Countering impaired glucose homeostasis during catch-up growth with essential polyunsaturated fatty acids: is there a major role for improved insulin sensitivity?

Catch-up growth, while aiding recovery from growth perturbations, has been associated with increased risks of type 2 diabetes and cardiovascular disease later in life. In rat models of semistarvation-refeeding, catch-up growth features disproportionately rapid fat recovery (preferential catch-up fat) with early hyperinsulinemia. Prior work showed that on a low-fat diet, despite skeletal muscle insulin resistance, glucose homeostasis can be maintained via increased insulin-stimulated glucose uptake and de novo lipogenesis in adipose tissue, acting as a glucose sink. Conversely, refeeding on a high-fat diet high in saturated and monounsaturated fatty acids (HF-SMFA) blunts adipose tissue insulin sensitivity and lipogenesis, leading to exacerbated hyperinsulinemia and glucose intolerance. Previous studies demonstrated that diets enriched in essential PUFA (linoleic and α-linolenic acids) prevent excessive fat deposition and impaired glucose homeostasis during catch-up growth. The present study tests the hypothesis that a high ePUFA diet improves glucose homeostasis primarily by enhancing insulin sensitivity in skeletal muscle and adipose tissue, assessed using hyperinsulinemic-euglycemic clamps with tissue-specific glucose utilization measurements.

The paper situates its hypothesis within evidence that catch-up growth is linked to later insulin resistance and cardiometabolic risk. In a rat semistarvation-refeeding model, low-fat refeeding maintains normoglycemia by diverting glucose to adipose tissue through increased insulin-stimulated uptake and de novo lipogenesis, despite skeletal muscle insulin resistance. High-fat diets rich in saturated and monounsaturated fat impair this adipose buffering and exacerbate hyperinsulinemia and glucose intolerance. Prior work by the authors showed that isocaloric high-fat diets enriched with essential PUFA (n-6 and/or n-3, notably linoleic and α-linolenic acids) attenuate excessive fat deposition and hyperinsulinemia during catch-up growth. These observations motivated testing whether improved insulin sensitivity in muscle and adipose tissue mediates ePUFA benefits.

• Animals: Male Sprague-Dawley rats (6 weeks), singly housed at 22 ± 1 °C, 12-h light/dark, standard chow (24% protein, 66% carbohydrate, 10% fat by energy) and water ad libitum prior to experiments. Ethical approvals from University of Fribourg and Canton of Geneva.
• Experimental design: Semistarvation for 2 weeks by feeding a fixed ration of 14 g chow/day (~50% of ad libitum intake), inducing growth arrest. After semistarvation, rats (230–250 g) were allocated into three groups (n = 8–9 per group) with matched body weights and refed isocaloric amounts (355 kJ/day/rat) of one of three diets for 1–2 weeks: (1) Low-fat (LF), (2) High-fat diet rich in saturated and monounsaturated fatty acids based on lard (HF-SMFA), (3) High-fat diet rich in essential PUFA (HF-ePUFA) using a 1:1 mix of safflower and linseed oils. High-fat diets provided 58% of energy from fat. All food provided was consumed daily. Four experiments were conducted: I) Energy balance and body composition over 2 weeks; glucose tolerance test (GTT) on days 7–8. II & III) Hyperinsulinemic-euglycemic clamp studies at high and low insulin doses on days 7–8. IV) Adipose tissue harvest after 1 week for de novo lipogenesis enzyme assays.
• Body composition: After sacrifice, carcasses dried at 70 °C to constant weight, homogenized; total fat by Soxhlet extraction; body water by weight loss upon drying; dry lean mass by difference (body weight – fat – water).
• Glucose tolerance test: Food removed at 07:00; after 6–7 h (post-absorptive), tail vein blood collected, then i.p. glucose 2 g/kg. Blood sampling every 30 min for 2 h; plasma separated and stored at −20 °C. Plasma glucose by Beckman Glucose Analyzer; insulin by ELISA (Crystal Chem).
• Hyperinsulinemic-euglycemic clamps: Anesthetized with Nembutal (50 mg/kg i.p.) on day 7–8. Body temperature maintained at 37 °C. Insulin infusion at high dose 18 mU/kg/min (200 mU/mL Actrapid) ensuring suppression of hepatic glucose production; in a separate cohort, low dose 9 mU/kg/min approximating peak insulin after glucose load in this model. Glucose infusion rate (GIR) adjusted to maintain euglycemia; plasma glucose and insulin measured under basal and clamp conditions (glucose oxidase method; enzyme immunoassay). At clamp end, 2-deoxy-D-[1-3H]glucose (30 µCi/rat) bolus via jugular vein; serial arterial samples collected. After 30 min, rats decapitated; tissues rapidly harvested and frozen. 2-deoxyglucose-6-phosphate in tissues quantified to compute tissue glucose utilization index (GUI; ng/min/mg tissue). Blood samples deproteinized for tracer specific activity using ZnSO4 and Ba(OH)2.
• De novo lipogenesis enzyme assays: Fatty acid synthase (FAS) and glucose-6-phosphate dehydrogenase (G6PDH) activities measured in white adipose tissue depots (methods per prior publications; details in SI 2).
• Statistics: All rats included; minimum n ≥ 7 per group based on prior power. Data as mean ± SE. One-way ANOVA across diet groups with Scheffé post hoc tests; p < 0.05 significant. Analyses via STATISTIX v8.0.
• Outcomes: Body weight and composition; GTT glucose and insulin curves and AUCs; clamp parameters (basal and clamp glucose/insulin, GIR); tissue GUI in multiple skeletal muscles and WAT depots; FAS and G6PDH activities.

