Effects on Insulin Resistance & Diabetes
Animal Studies
In rats, high SFA feeding with pork, but not with MC SFA promotes insulin resistance, compared with standard chow (higher in carbohydrate content), while both pork and coconut oil induce hepatic steatosis via SREBP-1.[88] Coconut fat feeding in rats has in fact been demonstrated to induce more hepatic steatosis than pork or a low-fat diet, while simultaneously maintaining insulin sensitivity in skeletal muscle and adipose tissue and even improving oxidative capacity and reducing adiposity.[89] Both high sucrose and high SFA (lard) feeding, but not PUFA, have been demonstrated to, besides steatosis, induce ER stress, caspase activity and liver injury.[90]
SFA & Insulin Resistance in Humans
In a human 5-week, randomized controlled crossover study, Summers et al. demonstrated that replacing SFA with PUFA (n-6) improves euglycemic clamp insulin sensitivity and reduces abdominal subcutaneous fat area.[91] We are presently conducting an RCT on men and women with abdominal obesity, which will elucidate effects on hepatic steatosis from a diet high in SFA from mostly butter compared with vegetable PUFA.[201] Earlier, short interventions in healthy women observed no difference between 12:0 and 16:0 consumption on glucose tolerance[92] but impaired glucose metabolism after SFA (mostly butter) intake compared with rapeseed oil.[93] In a randomized, multicenter intervention where a a high-SFA diet was compared with high-MUFA diet for 3 months, insulin sensitivity was impaired after the high-SFA diet, especially in individuals with a below median total fat intake (<37% energy).[94] This was, however, not confirmed in a recent British multicenter trial.[95]
Individual SFA & Insulin Resistance
In a Swedish cohort of healthy 50-year-old men with 20 years follow-up, 14:0 and 16:0 in serum cholesterol esters predicted the development of metabolic syndrome, independently of other metabolic and lifestyle factors.[96] In adipose tissue, two recent cross-sectional studies have demonstrated similar results on associations between individual SFA and insulin sensitivity. In a study of 59 healthy British men and women, adipose 16:0 was inversely (whereas 14:0 and 18:0 were positively) associated with insulin sensitivity by homeostasis model assessment (HOMA).[97] Since this was not reflected in dietary intake for any of these SFA, and 14:0 and 18:0 were inversely associated with adipocyte size, these associations could instead represent de novo lipogenesis in adipose tissue. These results are in line with a study of 795 Swedish men, in which adipose tissue 16:0 was inversely associated with insulin sensitivity by euglycemic clamp and HOMA, but only modestly reflected dietary intake (r = 0.22).[98] Adipose tissue 12:0, 14:0, 17:0 and 18:0 were independently associated with insulin sensitivity. For 14:0, 16:0 and 17:0, these associations were observed in overweight (BMI >25) men only, whereas 12:0 and 18:0 were associated with insulin sensitivity in all individuals. These results derived from adipose tissue could partly reflect endogenous synthesis, although de novo lipogenesis rate should be low in this population with a high fat intake. The separate patterns for different SFA are clear-cut (Figure 4) and although no association was observed for the dairy fat marker 15:0, the consistent positive associations with insulin sensitivity for 17:0, 14:0 and 12:0 are interesting and require further study with regard to endogenous versus dietary sources including dairy fat intake.[98]
Figure 4.
Crude association between insulin sensitivity determined by euglycemic clamp and proportions of individual saturated fatty acids in subcutaneous adipose tissue in 71-year-old men (n = 795).
M: Euglycemic clamp.
Data taken from [98].
Individual SFA & Diabetes
Regarding the effects from separate SFA, biomarker studies could provide valuable evidence, especially considering concerns with under-reporting of fat intake in obese individuals.[99] In human skeletal muscle phospholipids, SFA[100,101] and especially 16:0[102] have been negatively associated with insulin sensitivity and Type 2 diabetes,[103] which could partly reflect dietary intake. In a cohort of 50-year-old men, followed for 10 years, the risk of developing Type 2 diabetes was associated with serum cholesterol ester levels of 14:0 and 16:0, but not significantly with 18:0.[104] In a Finnish cohort study of 4 years follow-up, impaired fasting glucose and Type 2 diabetes incidence were associated with serum nonesterified 16:0 levels, but were not associated with baseline dietary 16:0 intake assessed from dietary records.[105] In the American Atherosclerosis Risk in Communities (ARIC) study, 2909 middle-aged men and women were followed for 9 years. The incidence of Type 2 diabetes was associated with total SFA levels of plasma cholesterol esters (also observed for 16:0 independently) and phospolipids (also for 16:0 and 18:0).[106] In a more recent 4-year case–cohort study from Australia, dietary intake of 16:0 and 18:0 assessed by a FFQ at baseline nonsignificantly predicted diabetes incidence, whereas dietary 15:0 was inversely associated with diabetes.[107] In baseline plasma phospholipids, total SFA and 18:0 were positively associated with diabetes risk, 16:0 was nonsignificantly associated, whereas 15:0 was negatively associated.[107] For 16:0 and 18:0, this could reflect dietary intake and/or metabolism. For 15:0, these results indicate an inverse association between diabetes incidence and dairy intake.
Clin Lipidology. 2011;6(2):209-223. © 2011 Future Medicine Ltd.
