Trends in Short Interpregnancy Interval Births in the United States, 2016–2022

The prevalence of short interpregnancy interval births remained relatively stable between 2016 and 2022 in the United States.

Abstract

OBJECTIVE:

To measure contemporary trends in the prevalence of short interpregnancy interval (IPI) births in the United States.

METHODS:

We conducted a repeated cross-sectional analysis using 2016–2022 natality data from the National Vital Statistics System. We included all singleton live births to individuals with at least one prior live birth. We examined trends over time in short IPIs less than 18 months, as well as for specific durations within this time frame (less than 6 months, 6–11 months, and 12–17 months), using linear probability models that estimated changes in the prevalence of each IPI duration over time. We then estimated the prevalence of each short IPI duration by maternal race and ethnicity, socioeconomic characteristics (age, education, insurance payer at delivery), and geography (U.S. census region, state of residence).

RESULTS:

The study sample included 14,770,411 singleton live births to individuals with at least one prior live birth in 2016–2022. Roughly a third (29.8%) of births had an overall IPI of less than 18 months (5.0% less than 6 months, 11.0% 6–11 months, and 13.8% 12–17 months). For IPIs less than 6 months, a slight statistical decline in prevalence was identified over the study period in unadjusted and adjusted models (adjusted annual percentage point change −0.02, 95% CI, −0.03 to −0.02). Slight statistical increases in the prevalence of IPIs of 6–11 and 12–17 months were identified in unadjusted models but were no longer significant and reversed direction in adjusted models, respectively. Long-standing inequities in the distribution of the shortest IPIs (less than 6 months) were stable compared with prior work across the indicators examined in this study.

CONCLUSION:

Overall, it appears the prevalence of short IPIs has remained stable between 2016 and 2022.

Short interpregnancy intervals (IPIs), defined as less than 18 months between a live birth and the beginning of a subsequent pregnancy, accounted for nearly a third of second or higher-order U.S. births as of 2014¹ and comprise an estimated 600,000 births annually. Recent literature suggests that, in high-resource settings such as the United States, IPIs less than 6 months are significantly associated with increased risk of preterm birth; infant death and IPIs of 6–11 and 12–17 months had smaller point estimates associated with these outcomes.² Some studies have attributed the increased risk of adverse outcomes associated with short IPIs to the different sociodemographic characteristics and social health risks experienced by individuals becoming pregnant within this time frame^3,4; others have implicated the causal effect of the IPI itself.^5,6 Although the degree to which these factors drive health outcomes remains debated,⁷ there is agreement that individuals who are from minoritized backgrounds, younger, and from lower income communities experience the greatest prevalence of very short IPIs (less than 6 months),¹ a large proportion of which are unintended (mistimed or unwanted).^8,9

Reducing short IPIs is a clinical and public health priority in the United States. The American College of Obstetricians and Gynecologists recommends that individuals be advised to avoid IPIs shorter than 6 months (grade 1B) and counseled about the risks and benefits of repeat pregnancy sooner than 18 months (grade 2B).¹⁰ The hypothesis that the United States may be on the way to achieving this goal is supported by recent events, including improvements in postpartum contraceptive access after the Affordable Care Act,¹¹ unbundling of Medicaid payments for immediate postpartum long-acting reversible contraception from global maternity payments,^12–14 fertility postponement associated with the coronavirus disease 2019 (COVID-19) pandemic,^15,16 and extensions of pregnancy-related Medicaid coverage beyond 60 days postpartum.^17–19

However, trends in short IPIs have not been comprehensively examined since 2014¹—data that would aid contemporary patient counseling and design of clinical and policy interventions aimed at reducing unintended pregnancy and associated health burdens among desired pregnancies with short IPIs, particularly very short IPIs. Our objective was to examine trends and sociodemographic differences in short IPI prevalence using the most recently available years of national data.

METHODS

We implemented a repeated cross-sectional study design using 2016–2022 natality data from the National Vital Statistics System, which includes a census of all live births in the United States,²⁰ including information captured on the birth certificate, based on a combination of medical records and self-report.^21–23 We included all singleton live births to individuals with at least one prior live birth because IPI occurs among multiparous individuals. We excluded multifetal gestations because IPI cannot be determined for multifetal gestations with the National Vital Statistics System data.²⁴ Furthermore, we excluded births for which there were missing data or an implausible value (0 months or less) for IPI.