• Body composition after 2 weeks refeeding: Semistarvation reduced body fat by ~45% without significant lean mass loss. HF-SMFA refeeding led to significantly higher fat gain than LF, resulting in +25% higher fat mass vs LF (p < 0.001), with similar lean mass gain. HF-ePUFA refeeding resulted in less fat gain and greater lean mass vs HF-SMFA, abolishing fat mass differences relative to LF and yielding higher lean mass than both other groups.
• Glucose tolerance (days 7–8): Basal post-absorptive plasma glucose did not differ among groups. After glucose load, plasma glucose response and AUC were higher in HF-SMFA vs LF and HF-ePUFA; LF and HF-ePUFA did not differ. Basal insulin similar (trend lower in HF-ePUFA). Post-load insulin response and AUC were higher in HF-SMFA vs LF and HF-ePUFA; LF and HF-ePUFA did not differ.
• Hyperinsulinemic-euglycemic clamps: Basal plasma glucose similar across diets; basal insulin significantly lower in HF-ePUFA vs HF-SMFA by ~24% (high-dose clamp) and ~18% (low-dose clamp) (p < 0.01 across diets). GIR markedly reduced in both HF groups vs LF (ANOVA p < 0.001). High-dose clamp: GIR 35.4 ± 0.9 (LF) vs 18.4 ± 0.6 (HF-SMFA, −48%) and 18.9 ± 0.6 mg/min/kg (HF-ePUFA, −47%). Low-dose clamp: GIR 31.0 ± 3.6 (LF) vs 21.1 ± 1.2 (HF-SMFA, −32%) and 23.8 ± 2.4 mg/min/kg (HF-ePUFA, −23%). GIR tended to be higher in HF-ePUFA vs HF-SMFA, especially at low dose, but not statistically significant.
• Tissue-specific GUI: Skeletal muscles (six different muscles) showed no increase in insulin-stimulated GUI with HF-ePUFA vs HF-SMFA; overall muscle insulin sensitivity was not improved by HF-ePUFA. White adipose tissues: HF-SMFA markedly reduced GUI vs LF (often by several-fold). HF-ePUFA only modestly attenuated this reduction, with GUI slightly higher than HF-SMFA in some depots, of borderline or non-significant differences.
• Adipose de novo lipogenesis enzymes: Activities of FAS and G6PDH were reduced in WAT of both HF groups vs LF, but the reduction was less pronounced in HF-ePUFA than HF-SMFA (e.g., significant decreases in HF-SMFA relative to LF; smaller decreases or partial preservation with HF-ePUFA).

The study tested whether ePUFA-rich high-fat refeeding improves glucose homeostasis during catch-up growth by enhancing insulin sensitivity in skeletal muscle and adipose tissue. Despite HF-ePUFA preventing the increases in fat mass and the elevated glucose and insulin responses seen with HF-SMFA during a glucose tolerance test, hyperinsulinemic-euglycemic clamp data indicate no improvement in skeletal muscle insulin sensitivity and only marginal improvements in adipose tissue insulin responsiveness. Whole-body insulin sensitivity (GIR) remained substantially lower in both high-fat groups compared to LF, with only a non-significant tendency toward higher GIR in HF-ePUFA vs HF-SMFA, particularly at the lower insulin dose. These findings argue that improved glucose homeostasis with HF-ePUFA is not primarily due to enhanced insulin-mediated glucose uptake. The authors discuss alternative mechanisms: insulin-independent pathways of glucose uptake (e.g., via AMPK, calcium, nitric oxide, reactive oxygen species), potentially modulated by neuroendocrine factors such as adiponectin and leptin, and transcriptional regulation by PPARs influenced by PUFA. Additionally, greater lean mass accrual with HF-ePUFA may increase glucose buffering capacity, contributing to better glucose handling. Thus, ePUFA likely improve glucose homeostasis through insulin-independent glucose disposal and increased protein retention rather than through major enhancements in insulin sensitivity.

During isocaloric refeeding after semistarvation, a high-fat diet rich in essential PUFA prevents excessive fat gain and normalizes glucose and insulin responses to a glucose load compared to a high-fat diet rich in saturated/monounsaturated fats. However, insulin sensitivity assessed in vivo shows no improvement in skeletal muscle and only marginal benefits in adipose tissue, with minimal impact on whole-body insulin sensitivity. The major contributors to improved glucose homeostasis with ePUFA likely include insulin-independent stimulation of glucose uptake in muscle and other tissues and increased lean mass, enhancing glucose buffering. Future research should elucidate the insulin-independent signaling pathways (e.g., AMPK, nitric oxide, ROS), the roles of adipokines (adiponectin, leptin), PPAR-mediated gene regulation, and the specific contributions of ePUFA and their long-chain metabolites to tissue-specific glucose metabolism and lean mass accrual.