Obesity and non-Hodgkin lymphoma survival in an ethnically diverse population: the Multiethnic Cohort study

Abstract

Purpose Obesity increases mortality for several malignancies, but for non-Hodgkin lymphoma (NHL) the association between body mass index (BMI) and survival is unclear. We examined the association of pre-diagnostic BMI with overall and NHL-specific survival in the Multiethnic Cohort (MEC) study of African Americans, Native Hawaiians, Japanese Americans, Latinos, and Caucasians. Methods MEC participants free of NHL at cohort entry and diagnosed with NHL during follow-up were included in the analyses (N=1331). BMI was based on self-reported weight and height at cohort entry and after 6.1 years of cohort entry. Cox proportional hazards regression was used to calculate hazard ratios (HR) and 95% confidence intervals (CI) with BMI as time-varying exposure in relation to all-cause and NHL-specific mortality while adjusting for known confounders. Results The mean age at NHL diagnosis was 70.5 (range 45–89) years. After a mean follow-up of 4.3±3.5 years, 667 deaths including 450 NHL-specific deaths occurred. In multivariable models, obese patients (BMI ≥30.0 kg/m2) had higher all-cause (HR=1.46, 95%CI 1.13–1.87) and NHL-specific (HR=1.77, 95%CI 1.30–2.41) mortality compared to patients with high-normal BMI (22.5–24.9 kg/m2). For overweight patients (BMI=25.0–29.9 kg/m2), the respective HRs were 1.21 (95%CI 0.99–1.49) and 1.36 (95%CI 1.06–1.75). Cases with low-normal BMI (<22.5 kg/m2) experienced a significant 45% higher all-cause and a 40% higher NHL-specific mortality. After stratification by NHL type, the adverse effect of BMI was stronger for chronic lymphocytic leukemia/small lymphocytic lymphoma than for diffuse large B-cell lymphoma and follicular lymphoma. Conclusions Pre-diagnostic BMI may be a suitable prognostic marker for NHL patients.

Body mass index as a prognostic factor in Asian patients treated with chemoimmunotherapy for diffuse large B cell lymphoma, not otherwise specified

“…However, contrary to the results reported by Carson et al [11] showing a linear, proportional relationship between BMI and estimated survival rate, we observed a non-linear trend between BMI and the estimated 3-year survival. Furthermore, in contrast to the previous report [11], we demonstrated that a very high BMI (≥30.0 kg/m 2 ) had an unfavorable prognosis, which is in agreement with the findings of a recent multiethnic cohort study [32]. The prognostic implications of a high BMI for survival of DLBCL patients should now be validated with further studies.…”

Racial differences in three major NHL subtypes: Descriptive epidemiology

“…Another potential risk factor for survival is BMI. The interaction between race and BMI and its effect on NHL survival was discussed in a recent study [37]. There are a few putative risk factors, including UV radiation and dietary and occupational exposures.…”

Excess body weight and colorectal cancer survival: the multiethnic cohort

“…Because the BMI in this cohort was obtained through self-reports, under-reporting, in particular among overweight and obese individuals, may have occurred despite documentation in a small MEC sample that under-reporting was low [25]. However, the fact that previous analyses within the MEC detected associations of self-reported BMI with survival for other cancer sites [26, 27] provides confidence in the accuracy of self-reported weight. Measured vs. self-reported BMI did also not constitute a major determinant of outcomes in previous studies [10, 12].…”

“…The following covariates (Table 1) were included into the models as potential confounders based on previous publications and on survival analyses for other cancers within the MEC [26, 27]: ethnicity, age at CRC diagnosis in 5-year age groups, education, family history of CRC, alcohol use, smoking status and number of pack-years, physical activity, SEER tumor stage, radiation treatment, chemotherapy treatment, and NSAID use, as well as red meat intake and fiber intake as log-transformed (with 1 added to allow for zero intake) continuous variables (g/1000 kcal). For covariates with missing values, i.e., education, smoking status, physical activity, NSAID use, tumor stage, treatment variables, a missing category was created.…”