Pathway

Insulin sensitivity

The Linchpin

33 blends on this site modulate it.

You want it high: strong insulin sensitivity is central to weight, energy and longevity.

In depth

Insulin sensitivity — how readily cells respond to insulin, the linchpin of metabolic health.

What drives it

Pooling observational studies, visceral fat mass showed the strongest correlation with HOMA-IR of any fat depot or anthropometric measure (r = 0.570, 95% CI 0.424-0.687), ahead of total fat, BMI and waist circumference, so greater visceral fat tracks with lower insulin sensitivity. doi 10.1038/srep18495 · PMID 26686961

Fasting metabolic hormone increases it.

Kharitonenkov 2005, J Clin Invest 115:1627 — FGF21 improves glucose and insulin; transgenic FGF21 yields a lean, insulin-sensitive phenotype. PMID 15902306. Gain-of-function in vivo: liver-directed FGF21 transgenic mice (ApoE promoter, ~50-fold higher hepatic FGF21 than fasted wild-type) showed significant reductions in serum cholesterol, glucose and insulin — lower insulin at lower glucose is the signature of improved sensitivity rather than suppressed secretion. Independently replicated by a second group generating FGF21 transgenics with the same promoter. Mechanism is partly central, not purely hepatic: beta-Klotho is expressed in discrete hypothalamic areas whose deletion abrogates FGF21's metabolic effects, and central FGF21 requires intact sympathetic tone to act at the periphery. PMID 15902306

Training load increases it.

Regular aerobic and resistance training improve insulin sensitivity and glycemia in type 2 diabetes — meta-analyses show 0.5-0.7% HbA1c reductions, with resistance training giving 10-15% improvements in insulin sensitivity. PMC8802999

Fasting response increases it.

Caloric restriction improves insulin sensitivity — a core, repeatedly reviewed metabolic effect of CR (Calorie restriction, SIRT1 and metabolism, Nat Rev Mol Cell Biol 2005). The 2-year human calerie trial confirmed the underlying program in skeletal muscle: sirtuin signaling, mitochondrial biogenesis, enhanced autophagy and reduced inflammation (doi 10.1111/acel.13963). Intermittent fasting consistently improves insulin sensitivity, lipid profiles and body composition by promoting fat oxidation while maintaining lean mass — the lean-mass clause matters, because it separates fasting from simple energy deficit. Leptin sensitivity improves alongside insulin sensitivity. The reviewers themselves state the limit: most data come from animal studies OR short-term human trials, raising questions about generalisability to long-term clinical outcomes, and the degree to which autophagy induction produces observable improvement in human metabolic health is still not fully understood. doi 10.1111/acel.13963

Ceramide load reduces it.

Holland 2007, Cell Metab 5:167 — the sphingolipid ceramide is the common intermediate linking saturated fat and glucocorticoids to insulin resistance; inhibiting ceramide synthesis restores glucose tolerance. PMID 17339025. Two independent and fully separable mechanisms, named to the residue: ceramides block AKT by (1) PP2A dephosphorylating its activating residues, and (2) PKC-zeta phosphorylating an inactivating T34 residue inside the AKT PH domain, which blocks AKT translocation rather than its phosphorylation. Downstream, GLUT4 translocation fails in muscle and adipose and hepatic gluconeogenesis is inadequately suppressed. Human, with clamp: in visceral and subcutaneous adipose from 439 subjects, CerS6 expression correlated positively with BMI, body fat and hyperglycemia and inversely with insulin sensitivity measured by EUGLYCAEMIC-HYPERINSULINAEMIC clamp — the reference method, not a surrogate index. Three limits that the sources state themselves. (1) much of the mechanism rests on artefact-prone tools: the early cultured-cell work used cell-permeable short-chain ceramides with cytotoxic properties, often at unphysiologic concentrations. (2) the PP2A arm fails its in-vivo test: pharmacological inhibition of PP2A in rats paradoxically worsened hepatic insulin resistance despite activating AKT — the opposite of what the model predicts. (3) which arm dominates depends on the cell: the PKC-zeta route predominates in L6 myotubes and in primary human muscle cells, whereas PP2A predominates in C2C12. And the causal link is explicitly unsettled: an enzyme generating ceramide close enough to the insulin receptor to suppress it directly has not yet been identified. Chain length decides the sign: long-chain C16:0/C18:0 ceramides track myocardial infarction, heart failure and cardiovascular mortality, while very-long-chain C22:0/C24:0 are neutral OR potentially protective. 'Ceramide' is not one thing. PMID 17339025

Calms inflammation reduces it.

