I want to calculate the sample size for a nested case-control study.
The outcome is the incidence of cardiovascular disease. The exposure is the metabolomic profiles measured from frozen serum samples collected ten years ago.
We are using nuclear magnetic resonance to measure the metabolomic profiles, which yields approximately 250 factors (of which about 40 have been clinically validated).
We plan to use the logistic regression to analyze the associations between metabolomic profiles and cardiovascular disease, and adjust for other confundings.
Due to limited budget, I cannot measure a large number of samples. Therefore, I want to know the minimum sample size. However, I am unsure about the appropriate method for calculating sample size with so many exposures.
This study is similar to the research we want to conduct, although we may not be able to measure such a large sample size, and we won’t use 1-SD as advised by Professor Frank.
Lipids, Lipoproteins, and Metabolites and Risk of Myocardial Infarction and Stroke - PMC (nih.gov)
Any suggestions would be appreciated.
Thank you very much.