Paleolithic diet fraction in post hoc data analysis of a randomized cross-over study comparing Paleolithic diet with diabetes diet

Abstract

<h2>Summary</h2><h3>Background</h3> Paleolithic Diet Fraction (PDF), calculated as the fraction of intake from Paleolithic food groups divided by the intake from all food groups, was developed as a measure of compliance hitherto lacking in interventional studies on Paleolithic diet. Mean PDFs in our previously published study comparing Paleolithic diet with a Mediterranean-like diet were around 80% and 40% for the two diets, respectively, and higher PDF was associated with healthier levels in several glucometabolic outcome measures. We now present PDF comparing a Paleolithic diet with a diabetes diet. <h3>Methods</h3> 13 adult patients (3 women and 10 men, mean age 64) with type 2 diabetes followed in 2005–07 a Paleolithic diet and a diabetes diet during two consecutive 3-month periods in a randomized order cross-over design study. Mean daily PDF was calculated using four-day weighed food records. Linear mixed models were used to assess associations between PDF and relative change in the outcome measures used in this study and in aggregated data from this study combined with data from the study comparing Paleolithic diet with a Mediterranean-like diet. <h3>Results</h3> Mean PDF for energy was 77% for the Paleolithic diet and 38% for the diabetes diet. The corresponding percentages for PDF for weight were 81% and 54%. A 10% higher PDF for energy was associated with 0.8% lower weight, 0.7% smaller waist circumference, 2.1% lower total cholesterol, 7.6% lower triglyceride, 2.0% lower HbA1C, 9.9% lower leptin and, in the aggregated data, also with 1.6% lower area under the curve (AUC) for glucose during an oral glucose tolerance test (OGTT). Similar results were found for PDF for weight. <h3>Conclusions</h3> A mean PDF of around 80% for a Paleolithic diet in both this study and our previous study comparing the Paleolithic diet with a Mediterranean-like diet indicate a high reproducibility of the PDF. The mean PDF of around 40% for both the diabetes diet and the Mediterranean-like diet indicates their similarity and how far removed both diets are from a Paleolithic diet in terms of food group content. The associations between higher PDF and healthier glucometabolic outcome measures indicate that this removal is unhealthy. The results from this post hoc data analysis should be corroborated in future studies. <h3>Trial registration</h3> ClinicalTrials.gov. NCT00435240. Registered 14 February 2007. Retrospectively registered, https://www.clinicaltrials.gov/ct2/show/NCT00435240?term=tommy+jönsson&cond=paleolithic&draw=2&rank=2.