Protein intake and its association with sleep: results from 3 cohort studies of U.S. American health professionals

Janine Wirth, Kun Wu, Lorraine BrennanProceedings of The Nutrition Society 2022 · DOI

TL;DR

While total protein intake doesn't seem to impact how well you sleep, the source of that protein matters significantly. In a massive study of nearly 100,000 health professionals, researchers found that higher vegetable protein intake is linked to better sleep quality, while animal-based proteins may have the opposite effect.

What Problem Does This Paper Address?

Sleep isn't just about feeling rested; it’s a critical pillar of health linked to chronic disease prevention and daily cognitive functioning. As sleep disorders become a growing public health crisis, researchers are looking toward modifiable lifestyle factors—like diet—to find solutions.

While we know that what we eat affects our metabolic health, the specific relationship between protein consumption and sleep quality has remained murky. Previous studies often looked at total protein in isolation. This paper seeks to fill that gap by analyzing whether the type of protein (animal, dairy, or vegetable) influences sleep quality across three large, long-term cohorts of U.S. health professionals.

Methods

The researchers conducted a secondary analysis of data from three major longitudinal studies:

  • The Nurses’ Health Study (NHS)
  • The Nurses’ Health Study II (NHS II)
  • The Health Professionals’ Follow-up Study (HPFS)

To track diet, the team used validated semiquantitative food frequency questionnaires (FFQs) administered every four years, capturing habitual intake over decades. Sleep quality was assessed using the Pittsburgh Sleep Quality Index (PSQI)—a gold-standard self-reporting tool—or adapted versions of it.

To find patterns, they used ordinal logistic regression to calculate Odds Ratios (OR). They grouped participants into five sleep quality categories, comparing the "best" sleepers (Category 1) against those with poorer sleep (Categories 2-5) based on their protein intake levels.

Datasets

The study leveraged three powerhouse datasets in the world of epidemiology:

  1. NHS (32,212 women): Data collected in 2014.
  2. NHS II (51,976 women): Data collected in 2017.
  3. HPFS (14,890 men): Data collected in 2012.

By excluding participants with chronic diseases or missing data, the researchers ensured a robust sample size of nearly 100,000 individuals, providing significant statistical power compared to smaller clinical trials.

Key Insights and Ideas

  • Total Protein is Neutral: If you only look at the "grams of protein" on a nutrition label, you won't find a correlation with sleep. The "total protein" metric masks the underlying effects of specific food sources.
  • The Plant-Based Advantage: The most consistent finding across the cohorts was that vegetable protein is a "win" for sleep. This suggests that the non-protein components of plant foods (like fiber or phytonutrients) or the specific amino acid profiles of plants might be more sleep-conducive.
  • Animal Protein's Negative Trend: There was a "significant negative association" between animal protein and sleep quality in the NHS II cohort. Essentially, the more animal protein consumed, the lower the odds of being in the "good sleep" category.
  • The Dairy Dilemma: Results for dairy were contradictory. It was associated with better sleep in the younger women of NHS II but worse sleep in the men of the HPFS. This suggests that the impact of dairy might be moderated by age, sex, or perhaps the specific type of dairy (e.g., fermented vs. non-fermented).

Key Results

  • Vegetable Protein: Participants with the highest intake showed significantly better sleep quality.
  • NHS: OR 1.10 (95% CI: 1.02–1.19)
    • HPFS: OR 1.14 (95% CI: 1.03–1.26)
  • Animal Protein (NHS II): Comparing the highest quartile of intake to the lowest, the Odds Ratio for good sleep was 0.93 (95% CI: 0.88–0.99), indicating a slight but statistically significant decrease in sleep quality for high animal protein consumers.
  • Total Protein: No significant association was found across any of the three cohorts.

Limitations and Future Work

The authors acknowledge several hurdles that prevent a "smoking gun" conclusion:

  • Observational Nature: Because this is a cohort study, it can show associations but cannot prove that vegetable protein causes better sleep.
  • Self-Reporting: Both diet (FFQs) and sleep (PSQI) rely on participant memory and perception, which can introduce bias.
  • Inconclusive Dairy Data: The conflicting results between men and women/different age groups regarding dairy protein require more granular investigation.
  • Mechanistic Gap: The study doesn't explore why these proteins affect sleep. Future research could look at how different protein sources affect the production of sleep-regulating hormones like melatonin or serotonin.

Final Takeaway: If you're looking to optimize your sleep through diet, swapping some animal-based proteins for plant-based sources (like beans, nuts, or lentils) appears to be a scientifically backed strategy.

Summary updated

Abstract

Sleep is related to chronic diseases and overall daytime functioning, and sleep disorders are a growing public health issue. Protein intake is a possible health beneficial factor influencing the quality of sleep. We investigated the association between protein intake and sleep quality in three U.S. American cohorts of health professionals. The Nurses' Health Study (NHS), the NHS II and the Health Professionals' Follow-up study (HPFS) were used for data analysis. Habitual dietary intake during the past year was assessed at baseline and every 4 years using validated semiquantitative food frequency questionnaires. In 2012 (HPFS), 2014 (NHS) and 2017 (NHS 2) sleep quality was measured in detail. In NHS, the original Pittsburgh Sleep Quality Index (PSQI) was used to measure sleep quality, while NHS II and HPFS used adapted versions. The sleep score was grouped into five similarly sized categories for ordinal logistic regression and Odds ratios (OR) and 95% confidence intervals (CI) were calculated to estimate the odds of good sleep quality (category 1) versus poorer sleep quality (categories 2-5) depending on protein intake. After excluding participants with chronic diseases, missing exposure or outcome data, our study populations comprised 32,212 women from NHS, 51,976 women from NHS 2, and 14,890 men from HPFS. Overall, no association between total protein intake and sleep quality was observed in the three cohorts. Women from the NHS 2 showed a significant negative association with sleep quality comparing quartile 1 of animal protein intake with quartile 4 (OR: 0.93, 95%CI: 0.88-0.99), although this trend was not significant in continuous analysis per 2% increase (0.99, 0.97-1.01).
BibTeX
@inproceedings{wirth2022protein,
  author = {Janine Wirth and Kun Wu and Lorraine Brennan},
  title = {{Protein intake and its association with sleep: results from 3 cohort studies of U.S. American health professionals}},
  booktitle = {Proceedings of The Nutrition Society},
  year = {2022},
  doi = {10.1017/s0029665122001112},
}
Figure