Hotamisligil 1993, Science 259:87 — adipose TNF-α (an NF-κB-driven inflammatory cytokine) blocks insulin action; neutralizing TNF-α restores insulin-stimulated glucose uptake. doi:10.1126/science.7678183. The demonstration is loss-of-function and that is its strength: TNF-alpha mRNA was induced in adipose tissue across four different rodent obesity/diabetes models, protein rose locally and systemically, and neutralizing TNF-alpha in obese fa/fa rats caused a significant increase in peripheral glucose uptake in response to insulin. Species — the landmark paper is entirely rodent. The human counterpart is a separate, later study (Hotamisligil 1995, PMID 7738205): adipose TNF-alpha mRNA in 18 control vs 19 obese premenopausal women, and that paper states plainly that the mechanisms linking obesity to insulin resistance and diabetes in humans remain largely unknown. And the review found disconfirming human evidence that must travel with this edge: subcutaneous adipose tissue was shown to release interleukin-6 but not TNF-alpha in vivo in humans (Mohamed-Ali 1997, J Clin Endocrinol Metab 82:4196-4200). If adipose TNF-alpha is not exported systemically in people, its action is LOCAL/PARACRINE and the systemic rodent story does not transfer intact. The edge stands on the rodent loss-of-function; the human systemic route is contested. doi 10.1126/science.7678183. · PMID 7738205

Slows carb absorption increases it.

By decreasing postprandial hyperglycemia and improving insulin sensitivity, acarbose therapy reduces fasting and postprandial serum insulin, fasting plasma glucose and HbA1c (PMID 12086760; doi 10.1016/S0140-6736(02)08905-5; PMID 14703948; doi 10.1002/14651858.CD005061.pub2; PMC8771266). doi 10.1016/S0140-6736(02 · PMID 12086760 · PMC8771266

Appetite & sugar hormone increases it.

Drucker 2006, Cell Metab 3:153 — GLP-1 stimulates glucose-dependent insulin secretion from pancreatic β-cells.

Fat-burning hormone increases it.

Yamauchi 2001, Nat Med 7:941 — adiponectin reverses insulin resistance by lowering muscle and liver triglyceride. doi 10.1038/90984

Blocks new fat cells increases it.

PPARγ–adiponectin review 2005 — PPARγ activation improves insulin sensitivity, largely via adiponectin. PMID 15711577

Calms inflamed fat tissue reduces it.

The two are now separate, and that separation is the point of the node. The anatomy is specific and visible: in lean tissue macrophages are evenly distributed and display limited inflammatory properties; in obese tissue they gather around dying adipocytes and form crown-like structures. Macrophage abundance rises to about 41 percent of the cell population. This is a histological finding you can point at, not an inference from a cytokine level. And the intervention arm is good: a dual CCR2/CCR5 antagonist significantly reduces M1 macrophage infiltration into adipose tissue in high-fat-diet mice, as well as improving adipose tissue inflammation and insulin resistance. Converging: IL-10 overexpressed in lean-mouse macrophages protected adipocytes from TNF-alpha-induced insulin resistance, and mice with impaired M2 polarization show higher inflammation, weight gain and insulin insensitivity. The direction is disputed and the conditional carries both sides. One literature says infiltration of pro-inflammatory macrophages precedes the insulin resistance. Another says insulin resistance may drive adipose tissue inflammation by promoting the recruitment of macrophages — and in one high-fat time course insulin resistance was evident by week 4 while M1 accumulation was only mildly increased by week 10. The proposed cause arriving six weeks after the effect is not a detail. And the human data do not show the central mechanism at all. The M1-to-M2 polarization switch organises this entire field. In human adipose tissue: despite increased macrophage abundance, there was NO evidence of a phenotypic shift with increasing BMI; the M1:M2 ratio was not different between BMI below 30 and above 30, and there was NO correlation between BMI and the ratio. More macrophages, yes. The switch that is supposed to make them harmful, not demonstrated. The authors offer that single-surface-marker characterization may have been insufficient — which is a limitation, not a rescue. The arrow points at insulin sensitivity, which is a polarity-inverted node: drawn inhibitory means inflammation reduces sensitivity. doi 10.3389/fimmu.2023.1153915 · PMC8321823

Supports weight management reduces it.

