Research identifies a unique protein fingerprint linked to very short sleep and increased diabetes risk

By Dr. Liji Thomas, MDReviewed by Benedette Cuffari, M.Sc.May 1 2024

Diabetes mellitus is a lifestyle disorder that is associated with various adverse health outcomes, including cardiovascular disease (CVD) and unhealthy sleep patterns. Epidemiologic studies and other research emphasize the association between both long and short sleep duration, CVD, and metabolic disease.

Both short sleep and a sleep duration exceeding eight hours are pro-inflammatory; however, the markers of inflammation are nonspecific, as they do not reflect sleep duration or health outcome.

A recent study published in the journal BMC Medicine explores the serum protein profile associated with sleep duration to identify any correlation between new-onset coronary heart disease (CHD) and DM.

Study: Very short sleep duration reveals a proteomic fingerprint that is selectively associated with incident diabetes mellitus but not with incident coronary heart disease: a cohort study. Image Credit: Sklo Studio / Shutterstock.com

About the study

The current study examines proteins that may be implicated in different sleep durations to generate proteomic scores that can predict the sleep duration category for each score and identify associations between these scores with incident DM (iDM) and CHD (iCHD).

Over 3,300 participants between the ages of 46 and 68 years were included in the study, from whom a total of 78 plasma proteins were measured. Although none of the study participants had a history of DM or CHD, any individual who developed iDM or iCHD was identified.

Sleep duration data were used to classify study participants into quintiles from Q1 to Q5, wherein Q3 was used as the reference. Q1 included those with the shortest mean sleep duration, whereas Q5 included those with the longest. Circulating protein levels were combined to obtain proteomic scores (PS), which were then analyzed for associations with Q1, Q2, Q4, and Q5.

Q3 had the lowest mean age, fewest symptoms of insomnia, and waist circumference. Health behaviors were least favorable in this cohort, including lowest physical activity, least non-drinkers, and most individuals with at least elementary school education who drank heavily.

Comparatively, Q1 participants had the fewest drinkers, highest insomnia scores, and highest low-density lipoprotein levels.

What did the study show?

Among the nearly 78 serum proteins analyzed, 16 were associated with one or more sleep duration quintiles, 13 of which were associated with only one quintile. Six and four proteomic markers were associated with Q1 and Q2, whereas five and six proteomic markers were associated with the long sleep duration quintiles Q4 and Q5, respectively.

PS and iDM

With all sleep duration quintiles, the risk of iDM was higher than in Q3. For Q1 and Q2, the risk of iDM was about 30% higher, compared to about 50% higher for Q4 and Q5, respectively.

The most accurate predictions of sleep duration were with proteomic scores for Q1 and Q5, which were denoted as PSQ1 and PSQ5. Increases in PSQ1 were accompanied by a 27% rise in iDM.

When these scores were included in the model parameters, there was no longer any significant correlation between Q1 and iDM. Thus, PSQ1 accounted for 30-50% of the association from years 11 to 27.

PS and iCHD

The iCHD risk was higher for Q1. No association was observed between PSQ1-PSQ5 and iCHD, and their inclusion did not significantly reduce the observed association between Q1 and iCHD.

These findings suggest that sleep duration is associated with both iDM and iCHD. The 16 proteomic markers identified in the current study successfully predicted specific quintiles of sleep duration as against baseline or Q3. The use of proteomic scores also showed differential associations between the scores and health outcomes.

Conclusions

The findings confirm the independent and positive association between both short and long sleep durations with incident DM…a 37% increased risk of CHD only was observed for those with the shortest sleep duration when compared to Q3.”

If sex stratification is used in future studies, the longest sleep duration might reveal an equal or greater association with CHD risk. This trend has been observed in a larger sample from the same cohort, wherein men who slept more than nine hours a night had a 33% higher CHD risk than was not observed in women with similar sleep patterns.

Another explanation for the absence of a clear correlation between Q5 and iCHD could be attributed to this quintile, which comprises sleep durations from eight to 14 hours, thereby affecting the lower limit of normal sleep duration.

The proteomic score for Q1 could be considered a proteomic fingerprint of very short sleep duration and might explain its association with iDM. Nevertheless, confounding factors may have prevented the identification of this type of association for the longest sleep quintile.

The absence of a significant association between iCHD and proteomic scores for Q1 and Q5 may indicate that the increased risk of iCHD in very short sleep cohorts is due to iDM itself. The biological mechanism may involve inflammation and apoptosis triggered by very short sleep patterns that lead to iDM and an increased risk of iCHD.

The proteomic markers identified in the current study suggest that proteins involved in inflammation and cell death are raised with very short sleep duration. Conversely, proteins associated with new vessel growth and anti-inflammatory factors are reduced.

Q5 had both positive and negative associations with several proteomic markers, including those involved in new vessel growth, cell proliferation, clot formation, and cell death. For example, tissue plasminogen activator (tPA) has been reported to be low in habitual short sleep but increased in obstructive sleep apnea (OSA). In the current study, tPA levels were raised in Q5, which may indicate the presence of OSA in this subgroup.

Of the proteomic markers analyzed in the current study, only follistatin and E-selectin were raised in short sleep durations.