Our primary outcome of interest was short IPI (less than 18 months). We calculated IPI using two measures available in the National Vital Statistics System data: interval since previous live birth (ie, the number of months between the last live birth and current live birth) and gestational age (in completed weeks). We converted gestational age to months and then subtracted this value from the number of months since the previous live birth, thereby measuring the number of months between the previous live birth and start of the current pregnancy that ended in a live birth.²⁵ We then created a binary variable to indicate whether the IPI was less than 18 months. Furthermore, we operationalized three additional binary indicators of IPI duration that allowed us to consider prevalence and trends in short IPIs of different durations (less than 6, 6–11, 12–17 months).

We considered six maternal sociodemographic characteristics for which differences in prevalence of short IPI have previously been identified: age at previous live birth (less than 20, 20–24, 25–29, 30–34, 35 or more years); race and ethnicity (Hispanic, non-Hispanic American Indian/Alaska Native, non-Hispanic Asian, non-Hispanic Black, non-Hispanic Native Hawaiian or other Pacific Islander, non-Hispanic multiple races, non-Hispanic White); education level (less than high school, high school, some college or associate's degree, bachelor's degree or more); payer at current delivery (Medicaid, commercial insurance, uninsured [self-pay], other); census region (Midwest, Northeast, South, West); and state of residence.^1,8,26 To provide additional information on which racial identities were represented within the “non-Hispanic multiple races” group, we also used a National Vital Statistics System–provided variable that disaggregates this group by indicating which multiple races were reported (American Indian/Alaska Native, Asian, Black, Native Hawaiian or other Pacific Islander, White). We included race and ethnicity as a measure of the social constructs of race and ethnicity in the United States to serve as an imperfect proxy for exposure to racism.²⁷

We described sample characteristics overall and for the first (2016) and last (2022) years of the study period. We then used unadjusted linear probability models to estimate the percentage point change in each sociodemographic subgroup between 2016 and 2022, converting these estimates to the percent change to provide a relative measure of the magnitude of change over the study period. Then, we assessed trends in the prevalence of each short IPI duration between 2016 and 2022 by 1) plotting the prevalence of short IPI by year to visualize the trend; and 2) estimating the change in prevalence of short IPI over time, presented as the overall percent change and the annual percentage point change between 2016 and 2022, using unadjusted and adjusted linear probability models. We adjusted for sociodemographic characteristics of the birthing individual (age at previous live birth, race and ethnicity, education level, payer at current delivery, and census region) to understand whether compositional changes in the birthing population were driving changes in the prevalence of short IPI births over the study period.

To evaluate contemporary differences in the prevalence of short IPI by sociodemographic characteristics, we limited our analysis to the most recent available year of natality data, 2022. First, we estimated the prevalence for each short IPI duration by sociodemographic characteristics (ie, birthing individual's age, race and ethnicity, education level, payer at current delivery, and census region). To visualize geographic variation at a more granular level, we also generated choropleth maps presenting short IPI prevalence at the state level. Then, we estimated differences in each short IPI duration using unadjusted and adjusted Poisson regression models to estimate prevalence ratios comparing the prevalence of short IPI between each sociodemographic subgroup and the stated reference group. We chose groups with the lowest prevalence of short IPI (less than 18 months) as the reference groups.

All analyses were conducted in Stata/SE 18.0 with a 5% significance level (α=.05) used for statistical estimation and hypothesis testing. We retained observations with missing covariate data by including an “unknown” category (ie, the missing indicator approach). We chose this approach because missingness was low (less than 2% of the analytic sample missing any given covariate: 0.8% race or ethnicity, 1.4% education level, 0.6% payer at delivery), and multiple imputation is generally considered unnecessary when missingness is rare (less than 5%).²⁸ This project was deemed exempt by the study site's IRB.

RESULTS

We identified 26,345,765 live births from 2016 to 2022. We limited the sample to singleton births (n=25,475,170) and to individuals with at least one prior live birth (n=15,525,926). We excluded births for which there were missing data (n=724,697) or an implausible value (n=30,818) for the outcome, short IPI, resulting in an analytic sample of 14,770,411 births (Appendix 1, available online at http://links.lww.com/AOG/D898).

