Investigating the association between cardiovascular diseases and cancer in South African adults from the GEMS database
No Thumbnail Available
Date
2021
Authors
Ramadhar, Anishka
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The incidence of non-communicable diseases is rising globally and in South Africa,
the epidemiologic transition shows a shift from communicable diseases to non communicable diseases. Cardiovascular disease and cancer are the NCDs with the
highest burden and prevalence in South Africa, and these conditions share common
risk factors including tobacco smoking, inflammation, type 2 diabetes and obesity.
Extensive data on CVD and cancer exist in South Africa but has not been previously
explored for an association, nor has the association been explored in South African
medically insured patients. In this study we explore the association between CVD
exposure and cancer outcome in South African adult patients insured by the
Government Employees Medical Scheme (GEMS) from 2014 to 2018 inclusive.
Methods
In this retrospective cohort study, routinely collected secondary data received from
GEMS were analysed for an association between CVD and cancer. The prevalence
of CVD, cancer, type 2diabetes, COPD, obesity and the CVD and cancer subtypes
was calculated in this study population. Univariate and multivariate logistic regression
was performed to investigate the association between CVD and cancer and the CVD
subtypes and cancer, odds ratios were analysed to indicate which NCD conditions
may increase the likelihood of cancer development. Stepwise regression was
performed to select variables for a parsimonious model. Time-to-event analysis was
conducted to measure the time to cancer occurrence and cancer incidence in patients
with and without CVD. Final flexible Generalised Linear Models with cubic splines,
were performed to indicate which NCD conditions are risk factors for cancer
development.
Results
The study analysis comprised 1 851 615 medical records representing 722 934 unique
beneficiaries. The mean age of the patients was 52.68 years with majority of the
patients being female (64.48%). In this study population, 9.17% of patients had cancer
and 79.31% of patients had CVD, 23.07% of all patients had T2D, 13.53% of patients
in the study population had both CVD and cancer. The top 6 cancer subtypes in
Anishka Ramadhar
2244064
5
descending order are breast cancer, skin neoplasms, prostate cancer, colon cancer,
cervical cancer, and lung cancer. The top 6 CVD subtypes in descending order are
hypertension, hyperlipidaemia, VET, angina, heart failure and ischaemia. The
univariate analysis indicated that age, gender, COPD, heart failure, ischaemia,
arrhythmia, atherosclerosis and hyperlipidaemia are all statistically significant risk
factors for the development of cancer. The multivariate analysis indicates that age and
COPD are statistically significant risk factors for cancer. The interaction between T2D
and HT, angina, HF, ischaemia, arrhythmia, cardiomyopathy, and atherosclerosis
respectively, increase the risk for cancer development. The interaction between HT
and VET, angina, HF, ischaemia, arrhythmia, cardiomyopathy, and atherosclerosis
respectively, increase the risk for cancer development.
Discussion
Collectively CVD is not a risk factor for the development of cancer, however, in this
study population, CVD subtypes such as hyperlipidaemia and ischaemia increase the
risk for cancer development. Interactions between CVD subtypes with T2D and HT,
increase the risk for cancer. Although COPD only comprised 1.34% of this study
population, regression analysis indicated that COPD is a risk factor for cancer and
may be responsible for an 8-fold increase in the likelihood of developing lung cancer.
Early treatment and controlling the onset and progression of the CVD subtypes and
T2D may delay the onset and progression of cancer. Awareness around the risks and
dangers of tobacco smoking may decrease the incidence and severity of COPD and
lung cancer, as it may be independently associated with both conditions and may be
a confounder. Further data collection and research is required to be conducted on
CVD subtypes and their association with cancer and the common risk factors between
CVD and cancer.
Description
A research report submitted in partial fulfilment of the requirements for the degree of Master of Science in Epidemiology in the field of Epidemiology and Biostatistics to the Faculty of Health Sciences, School of Public Health, University of the Witwatersrand, Johannesburg, 2021