Why Bad Sleep Makes You Crave Sugar (And Tanks Your Testosterone)

Huberman covered the metabolism angle. Here's the testosterone loop he left out.

What Huberman Got Right, and What He Left Out

Huberman's core claim holds: disrupted sleep measurably shifts your appetite toward sugar through altered metabolic pathways, not weakened willpower. That part is well-established. What the episode doesn't cover is the testosterone loop sitting underneath that mechanism, and for men over 35, that omission matters more than the episode suggests.

Here's what Huberman said: "Getting quality sleep each night helps regulate not only appetite, but also the specific forms of metabolism that drive specific appetites. It's been reported that when people are sleep deprived or the quality of their sleep is disrupted, their appetite for sugary foods increases." Accurate. But appetite regulation is downstream of hormonal regulation, and the hormone most relevant to men in this audience is the one the episode skips entirely.

Leproult & Van Cauter, *JAMA*, 2011 measured testosterone in healthy young men after one week of sleep restricted to five hours per night. Total testosterone dropped 10–15%. These were not middle-aged men with borderline labs, they were young and healthy, and one week was enough to produce a measurable hormonal decline.

The mechanism: sleep is when the brain fires its strongest pulses of luteinizing hormone (LH), the signal that tells the testes to produce testosterone. Compress sleep and LH pulsatility weakens. Weaken LH and testosterone output falls. Lower testosterone then independently worsens insulin sensitivity and glucose metabolism, which amplifies the sugar cravings Huberman describes. The metabolic and hormonal effects are not parallel problems; they compound each other.

For men whose testosterone is already borderline-low, this loop is not theoretical. It's the reason that TRT alone rarely solves the whole picture, and why sleep has to be part of the conversation before any hormone protocol makes sense.

How Each Stage of Sleep Shapes Your Metabolism

Sleep is not a single metabolic state. Slow-wave sleep (SWS) and REM sleep each run distinct physiological programs, and when either is cut short, specific hormonal systems fail, not sleep in general, but the exact systems that govern glucose and appetite.

Slow-Wave Sleep: Where Glucose Regulation Happens

SWS is the deepest, most restorative stage. It is also when your body does the bulk of its glucose regulation work. In a controlled study, Tasali et al., *PNAS*, 2008 selectively suppressed slow-wave sleep in healthy young adults for three consecutive nights, total sleep time stayed the same, only the depth changed. Insulin sensitivity dropped by 25%. You can spend eight hours in bed, feel like you slept fine, and still wake up metabolically closer to a pre-diabetic state if your deep sleep was fragmented.

This matters for men on any kind of hormone protocol. Insulin sensitivity is not a background variable; it directly determines how efficiently your cells respond to testosterone and how your body partitions calories between fat storage and muscle repair.

REM Sleep: Where Appetite Hormones Reset

REM sleep is where ghrelin and leptin, the hormones that drive hunger and signal fullness, get recalibrated for the next day. Spiegel et al., *PLOS Medicine*, 2004 restricted healthy men to four hours of sleep for two nights. Ghrelin rose 28%. Leptin fell 18%. The net effect was not general hunger; subjects specifically shifted toward calorie-dense, high-sugar foods.

Here is the mechanism-to-felt-experience bridge: ghrelin tells your brain you need fast fuel, leptin no longer counters that signal with satiety, and the result is that the 10:00 a.m. craving for a muffin or the post-lunch vending machine pull is not a willpower failure. It is a predictable output of the hormonal state your sleep created.

What neither of these studies measured is what happens to testosterone across the same disrupted nights, and that is where the story gets considerably worse.

The Testosterone Connection Huberman Didn't Mention

Sleep deprivation suppresses luteinizing hormone (LH) pulsatility, which cuts off the primary signal your testes need to produce testosterone. After just one week of five-hour nights, total testosterone in healthy men aged 24–38 drops 10–15%, equivalent to roughly a decade of normal aging compressed into seven days.

Why LH Pulsatility Depends on Sleep Architecture

The hypothalamic-pituitary-gonadal (HPG) axis runs on a schedule. The hypothalamus fires bursts of gonadotropin-releasing hormone, which prompts the pituitary to release LH, which tells the testes to produce testosterone. Most of that LH signaling happens during sleep, concentrated in the slow-wave stages. Axelsson et al., *Journal of Sleep Research*, 2005 found that the majority of daily LH pulses occur during these deeper stages, precisely the stages that fragment first when sleep is cut short or chronically disrupted.