Overall, 50.2% of the study sample was non-Hispanic White, 25.6% was Hispanic, 14.6% was non-Hispanic Black, 5.5% was non-Hispanic Asian, 0.9% was non-Hispanic American Indian/Alaska Native, and 0.3% was non-Hispanic Native Hawaiian or other Pacific Islander (Table 1). The non-Hispanic multiple races category (2.1%) was composed primarily of non-Hispanic Black and non-Hispanic White (39.9%), non-Hispanic Asian and non-Hispanic White (21.1%), and non-Hispanic American Indian/Alaska Native and non-Hispanic White (17.3%) individuals. Medicaid was the primary payer for roughly half of the included deliveries (45.6%), with most others paid by commercial insurance (45.5%); 4.7% were uninsured. Three-quarters of birthing individuals (75.0%) were younger than age 30 years at their previous live birth, and more than half (57.9%) had some formal education beyond high school. Over the study period, higher proportions of the sample fell into older age categories (25–29, 30–34, and 35 or more years); higher education levels (bachelor's degree or more); Hispanic, non-Hispanic Black, non-Hispanic Native Hawaiian or other Pacific Islander, and non-Hispanic multiple races racial and ethnic categories; and the South or Northeast census regions.

Table 1.

Sample Characteristics, 2016–2022 (N=14,770,411)

Nearly a third (29.8%) of births in 2016–2022 followed an IPI of less than 18 months (5.0% less than 6 months, 11.0% 6–11 months, and 13.8% 12–17 months) (Table 2). Overall, the prevalence of short IPIs less than 18 months overall and for each duration examined within less than 18 months appeared relatively stable between 2016 and 2022 (Fig. 1). Estimates from the unadjusted linear probability models suggested that the overall prevalence of a short IPI less than 18 months increased slightly (annual change 0.05 percentage points, 95% CI, 0.04–0.06; representing approximately 1,055 births per year) over the study period but reversed direction in adjusted models that accounted for temporal changes in the sociodemographic composition of the birthing population (annual change −0.08 percentage points, 95% CI, −0.09 to −0.07; representing approximately 1,688 fewer births per year), consistent with a decline of small magnitude.

Table 2.

Overall Percent Change and Annual Percentage Point Change in the Prevalence of Short Interpregnancy Interval Births, 2016–2022 (N=14,770,411)

Fig. 1. Trends in short interpregnancy interval (IPI) births, 2016–2022 (N=14,770,411). Short IPI births were relatively stable between 2016 and 2022 across all durations examined: less than 6 months, 6–11 months, 12–17 months, and less than 18 months overall.

Admon. Short Interpregnancy Interval Births. Obstet Gynecol 2025.

With respect to each IPI duration examined, estimates from the unadjusted linear probability models suggested that the prevalence of short IPIs less than 6 months declined slightly over the study period (annual change −0.02 percentage points, 95% CI, −0.03 to −0.02), whereas the prevalence of short IPIs 6–11 months and 12–17 months increased slightly (annual change 0.03 percentage points, 95% CI, 0.03–0.04; 0.04 percentage points, 95% CI, 0.03–0.05). In adjusted models, the estimated annual change in prevalence of short IPIs less than 6 months was the same (−0.02 percentage points, 95% CI, −0.03 to −0.02). However, the estimated annual change in prevalence of short IPIs 6–11 months (−0.01 percentage points, 95% CI, −0.02 to 0.00) was no longer statistically significant, and the estimated annual change in prevalence of short IPIs 12–17 months reversed direction (−0.05 percentage points, 95% CI, −0.06 to −0.04).

