Insulin resistance and metabolic health

Evidence and limitations

Written for a reader who wants the literature rather than the summary. Where the evidence is thin we say so and give both sides, and where a claim is mechanistic rather than demonstrated we label it as such.

Mechanism

Insulin resistance means peripheral tissues respond less to a given insulin concentration. Beta cells compensate by secreting more, which holds fasting glucose and glycated hemoglobin inside reference ranges for years. During that compensating phase the vascular endothelium — whose signaling is coupled to insulin action in the vessel wall — and the small unmyelinated nerve fibers are already being damaged. That is why a glucose-only panel detects the process at the point where compensation has failed rather than at the point where it began.

What the evidence shows

Optimal metabolic health, defined in the 2009–2016 National Health and Nutrition Examination Survey (n=8,721) as optimal waist circumference (under 102 cm in men, 88 cm in women), fasting glucose under 100 mg/dL and hemoglobin A1c under 5.7%, blood pressure under 120/80, triglycerides under 150 mg/dL and high-density lipoprotein cholesterol at or above 40 mg/dL in men and 50 mg/dL in women, with no medication for any of them, was met by 12.2% (95% CI 10.9 to 13.6) of American adults. Under the older Adult Treatment Panel III cut points the figure was 19.9% (95% CI 18.3 to 21.5). Removing waist circumference from the definition raised it to 17.6%.

Fewer than one third of normal-weight adults met the full definition; the proportion fell to 8.0% among overweight adults and 0.5% among adults with obesity. Body composition rather than weight is what separates those groups usefully: in 6,171 adults with a normal body mass index in an earlier survey, those in the highest sex-specific tertile of body fat had a four-fold higher prevalence of metabolic syndrome (16.6% versus 4.8%), and women in that group had a 2.2-fold adjusted increase in cardiovascular mortality (hazard ratio 2.2, 95% CI 1.03 to 4.67).

On the neuropathic consequences, nerve injury associated with impaired glucose tolerance and metabolic syndrome preferentially affects small fibers and is relatively invisible to nerve conduction studies, which is the reason small-fiber function is measured directly in the same visit.

Limitations of that evidence

• Cross-sectional prevalence describes how common a state is, not what happens to the people in it.
• The criterion drives the number. Two definitions applied to the same survey data produced 12.2% and 19.9%. Any figure of this kind is a statement about a threshold as much as about a population.
• Fasting insulin has no agreed diagnostic threshold and assays are imperfectly standardized between laboratories, so absolute values do not travel well between sites.
• The normal weight obesity analysis used tertile-based fat thresholds specific to that sample, and its mortality confidence interval has a lower bound of 1.03.
• Practice-reported clinic figures are not trial outcomes and describe patients who already sought care.

What remains uncertain

Whether adding fasting insulin to routine screening improves outcomes has not been tested in a randomized trial. The argument for it is mechanistic and temporal: the abnormality precedes the one currently measured, and the vascular and small-fiber consequences are already detectable during the compensating phase.

The genuinely encouraging part is also the least contested. Every component of the metabolic health definition — waist, glucose, blood pressure, triglycerides, high-density lipoprotein cholesterol — is responsive to changes in diet, activity and sleep, and every one of them can be re-measured. Triglycerides and fasting glucose move within weeks to months; waist and vascular measures more slowly; established arterial stiffness slowly if at all. That ordering is itself the argument for measuring early.