The authors state the problem plainly: reverse causality has made it difficult to establish the causal directions, and direct evidence of obesity causing systemic insulin resistance in humans is still lacking. So they used genetic instruments, in UK BIOBANK and METSIM, restricted to non-diabetic individuals so that established disease could not explain the result: both prediabetes and systemic insulin resistance ARE caused by obesity, p = 1.2 x 10^-3 and p = 3.1 x 10^-24 respectively. That second p-value is not a marginal finding. And they did not stop at the association — they went to the tissue: bulk and single-nucleus RNA sequencing on frozen human subcutaneous adipose biopsies showed that adipose mitochondrial gene expression and body fat percent ARE independently associated with insulin resistance, and together with adipose cell-type composition these three factors explain 44.39 percent of the variance and predicted it across three independent human cohorts. Here is the same tissue named as the mediator between body fat and systemic insulin resistance, from an entirely separate literature. Note the direction: the insulin pathway here stands for insulin sensitivity, so this link is drawn inhibitory — obesity reduces sensitivity. doi 10.1530/JOE-24-0269 · PMC7515203

Body clock increases it.

Controlled laboratory studies have shown that short-term circadian misalignment and sleep restriction independently impair physiological processes, including insulin sensitivity, energy expenditure, immune function, blood pressure and cardiac modulation by the AUTONOMOUS nervous system. The laboratory design separated them. The mechanism is meal timing and it is measured: feeding rodents only during the rest phase can phase shift peripheral oscillators in the liver, pancreas, kidney and heart while it has NO effect on the phase of the SCN. Meal timing moves the organs and not the master clock — which is precisely how internal desynchrony is manufactured, and why shift workers eating at 3am are not simply tired people. doi 10.1111/jsr.12770 · PMC3704149

Sleep increases it.

Sleep abnormalities were causally linked to impairments in glucose homeostasis, metabolic syndrome and type 2 diabetes mellitus, and neuro-hormonal changes accompanying behavioral sleep deprivation may lead to insulin resistance. Sleep duration is reported to have shortened from 8 to 6.5 hours over recent decades alongside the rise in obesity and diabetes - the reviewers treat that as hypothesis-generating, not proof. doi 10.1016/j.smrv.2011.05.001 · PMC3132857

What it drives

It reduces Type 2 diabetes.

Insulin resistance is the major pathophysiological feature of type 2 diabetes; obesity-driven release of non-esterified fatty acids, glycerol, hormones and pro-inflammatory cytokines from adipose tissue is what produces it. PMID 17167471

It reduces PCOS.

Insulin resistance with compensatory hyperinsulinemia is present in 65-95% of women with PCOS and is considered a major driver of its pathophysiology — it participates directly in the development of hyperandrogenemia and reproductive dysfunction, and insulin-sensitizing treatment is an established therapeutic route. PMC9832677

It reduces Fasting insulin.

As insulin sensitivity falls, the pancreas compensates by secreting more — which is why fasting insulin rises long before fasting glucose does, and why it is the earlier warning (PMC11906131, PMC9255858). Note, and it matters: the arrow is not settled in one direction. Hyperinsulinemia is conventionally read as the compensation for resistance, but evidence shows chronically elevated insulin can itself drive resistance — fasting insulin correlates negatively with insulin-receptor mRNA in human skeletal muscle across >450 samples, and prolonged insulin exposure attenuates AKT and ERK signaling in myotubes. Treat this as a reinforcing loop, not a one-way street. PMC11906131

It increases Reproductive trigger hormone.

Kisspeptin neurons are influenced by leptin and insulin (excitatory) and by ghrelin and adiponectin (inhibitory) — this is the junction where metabolic state reaches the reproductive axis, and it explains why obesity and metabolic disease suppress the axis centrally rather than at the testis (PMC8482951). stated as a regulatory influence on these neurons; the quantitative contribution relative to sex-steroid feedback is not established in the sources read here. PMC8482951

It increases Cardiac output.