Figure 2 shows the prevalence of short IPIs and each IPI duration examined within short IPIs by maternal sociodemographic characteristics. Appendix 2, http://links.lww.com/AOG/D898, shows prevalence ratios comparing differences in prevalence by sociodemographic subgroups. The greatest prevalence for each short IPI duration was among individuals aged 35 years and older (overall 43.3%, less than 6 months 5.8%, 6–11 months 15.7%, and 12–17 months 21.7%). With respect to education, the greatest prevalence of short IPIs less than18 months overall was among individuals with a bachelor's degree or higher (overall 32.3%, less than 6 months 2.6%, 6–11 months 11.4%, and 12–17 months 18.3%); however, the greatest prevalence of short IPIs less than 6 months was among individuals with less than high school or high school education (7.2% and 7.0%, respectively), and the greatest prevalence of short IPIs 12–17 months was among individuals with a bachelor's degree or higher (18.3%). Among individuals with commercial insurance at birth, 30.1% overall had IPIs less than 18 months (3.2% less than 6 months, 10.8% 6–11 months, and 16.1% 12–17 months). Among individuals with Medicaid insurance at birth, the overall prevalence of IPIs less than 18 months was similar (30.0%); however, a larger proportion of these births were very short IPIs: 7.0% less than 6 months, 11.4% 6–11 months, and 11.6% 12–17 months.

Fig. 2. Prevalence of short interpregnancy interval (IPI) births by duration and maternal sociodemographic characteristics, 2022. Short IPI births less than 6 months were most prevalent among individuals aged 35 years or older, reported as non-Hispanic Native Hawaiian or other Pacific Islanders (NH NHOPI), reporting less than a high school education, insured by Medicaid, and living in the South. Other includes Indian Health Service, military, and other government-sponsored plans or programs. NH, non-Hispanic; AIAN, American Indian/Alaska Native.

Admon. Short Interpregnancy Interval Births. Obstet Gynecol 2025.

Regarding maternal race and ethnicity, the greatest prevalence of short IPIs less than 18 months overall was among non-Hispanic Native Hawaiians and other Pacific Islanders (40.2%), followed by non-Hispanic White (34.0%), non-Hispanic American Indian/Alaska Native (32.5%), and non-Hispanic Black (29.0%) individuals. The greatest prevalence of short IPIs less than 6 months was among non-Hispanic Native Hawaiian and other Pacific Islander (11.6%), non-Hispanic American Indian/Alaska Native (7.9%), non-Hispanic Black (7.0%), and non-Hispanic individuals of multiple races (6.7%). Only non-Hispanic Native Hawaiian and other Pacific Islander individuals had greater prevalence of short IPIs 6–11 months than non-Hispanic White individuals. Non-Hispanic White individuals had the greatest prevalence of short IPIs 12–17 months (17.1%).

Finally, the Midwest had the greatest prevalence of short IPIs less than 18 months overall (33.5%), followed by the Northeast (29.8%), South (29.6%), and West (28.5%). The prevalence of short IPIs (less than 6 months and less than 18 months) can be seen at the state level in Figure 3. The prevalence of short IPIs less than 6 months ranged from 3.4% (Utah and Connecticut) to 7.4% (Mississippi).

Fig. 3. State variation in prevalence of short interpregnancy intervals (IPI) births less than 18 months and less than 6 months in duration, 2022 (n=2,053,605). Short IPI births less than 6 months represent more than 6% of multiparous births in eight states: Alabama, Arkansas, Kentucky, Louisiana, Mississippi, Missouri, Tennessee, and West Virginia.

Admon. Short Interpregnancy Interval Births. Obstet Gynecol 2025.

DISCUSSION

Using national natality data, we found that the rate of short IPIs less than 18 months was relatively stable from 2016 (29.9%) to 2022 (30.2%) (adjusted overall percent change in prevalence −2.3% between 2016 and 2022). These estimates are also relatively stable compared with prior work using similar methods and natality data to assess rates of short IPIs in 2014 (29.0%).¹ In unadjusted results, we observed a small shift toward short IPIs of longer duration (ie, 6–11 months and 12–17 months) from 2016 to 2022. These slight increases appear to be explained by changes in the composition of the birthing population. After adjustment for maternal sociodemographic characteristics, the change in the prevalence of short IPIs 6–11 and 12–17 months were no longer statistically significant and reversed direction, respectively. In other words, given the temporal trends in maternal sociodemographic characteristics (eg, toward older maternal age) from 2016 to 2022, short IPIs between 6 and 17 months actually decreased relative to what would have been expected. Although the proportion of short IPIs less than 6 months was slightly reduced over the study period in both unadjusted and adjusted models, these births still accounted for 5.0% of second or higher-order births in 2022 (from 5.1% in 2016), demonstrating little progress.

