THE WAKE-UP CALL
Sleep advice often walks into pregnancy with a clipboard and one question: how many hours?
That question is not useless. It is also not enough.
A new SLEEP study followed 1,186 pregnant women with gestational diabetes in the WeBirth cohort in Hangzhou, China. The design had one major strength: sleep and glucose were measured at the same time over 14 days. Participants wore accelerometers, and continuous glucose monitoring captured overnight glucose patterns.
The result was not a miracle story. It was more interesting than that.
Higher sleep regularity was associated with lower nocturnal glucose variability and less time above range. Higher sleep efficiency showed similar associations. In joint analyses, people with both irregular and inefficient sleep had higher mean nocturnal glucose than those with regular and efficient sleep.
That sounds practical, which is where we need to slow down.
This was an observational cohort analysis, not a treatment trial. It does not prove that changing bedtime, sleep efficiency, or regularity will control gestational diabetes. It does not replace glucose monitoring, nutrition care, medication, or prenatal care. It does not tell any pregnant reader what to do tonight.
What it does do is sharpen the question. In gestational diabetes, sleep may be more than a side note and more than a total-hours number. Timing and continuity may matter, too.
That is a useful signal, provided nobody tries to turn it into a wellness commandment by breakfast.
THIS WEEK IN SLEEP
Another SLEEP paper looked at a problem many clinics already see: obstructive sleep apnea and insomnia do not always stay in separate lanes.
Researchers used University of Michigan electronic medical records from 2012 through 2023. The cohort included 313,953 adults ages 40 to 85 who were insomnia-free for one year after entering the cohort. An obstructive sleep apnea diagnosis was associated with higher later risk of incident insomnia.
The equity detail matters. The association was stronger in Arab and Black patients than in White patients. The hazard ratio was 1.57 for Arab patients, 1.57 for Black patients, and 1.39 for White patients.
This is not proof that apnea caused insomnia. Electronic medical record studies depend on diagnosis, coding, health-system contact, and who gets evaluated in the first place. Arab ethnicity was identified by a surname algorithm, while Black and White categories came from EMR self-report. That method can help answer one question while raising others.
Still, the clinical point lands. COMISA, comorbid insomnia and sleep apnea, is not just a tidy acronym. It may develop over time, and the risk may not fall evenly across groups.
THE DEEP DIVE
The gestational-diabetes study is strongest when read as a measurement story.
For years, many public sleep messages have made duration the main event. Duration is easy to ask about. It also fits nicely in a headline, which is one reason the internet keeps feeding it snacks.
But sleep is built from several parts. Regularity asks whether sleep timing is stable. Efficiency asks how much time in bed is actually sleep. Continuity asks whether the night stays together or keeps breaking apart. In pregnancy, and especially in gestational diabetes, those parts may sit next to glucose regulation in ways worth studying.
The new SLEEP paper paired accelerometer sleep data with continuous glucose monitoring across the same 14-day window. That matters because it avoids a common mismatch: asking someone to remember sleep while measuring a body system with much finer detail.
The core findings were association-based. Per one standard deviation higher sleep regularity, nocturnal glucose variability and time above range were lower. Sleep efficiency showed comparable associations. The authors adjusted for potential confounders including maternal age, pre-pregnancy BMI, and gestational weeks.
Good adjustment does not erase all uncertainty. This was one regional cohort in China. The abstract does not establish birth outcomes, medication changes, long-term child health, or a tested sleep intervention. Residual confounding remains possible.
So the clean takeaway is not "sleep fixes glucose." It is this: if clinicians and researchers care about overnight glucose in gestational diabetes, they may need to look beyond total sleep time.
That is less catchy than a rule. It is also more honest.
REALITY CHECK
A new Sleep Health paper offers a useful reminder: mortality headlines need adult supervision.
The study examined sleep phenotypes in older adults using two large cohorts: 2,844 men with an average age of 76.4 years and 12.1 years of follow-up, and 2,967 women with an average age of 83.5 years and 5.7 years of follow-up.
