Sleep Quality and Glycemic Variability Among Nondiabetic African American Pregnant Women

To examine between-person and within-person associations between sleep health and glycemic outcomes among nondiabetic African American pregnant women.

Secondary analysis of baseline data from a randomized controlled trial.

Prenatal care clinics affiliated with a large urban academic medical center in the United States.

Nondiabetic African American pregnant women (16-22 weeks gestation) aged 18 to 42 years with singleton pregnancies and body mass indices from 24 to 45 kg/m² (N = 55).

We assessed baseline subjective sleep quality during the past month using the Pittsburgh Sleep Quality Index [PSQI]. We objectively measured sleep (duration, efficiency, timing, and regularity) using a wrist-worn Fitbit Inspire 2 over 7 consecutive days. We measured glycemic outcomes (glucose mean and standard deviation, mean amplitude of glycemic excursions [MAGE], and coefficient of variation [CV]) using 7-day continuous glucose monitoring. We evaluated between-person associations among participants using multivariable regression, and we evaluated within-person associations using linear mixed-effects models with repeated daily measures over 7 days.

Poorer subjective sleep quality, measured by the PSQI, was significantly associated with higher glucose SD (B = 0.42, 95% confidence interval (CI) [0.07, 0.76], p = .02), higher MAGE (B = 1.10, 95% CI [0.30, 1.91], p = .01), and greater CV (B = 0.44, 95% CI [0.12, 0.75], p = .01). No objective sleep dimensions measured by the Fitbit Inspire 2 were significantly associated with glycemic outcomes.

Among nondiabetic African American pregnant women, poorer subjective sleep quality was associated with greater glycemic variability. Assessment of perceived sleep quality may help to identify early pregnancy metabolic risk and warrants further investigation as a modifiable target in prenatal care.