A new study published in BMJ Open Diabetes Research & Care landed on a number that is oddly specific: 7 hours and 18 minutes of sleep per night is the sweet spot for insulin sensitivity.
Researchers analysed data from over 23,000 adults across more than a decade and found that sleep duration and insulin resistance follow an inverted U-shaped curve, with that precise point at the top. The fact that science can now quantify this to within minutes says something about how seriously metabolic research is taking sleep.
What the study measured
The metric at the centre of this research is estimated glucose disposal rate, or eGDR — a reliable proxy for insulin sensitivity that measures how efficiently your body clears glucose from the bloodstream. When eGDR falls, you're moving towards insulin resistance, which underlies type 2 diabetes and is a significant driver of weight gain, fatigue, and hormonal imbalance.
Both too little and too much sleep were associated with lower eGDR. Most sleep advice frames the problem as simple deficiency, as though the only question is whether you're getting enough hours. This study suggests the relationship is more nuanced.
Why sleep affects insulin sensitivity
During deep sleep, the body runs a metabolic maintenance cycle. Growth hormone is released, tissue repair begins, and cortisol (the hormone that raises blood glucose in preparation for waking) is kept low. Cutting that cycle short raises cortisol too early, spikes blood glucose without any food arriving, and over time makes cells progressively less responsive to insulin.
Less discussed is what happens at the other end. Extended sleep appears to disrupt circadian signalling, and glucose regulation is partly time-dependent: cells are primed to receive glucose at certain hours, and that timing gets distorted when sleep sprawls significantly beyond its natural window. The researchers also note that poor glycaemic control itself increases the likelihood of disrupted sleep, so cause and effect run in both directions, which makes it genuinely difficult to know where to intervene once the cycle is established.
The catch-up sleep finding
Weekend catch-up sleep has a different effect depending on where your weekday baseline sits. For people regularly undershooting during the week, 1 to 2 hours of extra weekend sleep was associated with better insulin sensitivity. For those already sleeping beyond the optimal threshold, more than 2 hours of catch-up was associated with worse outcomes. The implication is that more hours don't automatically help; what matters is whether you're compensating for a deficit or compounding an existing pattern.
Why the research particularly matters for women in midlife
The eGDR effects were most pronounced in women aged 40 to 59, mapping closely onto perimenopause and early postmenopause. Oestrogen decline disrupts insulin sensitivity directly during this period, and sleep disturbance becomes significantly more common. Hot flushes and night sweats fragment sleep architecture even when total hours look fine on paper, and the body's metabolic response to the night depends on quality and continuity as much as duration.
We’ve seen this in Nico members repeatedly: women in their mid-forties who describe sleeping reasonably well but waking in the night, whose glucose data shows morning variability their diet alone doesn't explain. Sleep quality rarely enters the perimenopause conversation, even though the research increasingly positions it as a primary metabolic lever.
Research is one thing, here’s what our sleep expert Sam Sadighi says
It's encouraging to see sleep and metabolism being studied with this level of precision. The fact that researchers can now point to an optimal window, even down to minutes, reflects how seriously the field is taking this connection.
That said, as with all large-scale studies, it's worth remembering that population-level findings describe averages, not individuals. Over 23,000 participants is impressive, but your sleep and your metabolism are shaped by factors no study can fully capture; your hormonal profile, stress levels, life stage, and even your chronotype (whether you're naturally an early riser or a night owl).
What this research is most useful for is giving us a framework, not a prescription. If you're consistently sleeping well under 7 hours or regularly pushing past 9, it's worth paying attention to. But if you're hitting somewhere in that range and still feeling metabolically off (fatigued, experiencing blood sugar swings, or struggling with weight) the quality of your sleep may matter just as much as the quantity.
My advice: use findings like this as a prompt for self-reflection rather than a target to hit. Track how you feel after different amounts of sleep, notice patterns, and if something feels off, work with a practitioner who can look at the full picture rather than optimising for a single number.
[Follow Sam for good sleep advice]
What to do with a precise number
Treat it as a signal rather than a target. What it contributes is evidence that sleep sits inside a metabolic range, and that drifting beyond its edges has measurable consequences for how the body handles glucose. Most people have a felt sense that a bad night costs them something the next day; the mechanism running underneath that feeling is worth understanding.
Are you paying attention to yours?
---
References
Fan, Z., et al. (2026). Association of weekday sleep duration and estimated glucose disposal rate: the role of weekend catch-up sleep. BMJ Open Diabetes Research & Care, 14(2), e005692. https://doi.org/10.1136/bmjdrc-2025-005692
The full paper is available at: https://drc.bmj.com/content/14/2/e005692