Does fasting (intermittent fasting) help improve your insulin levels and other hormones?
Fasting probably improves your insulin levels, but most studies are small and short-term. Eating early in the day appears to be most effective; be aware of possible muscle loss if you are older or physically vulnerable.
Fasting lowers your fasting insulin levels and blood sugar. This finding appears consistently across multiple randomised studies and systematic reviews of that research. The average study lasts only three months and includes around 38 participants, however, so little is certain about long-term effects.
The most striking result comes from a small study in men with prediabetes: eating early in the day (all meals before 3:00 p.m., within a six-hour window) improved insulin sensitivity and pancreatic function -- and this happened without anyone losing weight. That shows the effect does not run purely through the bathroom scale; the timing of meals counts in its own right. This was a single small study, so treat it as an indication, not as a firm conclusion.
Eating early aligns with your biological circadian rhythm. Your body processes carbohydrates considerably more efficiently in the morning and early afternoon than late in the evening. Concentrating calories and carbohydrates in the first half of the day appears to keep blood sugar more favourable, regardless of exactly what you eat.
One serious caveat: in several studies, fasting is accompanied by a loss of muscle mass. For people who are somewhat older or physically vulnerable, this is a genuine concern and not a minor footnote. Discuss it with your doctor or dietitian if you are in that situation. Concerns about disordered eating behaviour caused by fasting itself do not appear to be supported by the available short-term studies, but little is yet known about what happens over the longer term.
In summary: the evidence for insulin improvement is there, but most studies are small and short in duration. The quality of evidence is, in technical terms, low to moderate. Eating early and within a restricted time window appears to be the most consistent approach, even without strict fasting.
Based on multiple meta-analyses and umbrella reviews of RCTs, supplemented by a small crossover RCT in people with prediabetes. The GRADE quality of most underlying studies is low to very low; one association (weight loss with alternate-day fasting) has high quality evidence. Studies are on average short (median 3 months) and small (median 38 participants).