Association of maternal prenatal selenium concentration and preterm birth: A multicountry meta-analysis

Nagendra Monangi, Huan Xu, Rasheda Khanam, Waqasuddin Khan, Saikat Deb, Jesmin Pervin, Joan T. Price, Stephen H. Kennedy, Abdullah Al Mahmud, Yuemei Fan, Thanh Q. Le, Angharad Care, Julio A. Landero, Gerald F. Combs, Elizabeth Belling, Joanne Chappell, Fansheng Kong, Criag Lacher, Salahuddin Ahmed, Nabidul Haque ChowdhurySayedur Rahman, Furqan Kabir, Imran Nisar, Aneeta Hotwani, Usma Mehmood, Ambreen Nizar, Javairia Khalid, Usha Dhingra, Arup Dutta, Said Ali, Fahad Aftab, Mohammed Hamad Juma, Monjur Rahman, Bellington Vwalika, Patrick Musonda, Tahmeed Ahmed, Md Munirul Islam, Ulla Ashorn, Kenneth Maleta, Mikko Hallman, Laura Goodfellow, Juhi K. Gupta, Ana Alfirevic, Susan Murphy, Larry Rand, Kelli K. Ryckman, Jeffrey C. Murray, Rajiv Bahl, James A. Litch, Courtney Baruch-Gravett, Zarko Alfirevic, Per Ashorn, Abdullah Baqui, Jane Hirst, Cathrine Hoyo, Fyezah Jehan, Laura L. Jelliffe-Pawlowski, Anisur Rahman, Daniel E. Roth, Sunil Sazawal, Jeffrey Stringer, Ge Zhang, Louis Muglia

Research output: Contribution to journalArticlepeer-review

15 Citations (Scopus)


Background Selenium (Se), an essential trace mineral, has been implicated in preterm birth (PTB). We aimed to determine the association of maternal Se concentrations during pregnancy with PTB risk and gestational duration in a large number of samples collected from diverse populations. Methods Gestational duration data and maternal plasma or serum samples of 9946 singleton live births were obtained from 17 geographically diverse study cohorts. Maternal Se concentrations were determined by inductively coupled plasma mass spectrometry analysis. The associations between maternal Se with PTB and gestational duration were analysed using logistic and linear regressions. The results were then combined using fixed-effect and random-effect meta-analysis. Findings In all study samples, the Se concentrations followed a normal distribution with a mean of 93.8 ng/mL (SD: 28.5 ng/mL) but varied substantially across different sites. The fixed-effect meta-analysis across the 17 cohorts showed that Se was significantly associated with PTB and gestational duration with effect size estimates of an OR=0.95 (95% CI: 0.9 to 1.00) for PTB and 0.66 days (95% CI: 0.38 to 0.94) longer gestation per 15 ng/mL increase in Se concentration. However, there was a substantial heterogeneity among study cohorts and the random-effect meta-analysis did not achieve statistical significance. The largest effect sizes were observed in UK (Liverpool) cohort, and most significant associations were observed in samples from Malawi. Interpretation While our study observed statistically significant associations between maternal Se concentration and PTB at some sites, this did not generalise across the entire cohort. Whether population-specific factors explain the heterogeneity of our findings warrants further investigation. Further evidence is needed to understand the biologic pathways, clinical efficacy and safety, before changes to antenatal nutritional recommendations for Se supplementation are considered.

Original languageEnglish
Article numbere005856
JournalBMJ Global Health
Issue number9
Publication statusPublished - 13 Sept 2021


  • child health
  • environmental health
  • epidemiology
  • maternal health
  • nutrition


Dive into the research topics of 'Association of maternal prenatal selenium concentration and preterm birth: A multicountry meta-analysis'. Together they form a unique fingerprint.

Cite this