Title:The Impact of Methodology and Confounding Variables on the Association Between Major Depression and Coronary Heart Disease: Review and Recommendations
VOLUME: 9 ISSUE: 4
Author(s):Nicolas J.C. Stapelberg, David L. Neumann, David H. K. Shum, Harry McConnell and Ian Hamilton-Craig
Affiliation:School of Applied Psychology and Griffith Health Institute, Griffith University, Parklands Drive, Southport Queensland 4215, Australia.
Keywords:Bias, Confounders, Coronary Heart Disease, Major Depressive Disorder, Methodology, Study design.
Abstract:Background: A reciprocal association exists between Major Depressive Disorder (MDD) and Coronary Heart
Disease (CHD). A quantitative evaluation of this association is necessary to identify potential areas of clinical
intervention. However, the association is unclear because of methodological differences and confounders across studies.
This review examines the impact of methodology and confounding variables on the magnitude of the relationship between
MDD and CHD.
Methods: The search terms “major depression AND coronary heart disease” were entered into an electronic multiple
database search engine. Abstracts were screened for relevance and individually selected articles were collated.
Results: Nine methodological issues and three confounders are identified, which have contributed to uncertainty in the
quantitative relationship between MDD and CHD. More quantitative, prospective longitudinal studies are needed, which
use standard definitions for MDD and CHD and define clear outcomes. Studies should clearly establish the temporal
relationship between the onset of depressed mood and one or more adverse cardiac events, should use quantitative
measures of depression which are treated as continuous data, and have frequent measures of mental state over time,
correlated with measures of cardiac health. Study design should avoid confounding by considering demographic factors,
cardiac risk factors and management of MDD in CHD patients.
Conclusions: This review raises the need for a standardised methodology in future research, taking into account the biases
and confounders listed. Adopting a consensus approach to methodology will facilitate the quantitative exploration of the
causal network linking MDD and CHD.