In the early stages of diabetic cardiomyopathy, metabolic disturbances such as defective insulin signaling, excessive circulating insulin, impaired glucose uptake, elevations in myocardial fatty acid uptake, and mitochondrial dysfunction promote cardiac morphological and functional changes. The route is specified three ways: hyperglycemia, insulin resistance, and hyperinsulinemia have all been associated with calcium overload; insulin resistance is detrimental to mitochondria via excessive ROS production and altered cardiomyocyte calcium handling; and it can increase FFA flux through CD36. This is heart disease that begins before any coronary narrowing, which is precisely what a map built around atherosclerosis alone would miss. doi 10.1016/j.jacc.2006.07.007 · PMC10932393

It increases Memory & focus.

Brain insulin resistance sits at the CROSSROAD between metabolic and neurodegenerative disease. Insulin acts on the hippocampus, a brain area playing a pivotal role in learning and memory and primarily affected in Alzheimer’s disease. A named molecular step: insulin induces phosphorylation of GSK3-beta on the inhibitory serine 9 residue, thus reducing its enzymatic activity. doi 10.1007/s10522-025-10300-4 · PMC6685093

It increases Brain adaptability.

Molecular evidence links alteration of hippocampal insulin sensitivity with changes of both adult neurogenesis and synaptic plasticity, with crosstalk between insulin and neurotrophin signaling - which is BDNF, already drawn into this pathway. doi 10.1007/s10522-025-10300-4 · PMC6685093

It increases Muscle.

Insulin, another important anabolic hormone, is also involved through a similar pathway. However, myocyte sensitivity to insulin is significantly reduced during aging, which may be related to the deposition of ectopic lipids such as ceramides in muscle, impaired mitochondrial function, and decreased endothelial function. Three named causes of the resistance, and all three are nodes here. doi 10.3390/ijms26146740 · PMC13074595

It increases Kidney.

The reviewers say this has been missed: the impact of insulin resistance and compensatory hyperinsulinemia on kidney function, and their contribution to CKD, have largely been overlooked in people with OR without diabetes. And the kidney is an insulin target organ in its own right: beyond its classical metabolic targets - muscle, liver and adipocytes - insulin has an essential role in kidney metabolism and physiology, regulating podocyte integrity, tubular sodium handling and gluconeogenesis. In resistance these are disrupted, leading to glomerular hyperfiltration, proteinuria, sodium retention, progressive kidney injury. doi 10.1038/s41581-026-01066-0 · PMC11725408

It reduces Clears liver fat.

Named across five reviews of the cardiovascular-renal-hepatic-metabolic syndrome as part of the shared mechanism set: ectopic fat accumulation triggered by metabolic reprogramming, oxidative stress and inflammation induced by energy overload, modified lipids, UREMIC toxins, and senescence, as well as insulin resistance pathways activated by pro-inflammatory factors and lipotoxic products, collectively exacerbate simultaneous hepatic and renal injury. doi 10.1038/s41581-026-01066-0 · PMC11725408

It reduces Attention regulation.

The split of labor is the finding, and it is drug-resolved: DA may be more essential to attaching attention; NE may contribute more to executive function. Methylphenidate blocks DA reuptake; amphetamines block reuptake and increase production of both DA and NE, with the noradrenergic effects contributing to increased energy and enhanced executive function; atomoxetine acts primarily on NE with secondary effects of increasing prefrontal cortical DA. Four drugs, four profiles, one dissociation. And the reviewers refuse the simple story: a hypothesis suggesting too much OR too little of a single neurotransmitter will NO longer suffice. Anything selling a dopamine-deficit account of attention is selling what the field explicitly abandoned. (PMC4033238; doi 10.1016/j.biopsych.2011.03.003; PMC2265273. doi 10.1016/j.biopsych.2011.03.003 · PMC4033238

It increases Sustained focus.

It increases Liver.

Named across six gut-liver axis reviews. The main alteration is dysbiosis - A decrease in beneficial bacteria, an increase in potentially harmful microbes, and a loss of overall microbial diversity - with intestinal barrier integrity determining the levels of microbial metabolites, ENDOTOXINS and ammonia that reach the liver and systemic circulation. doi 10.1038/s41522-022-00352-6 · PMC11265038

It increases Digestion.