Main Findings and Interpretation
As reported in previous studies, results of our analyses suggest that there are racial/ethnic differences in the prevalence of PDM and GDM. However, our findings expand on previous studies by reporting data on trends in pregnancy-associated diabetes by race/ethnicity. Consistent with earlier HCUP NIS studies, our results show an increasing trend in the prevalence of PDM and GDM from 2002 to 2017.10 In addition, our data suggest the prevalence of PDM and GDM is increasing among NH-White, NH-Black and Hispanic hospitalized mothers. These findings correlate with increasing trends in risk factors for diabetes such as obesity, inactivity, and hypertension observed among specific racial/ethnic minority women.25–30 Another possible contributing factor is an increase in advanced maternal age pregnancies.31–33
Note that the graphs in Figure 1 show an inflection point in trends for PDM and GDM in 2015. This is likely caused by two factors. First, the implementation of U.S. Prevent Services Task Force recommendations for routine GDM screening in asymptomatic pregnant women after 24 weeks of pregnancy in 2014.34 Secondly, a change in disease reporting in 2015 due to the transition from the ICD-9 to ICD-10 coding set.35 These changes likely increased screening for PDM and GDM as well as disease reporting.
The majority of associations we found between pregnancy-associated diabetes, race/ethnicity and maternal characteristics are consistent with previously reported data.10,36,37 One exception is the increase in GDM prevalence with increasing zip code income observed among hospitalized NH-Black and Hispanic women. A previously published study conducted in California found that high-income zip codes had a lower prevalence of GDM compared to low-income zip codes.38 These conflicting results suggest that factors such as diabetes screening, exercise, and diet could be more important than previously suspected to the prevalence of GDM in certain racial/ethnic groups.
The highest prevalence rates for PDM and GDM were observed among Hispanic and NH-Black pregnant mothers discharged against medical advice (DAMA). Previously published data suggest DAMA is more common among hospitalized pregnant women with PDM and GDM compared to those without diabetes.39 DAMA among hospitalized pregnant women is also associated with Black race, public insurance, substance abuse, mental illness, chronic hypertension, neonatal morbidity and fetal death.40 Additional research is needed to better characterize this vulnerable group and develop screening tools for intervention.
Another notable finding is the prevalence of PDM among NH-Black pregnant women who died during hospitalization. Among pregnant women who died during hospitalization, NH-Black had the highest prevalence of PDM when compared to the other race/ethnic groups. Our results are consistent with previously published NIS data.41 The causal pathway between PDM and maternal mortality among hospitalized NH-Black women is not clear but studies report African American race/ethnicity is a significant independent risk-factor for pregnancy-related cardiovascular conditions including venous thromboembolism and stroke that can be fatal.42,43 Further, type 1 diabetes has been associated with an increased risk of maternal mortality due to complications such as hypoglycemia.44 These data suggest NH-Black pregnant women with PDM could be predisposed to glycemic and cardiovascular conditions that increase risk for maternal mortality. However, additional research is needed.
Consistent with previously published data, stillbirth rates for hospitalized pregnant women in our study were higher for women with PDM than those with GDM.12,14 Unlike previous studies, results of our analyses show that racial/ethnic differences in stillbirth rates exist. Additionally, we are the first to report adjusted odds of stillbirth for Hispanic mothers with PDM. Hispanic mothers with PDM had the highest adjusted odds of stillbirth when compared to their non-diabetic counterparts in our study, a previously unreported finding. Reasons for differences across racial/ethnic groups are likely underlying maternal and fetal characteristics that have not been well-studied. Previously reported risks factors for stillbirth among woman with PDM include maternal age, BMI, gestational infant size, male gender, parity, tobacco use and type of delivery.18,35,37,45 The extent to which these factors are influenced by race/ethnicity is not clear and should be taken into consideration.
Interestingly, after adjustment for maternal and hospital characteristics, we found that GDM was protective for stillbirth in each race/ethnic group, decreasing the odds of stillbirth by approximately 50%. Similar findings are reported by Lavery et al.46One possible explanation could be related to the large percentage (70-78%) of women in our study with private or public medical coverage. Medical coverage may result in more aggressive diabetes screening during pregnancy and better access to follow-up care decreasing the occurrence of stillbirth.