Poor sleep quality and lipid profile in a rural cohort (The Baependi Heart Study)

Glaucylara Reis Geovanini*, Geraldo Lorenzi-Filho, Lilian K. de Paula, Camila Maciel Oliveira, Rafael de Oliveira Alvim, Felipe Beijamini, André Brooking Negrão, Malcolm von Schantz, Kristen L. Knutson, José Eduardo Krieger, Alexandre Costa Pereira

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

22 Citations (Scopus)
11 Downloads (Pure)


Aim: To test the association between cardiometabolic risk factors and subjective sleep quality assessed by the Pittsburgh sleep quality index (PSQI), independent of obstructive sleep apnea (OSA) and sleep duration. Methods: A total of 573 participants from the Baependi Heart Study, a rural cohort from Brazil, completed sleep questionnaires and underwent polygraphy for OSA evaluation. Multivariable linear regression analysis tested the association between cardiovascular risk factors (outcome variables) and sleep quality measured by PSQI, adjusting for OSA and other potential confounders (age, sex, race, salary/wage, education, marital status, alcohol intake, obesity, smoking, hypertension, and sleep duration). Results: The sample mean age was 43 ± 16 years, 66% were female, and mean body mass index (BMI) was 26 ± 5 kg/m 2 . Only 20% were classified as obese (BMI ≥30). Overall, 50% of participants reported poor sleep quality as defined by a PSQI score ≥5. A high PSQI score was significantly associated with higher very-low-density lipoprotein (VLDL) cholesterol levels (beta = 0.392, p = 0.012) and higher triglyceride levels (beta = 0.017, p = 0.006), even after adjustments, including the apnea–hypopnea index. Further adjustments accounting for marital status, alcohol intake, and medication use did not change these findings. No significant association was observed between PSQI scores and glucose or blood pressure. According to PSQI components, sleep disturbances (beta = 1.976, p = 0.027), sleep medication use (beta = 1.121, p = 0.019), and daytime dysfunction (beta = 1.290, p = 0.024) were significantly associated with higher VLDL serum levels. Only the daytime dysfunction domain of the PSQI components was significantly associated with higher triglyceride levels (beta = 0.066, p = 0.004). Conclusion: Poorer lipid profile was independently associated with poor sleep quality, assessed by the PSQI questionnaire, regardless of a normal sleep duration and accounting for OSA and socio-economic status.

Original languageEnglish
Pages (from-to)30-35
Number of pages6
JournalSleep Medicine
Early online date30 Jan 2019
Publication statusPublished - 1 May 2019
Externally publishedYes


Dive into the research topics of 'Poor sleep quality and lipid profile in a rural cohort (The Baependi Heart Study)'. Together they form a unique fingerprint.

Cite this