Abstract
Background Diffuse large B-cell lymphoma (DLBCL) is the most common aggressive lymphoma and has shown a significant global rise in incidence. While disease-related characteristics are well-established risk factors for pathogenesis and survival, recent studies suggest that obesity and limited physical activity are associated with worse overall survival. However, there is limited population-level evidence exploring modifiable, patient-related risk factors for DLBCL incidence and survival. Neighbourhood walkability (a measure of the built environment and proxy for physical activity) has been previously linked to reduced rates of obesity-related cancers and other important health outcomes but has not yet been studied in relation to DLBCL diagnosis and survival.
Objectives The primary objective of this study was to determine the association between neighbourhood walkability and DLBCL using population-based data. The secondary outcome of this study was to evaluate the impact of neighbourhood walkability on overall survival (OS) amongst patients with DLBCL.
Methods This was a population-based study in Ontario, Canada using linked administrative healthcare databases. Adults (≥18 years) with primary or transformed DLBCL who received curative-intent, rituximab-containing chemotherapy from Jan 2005 to Dec 2021 were included. Neighbourhood walkability was assessed using a validated index which includes 4 equally weighted components: residential density, population density, walkable destinations, and street connectivity and was derived from geographic information software (ArcGIS) and categorized into quintiles (Q1: least walkable to Q5: most walkable). A case-control design was used for the primary objective, where DLBCL cases were matched 1:4 to cancer-free controls by age (birth year) and sex, with a 5-year lookback for assessing neighbourhood walkability from the index date (date of first rituximab treatment). Conditional logistic regression was used to assess for the association between neighbourhood walkability and DLBCL diagnosis adjusting for comorbidities and marginalization (Ontario Marginalization Index; OnMarg). The OnMarg consists of 4 dimensions: material resources, racialized and newcomer population, age and labour force, and household and dwellings. A retrospective-cohort design was used for the secondary objective, to compare the 5-year OS among DLBCL cases with neighbourhood walkability determined at the index date. Cox regression explored the association between neighbourhood walkability and OS adjusting for age, sex, income quintile, comorbidities, and Ann Arbor stage.
Results For the primary analysis, 6,445 patients with DLBCL (median age 61 years, 46% female) were included and matched to 25,780 cancer-free controls. On univariate analysis, those residing in neighbourhood walkability quintiles Q1-Q3 had a significantly higher proportion of individuals who developed DLBCL ompared to Q5 (OR 1.12, 95% CI 1.04 – 1.20), however, individuals in Q4 did not (OR 1.05, 95% CI 0.96, 1.14). This association remained significant after adjusting for comorbidities and marginalization (OR 1.09, 95% CI 1.01 – 1.17). The overall effect of neighbourhood walkability on the hazard of developing DLBCL was significant (Type III Wald χ² = 10.39, df = 2, p = 0.005).The same association was seen when neighbourhood walkability was determined 10-years prior to index.
Within our univariate OS analysis, neighbourhood walkability quintiles Q1-3 had an inferior 5-year OS compared with Q5 (HR 1.11, 95% CI 1.00 –1.23) though this was not significant when adjusting for disease and sociodemographic covariates (HR 0.97, 95% CI 0.87 – 1.09), whereas age (HR 1.35, 95% CI 1.31–1.40), individuals from the lowest income quintile (HR 1.39, 95% CI 1.23–1.58), comorbidities (HR 1.29, 95% CI 1.25 – 1.34), and advanced DLBCL stage (HR 2.91, 95% CI 2.37 – 3.58) were independently associated with inferior 5-year OS.
Conclusion In this large population-based study, controlling for other sociodemographic factors, individuals residing in the least walkable neighbourhoods had an increase in DLBCL diagnosis. This suggests a potential association between the built environment, physical activity and DLBCL oncogenesis, however, an association was not observed between neighbourhood walkability and overall survival. Further studies will be directed at investigating associations between neighbourhood walkability and healthcare utilization in patients with DLBCL.