In addition, when examining 2022 births, we found differences in short IPI prevalence by the birthing person's age, race and ethnicity, education level, insurance status at delivery, and census region. It is important to note that these differences in prevalence of each short IPI duration were identified across sociodemographic subgroups with one exception: age. Birthing individuals aged 35 years and older had the highest prevalence of all IPI durations examined (less than 18 months; less than 6, 6–11, and 12–17 months). An IPI less than 6 months, however, was generally more common among less socioeconomically advantaged identities (ie, individuals who are racially minoritized, have completed fewer years of education, and had Medicaid insurance at birth) and among those in the South. In contrast, IPIs of 12–17 months were generally more common among more individuals with more socioeconomically advantaged identities (non-Hispanic White, bachelor's degree or higher, and commercial insurance at birth) and among those in the Midwest census region. These findings are largely consistent with earlier work using the National Vital Statistics System natality data¹ and reveal stable to widening inequities, depending on the indicator.

Our findings also have important implications for health policy, research, and clinical practice. Despite an environment that we hypothesized would have contributed to lower rates of short IPIs in recent years (eg, increased funding for postpartum contraception,^11–14 postpartum Medicaid eligibility extensions,^17–19 COVID-19–related fertility postponement^15,16), we instead found that rates of short IPIs, including very short IPIs, have been relatively stagnant. Further work is needed to understand what policy and practice changes may help enhance the potential of existing policies, as well as inform the development of additional complementary strategies, to address short IPIs, particularly those of the shortest durations (less than 6 months). Furthermore, although it is clear that IPIs less than 6 months present the greatest risk of adverse outcomes for childbearing individuals and newborns, research that sheds light on the mechanisms specifically leading to very short IPIs would be helpful for understanding how to mitigate adverse outcomes while supporting patients in meeting their reproductive goals, particularly when pregnancy is undesired.

Our findings must be interpreted in the context of several limitations. First, we were unable to assess pregnancy intention or desired family size among short IPIs, which is essential to understand in patient counseling. Prior work has demonstrated high rates of unwanted or mistimed pregnancies among very short IPIs (less than 6 months, nearly all),⁹ representing an important target for clinical and public health interventions. The same study revealed lower rates of unwanted or mistimed pregnancies among individuals with IPIs 6–11 and 12–17 months in duration,⁹ which, as we also found in the present study, tended to occur among individuals in older age ranges. Presumably, some birthing people at older ages have short IPIs because of a desire to reach their desired family size while they remain capable of becoming pregnant. Second, the National Vital Statistics System data includes information only on pregnancies ending in live birth, so our results may not generalize to individuals who had a short IPI followed by other pregnancy outcomes (eg, abortion, ectopic pregnancy, stillbirth). Third, we were unable to examine causal mechanisms contributing to short IPIs of different lengths in different groups (eg, inadequate access to contraception or abortion, reproductive coercion, and, again, the desire to balance perceived remaining years of fertility with fertility desires).

This study also has several strengths. We updated estimates of short IPI prevalence and associated racial and ethnic, socioeconomic, and geographic inequities, something that has not been done since 2014. Furthermore, we used data from the National Vital Statistics System, which encompasses all singleton births to multiparous individuals occurring in the United States for each study year. Many prior studies and the Healthy People 2030 surveillance data have used smaller survey samples to measure progress in meeting public health goals to reduce short IPIs^8,29 and are thus susceptible to selection bias (including underrepresentation of certain groups at high risk) and recall bias. Studies with smaller samples have also generally pooled all IPI lengths less than 18 months to increase the precision of their estimates rather than looking at short IPIs of different lengths and risk categories, whereas we were able to consider multiple short IPI durations using the National Vital Statistics System data. In addition, our study disaggregates race and ethnicity to a higher degree than previous studies and considers geographic disparities across the United States.

Using national U.S. birth record data from 2016 to 2022, this study provides important information on contemporary rates of short IPIs. Overall, it appears that little progress was made between 2016 and 2022 in reducing the prevalence of or addressing long-standing racial and ethnic, socioeconomic, or geographic inequities in short IPIs, including IPIs less than 6 months, which are associated with the greatest risk for adverse maternal and infant health outcomes. Future work should further consider pregnancy intendedness, desired family size, unmet need for family planning services, and how to best design clinical and public health interventions that can support individuals in achieving the healthiest possible pregnancies that are aligned with their reproductive goals.