Background: Measuring and monitoring the true prevalence of risk factors for chronic conditions is essential for evidence-based policy and health service planning. Understanding the prevalence of risk factors for cardiovascular disease (CVD) in Australia relies heavily on self-report measures from surveys, such as the triennial National Health Survey. However, international evidence suggests that self-reported data may substantially underestimate actual risk factor prevalence. This study sought to characterise the extent of misreporting in a large, nationally-representative health survey that included objective measures of clinical risk factors for CVD.
Methods: This study employed a cross-sectional analysis of 7269 adults aged 18 years and over who provided fasting blood samples as part of the 2011-12 Australian Health Survey. Self-reported prevalence of high blood pressure, high cholesterol and diabetes was compared to measured prevalence, and univariate and multivariate logistic regression analyses identified socio-demographic characteristics associated with underreporting for each risk factor.
Results: Approximately 16 % of the total sample underreported high blood pressure (measured to be at high risk but didn't report a diagnosis), 33 % underreported high cholesterol, and 1.3 % underreported diabetes. Among those measured to be at high risk, 68 % did not report a diagnosis for high blood pressure, nor did 89 % of people with high cholesterol and 29 % of people with high fasting plasma glucose. Younger age was associated with underreporting high blood pressure and high cholesterol, while lower area-level disadvantage and higher income were associated with underreporting diabetes.
Conclusions: Underreporting has important implications for CVD risk factor surveillance, policy planning and decisions, and clinical best-practice guidelines. This analysis highlights concerns about the reach of primary prevention efforts in certain groups and implications for patients who may be unaware of their disease risk status.
Keywords: Cardiovascular disease/epidemiology; Diabetes mellitus/epidemiology; Health surveys; Hypercholesterolemia/epidemiology; Hypertension/epidemiology; Logistic models; Multivariate analysis; Odds ratio; Self disclosure.