Aims: Cheese may have a different effect on lipids and lipoproteins than expected from the saturated fat content. We performed a systematic review and meta-analysis of randomized controlled trials (RCTs) to examine the effect of cheese consumption on blood lipids and lipoproteins in healthy populations. Methods: A systematic search in MEDLINE, EMBASE, Scopus, Cababstracts, Cochrane Controlled Trials Register, Clinicaltrials.gov was performed to identify RCTs of cheese supplementation in human adults with total cholesterol, LDL-cholesterol, HDL-cholesterol, triglycerides as a primary or secondary outcome (published until September 2013). A quantitative meta-analysis was performed if more than four RCTs with a comparable control treatment were available. Within person-differences of lipids with corresponding standard errors caused by the cheese compared to the control treatment were pooled (random effects model, STATA 11.0). Results: We identified 15 RCTs, published between 1978 and 2012. We pooled four RCTs comparing the effect of cheese intake to butter with a similar fat content on plasma levels of total cholesterol, LDL-cholesterol, HDL-cholesterol and triglycerides. The amount of cheese used in these trials was rather large, ranging between 120 and 205 g/d. This is approximately equivalent to 3 to 5 cheese servings per day. Intake of cheese (weighted mean difference: 142.6 g/d) reduced total cholesterol significantly by -0.27 mmol/l (95% CI: -0.36 to -0.18), LDL-C by -0.21 mmol/l (95% CI: -0.29 to -0.13), and HDL-C by -0.05 (95% CI: -0.08 to -0.02) compared to butter. The pooled effect on triglycerides was 0.004 (95% CI: -0.058 to 0.065). No heterogeneity was observed (all I2=0%). Cheese was also compared with tofu (n=4 RCTs), fat-modified cheese (n=3), CLA-rich cheese (n=3), milk (n=2), fish (n=1), egg white (n=1). Trials that compared cheese with tofu or fat-modified cheese suggest that differential effects of the products can mainly be attributed to the differences in fatty acid content of the diets. Comparisons with CLA-rich cheese were of limited value because those studied the effects of CLA (and not cheese). Too few trials with milk, egg white, and fish were available to draw conclusions. Conclusions: Based on a limited number of trials, cheese appears less hypercholesterolemic than butter with a similar fat content. Differences in plasma lipids based on cheese compared with tofu and fat-modified products are likely to be caused by the different fat content of the total diets.
|Publication status||Published - 2014|