Body mass index and incident cardiometabolic conditions in relation to obesity-related cancer risk: A population-based cohort study in Catalonia, Spain

Abstract

Background: We investigated the association between body mass index (BMI) and obesity-related cancer risk among individuals with/without incident hypertension (HTN), type 2 diabetes mellitus (T2DM), and cardiovascular disease (CVD) and the joint associations of overweight/obesity (BMI ≥25 kg/m² ) and each cardiometabolic condition with obesity-related cancer risk METHODS: We conducted a population-based cohort (n = 1,774,904 individuals aged ≥40 years and free of cancer and cardiometabolic conditions at baseline) study between 2010 and 2018 with electronic health records from Spain. Our main outcome measures were hazard ratios (HRs) for incident obesity-related cancers and relative excess risk due to interaction (RERI).

Results: A total of 38,082 individuals developed obesity-related cancers after a median of 8 years of follow-up. The positive association between BMI and obesity-related cancer risk was similar among individuals free of cardiometabolic conditions (hazard ratio, HR per 5 kg/m² : 1.08, 95% confidence interval, CI: 1.06-1.10) and with incident HTN (1.05, 1.01-1.08). The association among those with incident T2DM was null (0.98, 0.93-1.03). There was a positive additive interaction between overweight/obesity and CVD (relative excess risk due to interaction [RERI]: 0.19 [0.09, 0.30]), meaning that the combined association was 0.19 more than the sum of the individual associations. In contrast, a RERI of -0.24 (-0.28, -0.20) was observed for the combined association between overweight/obesity and T2DM.

Conclusions: Public health strategies to reduce overweight can help prevent cancer cases among the general population and individuals with incident HTN/CVD. Further, weight-loss interventions seem to lead to a greater cancer risk reduction among individuals with CVD.

Keywords: adiposity; cancer; cardiovascular disease; electronic health record; hypertension; multimorbidity; type 2 diabetes.

FIGURE 1

Flowchart with the inclusion and exclusion criteria of the study participants. History of cancer considers any type of cancer (C00‐C97) except other and unspecified malignant neoplasm of skin (C44). Causes of end‐of‐follow‐up include transferral out of SIDIAP, cancer diagnosis, death, or end‐of‐study period. Individuals with less than 12 months of follow‐up were excluded because the follow‐up of the participants started 1 year after study entry to avoid potential reverse causality (e.g., BMI affected by undiagnosed cancer). The proportion of excluded individuals due to prevalent cancer or the cardiometabolic conditions should not be interpreted as the proportion of individuals with that specific prevalent disease/condition in SIDIAP given the sequence of excluding participants with prevalent conditions and the overlap between individuals with more than one disease/condition. CVD, Cardiovascular disease; HTN, Hypertension; SIDIAP, Information System for Research in Primary Care; T2DM, Type 2 diabetes mellitus.

FIGURE 2

Framework for the definition of the time‐varying variable “cardiometabolic conditions”. Red arrows represent the disease‐trajectories that were the main interest of this study, while gray dashed lines represent the ones that were not investigated given the research question of this study. CVD, cardiovascular disease; HTN, hypertension; T2DM, type 2 diabetes mellitus.

FIGURE 3

Hazard ratios (HR) with 95% confidence intervals (CI) for associations between body mass index (BMI) and the risk of obesity‐related cancers by cardiometabolic conditions. N cases are obesity‐related cancer cases. The model included BMI as a continuous variable with an interaction term with the time‐varying “cardiometabolic conditions” variable and was adjusted by sex, the geographic region of nationality, the MEDEA deprivation index, smoking status, alcohol intake, and stratified by age (5‐year categories). We evaluated the interaction between BMI and the variable of cardiometabolic conditions by comparing the difference in log‐likelihood of models with and without the interaction term (p = 0.007). The p‐values for the interaction between BMI and each cardiometabolic condition (as extracted directly from the model output) were: 0.067 (HTN), 0.001 (T2DM), 0.980 (CVD), 0.034 (HTN & T2DM), 0.373 (HTN & CVD), 0.577 (T2DM & CVD), 0.790 (HTN, T2DM, & CVD). BMI, body mass index; CI, confidence interval; CVD, Cardiovascular disease; HTN, Hypertension; T2DM, Type 2 diabetes mellitus.