Cut the slow-wave sleep and you cut the LH pulses. Cut the LH pulses and your testes receive a weaker signal to make testosterone. Free testosterone, the fraction your body can actually use at the tissue level, falls first, often before total testosterone moves enough to trigger a flag on a standard lab report.

Put differently: you can lose meaningful testosterone function before your numbers look off on paper, which is one reason the [Foundation panel](/bloodwork) measures free testosterone alongside total.

The 10–15% Drop: What It Feels Like in Practice

Leproult & Van Cauter, *JAMA*, 2011 measured this precisely: five hours of sleep per night for one week produced a 10–15% reduction in daytime testosterone in young, healthy men. For context, Brambilla et al., *Clinical Endocrinology*, 2009 estimated testosterone declines roughly 1–2% per year after age 30. One bad week replicates months of that gradual slide.

For a 44-year-old sitting at 420 ng/dL, a 15% drop puts him at 357. That number is not dramatic enough to prompt most physicians to recommend TRT, but it is enough to widen the gap between sharp and flat, between motivated and indifferent, between recovering from Monday's workout by Wednesday and still feeling it on Friday.

That metabolic drag is where this gets worse, because low testosterone does not just make you feel sluggish in isolation.

The Compounding Loop: Low T, Poor Insulin Sensitivity, and Sugar Cravings

Low testosterone independently worsens how your body handles glucose, it is not just a passenger in this loop, it is an accelerant. A Grossmann, *Clinical Endocrinology*, 2011 meta-analysis found that low testosterone associates with insulin resistance and increased visceral adiposity independent of BMI, meaning the metabolic penalty shows up even in men who are not overweight.

How Low Testosterone Worsens Glucose Metabolism

Pitteloud et al., *Diabetes Care*, 2005 put numbers to this: insulin sensitivity correlated positively with testosterone levels in men, and those in the lowest testosterone quartile showed significantly worse HOMA-IR (homeostatic model assessment of insulin resistance) scores than men in the upper quartiles. HOMA-IR is the ratio your body uses to signal how hard the pancreas is working to keep blood sugar in range, a higher number means your cells are ignoring insulin's signal, which means your body compensates by releasing more insulin, which means more fat storage and more cravings for fast glucose.

Put differently: low testosterone makes your cells less responsive to insulin, so your body demands more sugar to get the same energy output, and the craving you feel mid-afternoon is partly a hormonal problem, not a willpower problem.

The mechanism compounds through visceral fat. Belly fat carries high concentrations of aromatase, the enzyme that converts testosterone to estradiol. More visceral fat means more aromatase activity, which means more testosterone gets converted and less remains available as free testosterone. That conversion also feeds back through the HPG axis, suppressing the LH signal that tells the testes to produce testosterone in the first place. The loop closes on itself.

Why Men with Borderline-Low T Are Especially Vulnerable

Consider what the trajectory looks like across a decade for a man sitting at 320 ng/dL total testosterone versus one optimized to 700 ng/dL:

Men with borderline-low testosterone, the 300s and low 400s, are in the worst position here. They are low enough to face the metabolic penalties but high enough that a standard primary care visit flags their labs as "normal." Their SHBG may be elevated, which further reduces free testosterone and leaves the metabolic picture even worse than total T alone suggests.

This is why a man who cleans up his sleep, cuts sugar, and still feels like he is fighting upstream deserves a closer look at his full hormone panel, not just total testosterone, but free testosterone, SHBG, and fasting glucose together. The $49 Foundation panel at PMM captures that picture, and the Primal Health Playbook that accompanies it uses functional optimal ranges than population-based "normal" cutoffs that include metabolically compromised men.

If the loop is running, poor sleep lowering T, low T worsening insulin sensitivity, worse insulin sensitivity driving cravings that further disrupt sleep, the question is where you intervene first.

How to Tell If This Loop Is Already Running in You

The clearest signal is a cluster of symptoms that show up together: afternoon energy that falls off a cliff, carb and sugar cravings that spike after a short night, gym recovery that takes longer than it used to, and morning erections that are less frequent or gone. Any one of these alone is noise. All four together points at the loop.

The symptom pattern matters because each one maps to a specific biomarker. Afternoon crashes and sugar cravings correlate with impaired insulin sensitivity and elevated fasting glucose. Slower recovery correlates with suppressed free testosterone. Reduced morning erections correlate with low LH output, which is the upstream signal the loop disrupts first. Brain fog sits at the intersection of all three.