The authors looked at insomnia with objective short sleep duration, objective long sleep duration, insomnia with normal sleep duration, and asymptomatic short sleep. Sleep duration came from actigraphy.
In demographic-adjusted models, insomnia with objective short sleep duration was linked with higher mortality compared with normal sleep in both men and women. But after adjustment for comorbidity, that association was no longer significant in either cohort.
Long sleep was different. In women only, objective long sleep remained associated with higher mortality after adjustment for demographics and comorbidity.
This is not a reason to panic about one long night. It is not a home risk calculator. It is also not proof that treating sleep length lowers mortality.
The careful version is better: in older adults, some sleep phenotypes may mark health risk, but chronic disease may explain part of the link. Sleep can be a signal without being the sole suspect.
NIGHT SHIFT
The brain item this week comes from the Journal of Neuroscience, and it is a good antidote to the idea that falling asleep is just willpower in pajamas.
Twenty healthy young adults completed two afternoon nap sessions with simultaneous polysomnography and magnetic resonance spectroscopy of the medial prefrontal cortex. The researchers used the first-night effect, the way a new sleep setting can disturb sleep, to change sleep onset within participants.
In the second session, glutamate plus glutamine, called Glx, fell during the transition from wakefulness to sleep. That reduction was absent during disturbed sleep in the first session. Across sessions, larger Glx reductions were associated with shorter sleep-onset latency. GABA did not show the same pattern.
This is elegant, early human neuroscience. It does not diagnose insomnia. It does not prove a treatment. It does not mean supplements, gadgets, or breathing tricks can tune Glx in a useful way.
But it does give one useful frame: sleep onset is an active biological transition. The off switch has chemistry.
ONE LAST THING
A second COMISA paper, also in SLEEP, looked under the hood.
Researchers analyzed 124 male participants: 45 controls, 58 with obstructive sleep apnea, and 21 with comorbid insomnia and sleep apnea. They used gray matter data and resting-state fMRI, then applied graph metrics to structural and functional networks.
Both patient groups showed functional hyper-connectivity compared with controls. The COMISA group showed a distinct functional pattern compared with OSA alone, with elevated nodal metrics in thalamostriatal circuits involving the bilateral thalamus and putamen. Exploratory analyses suggested possible links with insomnia and depressive symptom severity.
This does not mean COMISA is now solved. The sample was male, the COMISA group was small, and the symptom links were exploratory. But it does support a larger point from this week's evidence: mixed sleep problems are not just two labels stacked on each other. Sometimes the overlap may have its own biology.
SOURCES
Sleep regularity and efficiency are associated with nocturnal glycemic control in pregnant women with gestational diabetes:
https://pubmed.ncbi.nlm.nih.gov/42784578/
https://doi.org/10.1093/sleep/zsag256
Sleep Apnea and Risk of Comorbid Insomnia in Arab, Black, and White Women and Men:
https://pubmed.ncbi.nlm.nih.gov/42784281/
https://doi.org/10.1093/sleep/zsag257
Prefrontal Glx reduction during the sleep transition is associated with sleep onset latency:
https://pubmed.ncbi.nlm.nih.gov/42785995/
https://doi.org/10.1523/JNEUROSCI.0372-26.2026
Distinct Thalamostriatal Functional Network Dysregulation in Comorbid Insomnia and Sleep Apnea:
https://pubmed.ncbi.nlm.nih.gov/42789003/
https://doi.org/10.1093/sleep/zsag252
Insomnia with objective short sleep duration, objective long sleep duration, and mortality in older persons:
https://pubmed.ncbi.nlm.nih.gov/42791154/
https://doi.org/10.1016/j.sleh.2026.08.013
The Sleep News provides general information, not personal medical advice. Pregnancy, gestational diabetes, glucose management, sleep apnea, insomnia, depression, medication questions, breathing problems during sleep, and child or adult health concerns should be discussed with a qualified clinician.