The epidemiology of pregnancy loss: global burden, variable risk factors, and predictions

Is the incidence of pregnancy loss correlated with various geographic, socio-demographic, and age stratifications at the societal and national levels, and what are the risk factors associated with pregnancy loss at the individual level?

The epidemiological trends and disease burden of pregnancy loss were correlated with various geographic, socio-demographic, and age stratifications, and we identified that poor health condition, smoking, sedentary behaviour, lower educational level, and lower maternal birth weight may significantly increase the risk of pregnancy loss.

Several studies have used national, regional, or single-centre data to describe trends in the burden of pregnancy loss, and previous observational studies have identified some variable factors possibly associated with pregnancy loss. However, a comprehensive analysis of global trends and predictions of pregnancy loss are lacking, and the conclusions have been inconsistent.

We have utilized the data from Global Burden of Disease (GBD) 2019 to provide an overview of the trends in pregnancy loss in 204 countries and regions worldwide from 1990 to 2019, and have made a forecast for the next 10 years. Moreover, we applied a variety of statistical genetics methods to analyse 34 239 pregnancy loss and 89 340 non-pregnancy loss cases from the FinnGen consortium to comprehensively assess the bidirectional causality of variable factors with pregnancy loss from an individual perspective.

We analysed trends in the incidence, disability-adjusted life years (DALYs), and maternal mortality of pregnancy loss at global, regional, national, socio-demographic index (SDI), and age levels. The autoregressive integrated moving average (ARIMA) model was used to predict trends by 2030. Finally, we used two-sample Mendelian randomization (MR) and multivariate MR (MVMR) analyses to explore the relationship between the pregnancy loss and variables closely related to physical condition, physical activity, lifestyle, sleep conditions, basic conditions.

In 2019, there were approximately 42.39 million cases of pregnancy loss worldwide. Globally, the incidence, DALYs, and mortality of pregnancy loss showed a decreasing trend between 1990 and 2019, although the number was increasing in some countries. The age-standardized incidence, DALYs, and mortality rate were negatively correlated with SDI level and show a further decline by 2030. Based on MR analyses, we confirmed that genetically predicted overall health rating (inverse-variance weighted (IVW) odds ratio (OR), 1.68; 95% CI, 1.34-2.13; P = 5.10 × 10-6), smoking initiation (IVW OR, 1.26; 95% CI, 1.16-1.38; P = 1.90 × 10-9), sedentary behaviour (IVW OR, 1.56; 95% CI, 1.20-2.01; P = 2.76 × 10-5), educational level (IVW OR, 0.64; 95% CI, 0.55-0.73; P = 6.56 × 10-10), and maternal birth weight (IVW OR, 0.70; 95% CI, 0.58-0.85; P = 2.98 × 10-4) were significantly related to the risk of pregnancy loss, whereas body mass index (IVW OR, 1.10; 95% CI, 1.03-1.17; P = 5.31 × 10-3), alcohol consumption (IVW OR, 1.74; 95% CI, 1.03-2.95; P = 0.04), insomnia (IVW OR, 1.66; 95% CI, 1.14-2.42; P = 7.00 × 10-3), and moderate-to-vigorous physical activity (IVW OR, 0.59; 95% CI, 0.37-0.95; P = 2.85 × 10-2) were suggestively associated with the risk of pregnancy loss. These results were supported by sensitivity and directional analyses.

Despite efforts to standardize GBD data from all over the world, uncertainties in data quality control regarding ascertainment of pregnancy loss, medical care accessibility, cultural differences, and socioeconomic status still exist. Furthermore, the population in the MRstudy was limited to Europeans, which means that the results may not be extrapolated to people of other origins.

Our study provides for the first time an overview of the epidemiological trends and disease burden of pregnancy loss related with SDI, region, country, and age, and predicts changes in future trends up to 2030. In addition, findings support that genetic susceptibility, smoking, health condition, and sedentary behaviour may be powerful indicators of an increased risk of pregnancy loss. These results would be beneficial for policy makers of different countries and regions to improve prevention implementation.

This work was supported by grants 2021JH2/10300093, from the Science and Technology Projects of Liaoning Province, China. All authors declare no conflicts of interest.

N/A.