This article is only available in the PDF format. Download the PDF to view the article, as well as its associated figures and tables.
On our study of diet and high-density lipoprotein cholesterol subfractions, Dr Passey has focused on the negative low-density lipoprotein (LDL) cholesterol results. We designed our study with β error equal to .20 specifically to detect differences in highdensity lipoprotein cholesterol and cholesterol subfractions. We obtained significant differences for these variables and therefore think that the study groups were of sufficient size to report these positive results appropriately.While change in LDL cholesterol was not part of our primary hypothesis, clearly LDL cholesterol has a greater SE and greater differences would be needed with this sample size to obtain significance. With 80% power we could detect significance with an LDL cholesterol difference of 0.28 mmol/L, below which are levels we consider clinically insignificant. We consider the nonsignificant change in LDL cholesterol of 0.10 mmol/L to be of little clinical significance.We faithfully pointed out not only our P value but
Dreon DM, Vranizan KM, Wood PD. The Effects of Polyunsaturated Fat vs Monounsaturated Fat on Plasma Lipoproteins: The Power of a Study-Reply. JAMA. 1990;264(16):2071. doi:10.1001/jama.1990.03450160039010