Plasma non-esterified fatty acids: why are we not measuring them routinely?

Abstract

Non-esteri¢ed fatty acids (NEFAs), also known as free fatty acids (FFAs), are released from the hydrolysis of triglyceride in adipose tissue. Their concentration in plasma is usually in the range 100 mmol/L--1mmol/L. NEFAs are a normal part of metabolic adaptability during daily life. Their concentrations in plasma fall after eachmeal as insulin suppresses adipose tissue lipolysis, and rise during the night. As this occurs other tissues, notably skeletal muscle, adapt their substrate utilization accordingly. There is an emerging view that the ability of skeletal muscle (and possibly other tissues) to alter its metabolism appropriately for the prevailing substrate supply is associated with insulin sensitivity and good ‘metabolic health’. This capacity has been termed ‘metabolic £exibility’. A quick glance at current literature on insulin resistance, type 2 diabetes or cardiovascular disease (CVD), however, will give quite a diierent impression of the role of NEFAs. They are increasingly portrayed as the villains in a web of metabolic disturbances that link these diierent conditions in the metabolic syndrome. This view of NEFAs as ‘baddies’ could be said to start with the description of the glucose--fatty acid cycle in 1963. Over the subsequent years, elevated NEFA concentrations have been associated with impaired glucose utilization, impaired suppression of glucose production, impaired insulin secretion (although fatty acids are in fact crucial for normal insulin secretion), impaired endothelial function and vascular reactivity, accelerated hepatic very-low-density lipoprotein-triglyceride secretion, activation of cholesteryl-ester transfer protein (contributing to the dyslipidaemia of insulin resistance), and almost everything else that is bad metabolically. It is rather easy, on this basis, to make a convincing story linking elevated NEFA concentrations with most features of the metabolic syndrome, and certainly with increased risk of type 2 diabetes and CVD. NEFA concentrations are typically elevated as adipose tissue accumulates, especially with abdominal obesity. This is a tempting explanation for the many adverse eiects of abdominal obesity. If all this is true, then measurement of NEFAconcentrations should be highly informative about the risk of type 2 diabetes and CVD.Why, then, are plasma NEFA concentrations notmeasuredmore regularly in clinical practice? After all, a simple enzymatic method has been available for their measurement for several decades: we are no longer dependent upon solvent extraction and solvent-phase titration. In fact, the literature on NEFA as a risk factor is far less convincing than one might at ¢rst expect. Elevated NEFAconcentrations have been shown to be a riskmarker for the future development of type 2 diabetes in several prospective studies in various ethnic groups, but they have come out far less strongly, in fact hardly at all, as a marker for future CVD. Only in the QueŁ bec cardiovascular study were elevated NEFA concentrations found to be a risk marker for future ischaemic heart disease. In the Paris prospective study, elevated NEFA concentrations were predictive of later development of hypertension and, unexpectedly, of cancer, but not of CVD mortality. So why is it that elevated NEFA concentrations do not strongly predict CVD, other than that they have not often been measured in prospective studies? The ¢rst reason is statistical. It is similar to the reason that plasma triglyceride concentrations do not independently predict CVD risk in many studies. Plasma triglyceride concentrations are very variable from day to day in any one individual, giving a high imprecision of estimate of ‘usual’ concentration from a single fasting sample. In contrast, HDL-cholesterol concentrations, which are fairly closely inversely related to plasma triglyceride, are far more stable from day to day. Thus in a multivariate analysis, low HDL-cholesterol will appear to be a far stronger risk marker for CVD than triglyceride concentrations. The day-to-day coe⁄cient of variation of fasting plasma triglyceride concentrations is typically 20%. But for NEFA concentrations, the ¢gure is about twice that (43% in the study byWidjaja et al.). Thus, statistically, NEFA concentrations are not good markers for epidemiology. Beyond that, it is in any case unlikely that NEFA concentrations would appear as independent markers of CVD risk. As argued above, elevated NEFA concentrations are likely to be linked to CVD through other mechanisms that are themselves well-recognized risk markers, such as plasma triglyceride and lowHDL-cholesterol, impaired endothelial function, pro-thrombotic