Normative serum cortisol levels in second and third trimesters and their associated factors: A prospective cohort study from Sri Lanka.

OBJECTIVE

To describe the normative serum cortisol levels during 25-29 weeks of POG and the association of maternal, psychological, and social factors on serum cortisol in the second and third trimesters in a cohort of pregnant women.

METHODS

All eligible pregnant women registered in the maternal care program in Anuradhapura district, Sri Lanka, from July to September 2019 were invited to the Rajarata Pregnancy Cohort (RaPCo). An interviewer-administered questionnaire-based symptom analysis and clinical assessment were conducted at baseline in the first trimester and at follow-up from 25 to 29 weeks POG. We assessed fasting early morning serum cortisol levels at the follow-up visit.

RESULTS

The study sample included 1010 pregnant women with a mean age in years and POG in weeks at baseline of 28 (±6) and 10 (±3), respectively. The mean (SD, 97% percentile) serum cortisol level in all pregnant women was 10.93 (±3.83, 20.95) μg/dL, with no significant difference between singleton and twin pregnancies (p = 0.138). None of the study participants had a cortisol level exceeding the upper limit of 42 μg/dL, and 464 (45.9%) had levels less than 10 μg/dL. Serum cortisol levels were higher in women with an advanced POG, with a mean of 10.33 µg/dL (95%CI: 9.68-10.98) at 24 weeks POG and 12.23 µg/dL (95%CI: 11.15-13.32) at 29 weeks POG (p = 0.049). Primi-gravidity (p = 0.004), history of miscarriage (p = 0.010), BMI categories (p = 0.044), and POG (p = 0.002) were independently associated with serum cortisol levels in robust regression. An EPDS score of more than 9 was not associated with serum cortisol (p = 0.633).

CONCLUSION

The pregnant women in rural Sri Lanka reported a low mean serum cortisol level, which gradually increased with the POG. Significantly higher mean serum cortisol was associated with primi-gravidity, history of miscarriage, pre-pregnancy BMI, and POG at cortisol test, but not psychological factors.