Here is what to check on labs:

HOMA-IR requires both fasting glucose and fasting insulin to calculate. The Performance panel at $99 includes HbA1c and the lipid markers that round out the metabolic picture; the Foundation panel at $49 covers the testosterone and LH side of the loop. If you are not sure which applies to you, the quiz takes three minutes and points you toward the right starting point.

You do not need to guess whether the loop is active. The labs tell you.

Once you know your numbers, the intervention question becomes more specific than "sleep better and eat less sugar."

The Sleep Protocol That Actually Moves the Needle on Testosterone

Three variables account for most of the testosterone recovery you can get from sleep: total duration above 7 hours, slow-wave sleep quality, and wake-time consistency. Get those right and the LH pulsatility that drives overnight testosterone synthesis follows. Everything else is secondary.

Anchoring Your Circadian Rhythm for LH Pulse Recovery

Within two weeks of locking in a consistent wake time with morning light exposure, most men notice steadier afternoon energy and faster sleep onset at night. Here's the mechanism: the cortisol awakening response (CAR) peaks 20–30 minutes after waking and anchors the master clock in your hypothalamus. That clock sets the timing of nocturnal LH pulses, the same pulses that drive testicular testosterone synthesis while you sleep. Shift the anchor and the pulses blunt, which means your testes get a weaker signal on the nights that should be their most productive.

The protocol: bright light in your eyes within 30 minutes of waking, a fixed wake time seven days a week, and no overhead lighting in the 90 minutes before bed. Mah et al., *Sleep*, 2011 showed that extending sleep toward 10 hours in athletes produced significant performance improvements, directional evidence that most men are operating well below their sleep ceiling.

Temperature, Alcohol, and the SWS Killers

Your core body temperature needs to drop roughly 1°C to initiate slow-wave sleep. A bedroom between 65 and 68°F (18–20°C) is the evidence-based target; most American bedrooms run 70–72°F, which suppresses SWS through room temperature alone.

Alcohol compounds the problem. Even moderate intake cuts SWS by up to 20% per Ebrahim et al., *Alcoholism: Clinical and Experimental Research*, 2013. Two drinks before bed doesn't just produce a rough morning, it directly reduces the sleep stage where growth hormone and testosterone synthesis peak.

If your testosterone is already borderline low, these gaps are where the protocol starts, and knowing your actual number tells you how much room you have to recover without additional intervention.

When Sleep Fixes Aren't Enough: The PMM Clinical Perspective

For some men, optimizing sleep moves the needle on testosterone, but not far enough. When years of sleep debt have compounded with borderline-low T, the hormonal deficit doesn't fully repay itself through lifestyle changes alone. Labs tell you when you've crossed that threshold.

Consider a case from PMM's practice: a 44-year-old patient presented with persistent afternoon energy crashes and near-daily sugar cravings. His numbers told a clear story, total testosterone 318 ng/dL, SHBG 14 nmol/L, fasting glucose 101 mg/dL, and an average nightly sleep of 5.5 hours. Sleep was the first intervention. Over eight weeks, consistent 7.5-hour nights pushed his total T to approximately 380 ng/dL. His fasting glucose improved slightly. He was still symptomatic, still crashing at 2 p.m., still raiding the pantry.

That's the threshold Dr. Jacob Egbert, PMM's medical director, watches for: "Sleep is the first protocol we address, but for men who've been sleep-deprived for years, the hormonal debt doesn't fully repay itself with sleep alone. The labs tell you when you've crossed that threshold."

At that point, TRT became the appropriate conversation. PMM's standard starting protocol, testosterone cypionate 100 mg/week split into twice-weekly subcutaneous injections, with labs drawn at 6 and 12 weeks, brought his total T to 720 ng/dL at the 12-week mark. Fasting glucose normalized to 88 mg/dL. The afternoon crashes stopped.

The pattern this case illustrates is consistent enough to name directly:

• ▶Sleep alone: adequate if total T is above 400 ng/dL and symptoms are mild
• ▶Sleep + metabolic support: appropriate if T is 350–450 ng/dL with moderate symptoms and no prior treatment
• ▶Sleep + clinical evaluation: warranted if T is below 400 ng/dL, SHBG is low, fasting glucose is trending above 100, and symptoms persist after 6–8 weeks of sleep optimization

If you don't know your number, the $49 Foundation lab panel is where that conversation starts.

MORE FROM THE JOURNAL