger individuals(5, 6) whereas those focusing on older adults reported an inverse obesity-mortality association in sepsis(4).

It is currently unknown why obesity might protect against mortality in sepsis, particularly in older adults. CD4⁺ T lymphocytes and B lymphocytes are depleted in severe sepsis(7), and aging-related immune dysfunction impairs pathogen clearance(8). Adipose tissue is now believed to have many functions, including immune modulation. Accordingly, obese, older adults may chronically live in a relatively immune-stimulated rather than immune-suppressed state, which might improve pathogen clearance; however, this hypothesis requires further study.

One limitation is that obesity status was based on validated ICD-9-based chronic disease indicator and was not calculated from measured or self-reported height and weight. Visits by some obese individuals, particularly those with a BMI toward the lower limit of obese, may not have been classified as obese according to ICD-9 code and thus would have been misclassified as non-obese. The non-obese group included underweight individuals who, because of frailty and underlying comorbidity, may have been more likely than normal-weight individuals to die from sepsis (thus modestly increasing the overall mortality of the nonobese group). Nonetheless, our study reports the novel finding that age may modify the obesity-mortality association in sepsis. Future studies are needed to elucidate the relationship between obesity and sepsis outcomes and to determine the biological effects of adipose tissue immune function in aging and sepsis.