Impact of country income status on the relationship between interpregnancy interval and maternal complications of pregnancy: a systematic review and meta-analysis protocol

Abstract

Background

Both short and long interpregnancy intervals (IPIs) have been associated with adverse maternal outcomes, including uterine rupture, gestational diabetes mellitus (GDM), gestational hypertension and postpartum haemorrhage (PPH). However, the strength and consistency of these relationships remain uncertain, and there is limited information on how these associations differ according to country income status. This review is designed to examine the impact of the income status of a country on the relationship between IPI and maternal complications of pregnancy.

Methods and analysis

CINAHL, MEDLINE, EMBASE via Ovid, Global Health, LILACS and African Index Medicus will be searched using combinations of MeSH and Emtree terms with keywords. Studies included are expected to be observational studies; existing systematic reviews with no new data will be excluded. The study population will consist of women with two or more singleton births and the exposure assessed will be short IPI (<24 months, <18 months and <12 months between the birth of a child and conception of the next pregnancy) and long IPI (>60 months). The comparator group will be an IPI of 24–60 months, as recommended by several international organisations. The co-primary outcomes are preterm pre-labour rupture of membranes, PPH, pre-eclampsia, gestational hypertension, GDM, uterine rupture and maternal mortality. The income status will be categorised according to the World Bank income-based classifications for low-, middle and high-income countries. Two reviewers will independently perform study screening, data extraction and risk of bias assessment using the Newcastle-Ottawa Score. The Grading of Recommendations, Assessment, Development and Evaluations (GRADE) framework will be used to assess the certainty of evidence. A random effects model will be used to calculate pooled ORs with 95% CIs.

Ethics and dissemination

Ethics approval is not applicable for this study as no original data will be collected. The results will be disseminated through peer-reviewed publication and conference presentations.

PROSPERO registration number

CRD420251075283.

Available from: https://www.crd.york.ac.uk/PROSPERO/view/CRD420251075283

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study will use a broad search of multiple databases, including those specific to low- and middle-income countries.

• This study has a detailed analytic plan with sensitivity analyses to address known confounders to the relationship between interpregnancy interval and maternal outcomes.

• The stratification of countries is based on the World Bank classification at the time of patient analysis and recruitment.

• The study does not allow the examination of confounders or mechanisms as individual patient data is not available for the studies in this meta-analysis.

Introduction

Interpregnancy interval (IPI) is defined as the duration between a live birth and the conception of a subsequent pregnancy.¹ It has been increasingly recognised to impact on maternal and perinatal health outcomes, and both short and long IPIs have been associated with increased risks of adverse maternal outcomes. These include gestational diabetes mellitus (GDM), uterine rupture, preterm pre-labour rupture of membranes (PPROM) and PPH, all of which substantially contribute to global maternal morbidity and mortality.² The WHO recommends an optimal IPI of at least 24 months³; in addition, other international organisations recommend a maximum IPI of 60 months to reduce risk of poor maternal and perinatal outcomes.^{4 5}

PPH is defined as blood loss of more than 500 mL following vaginal birth or more than 1000 mL following caesarean section, within 24 hours of birth.⁶ PPH and other obstetric haemorrhage types remain the leading cause of maternal death, especially in low and middle income countries (LMICs⁷). Hypertensive disorders, such as pre-eclampsia (new onset hypertension with proteinuria or evidence of end-organ dysfunction), eclampsia and gestational hypertension (>140/90)^{8 9} remain the second most common cause of maternal death. These obstetric complications, along with maternal infection, delivery complications and unsafe abortions, make up 75% of all maternal deaths globally.¹⁰

The impact of a short IPI is a well-documented risk factor for a variety of adverse perinatal outcomes, and previous reviews have primarily focused on neonatal outcomes such as low birth weight, preterm birth and perinatal mortality.^11–13 While the association between IPI and maternal health outcomes is less consistently reported, Conde-Agudelo et al¹⁴ found that both long and short IPIs increase the risk of adverse maternal outcomes including PPROM, pre-eclampsia and eclampsia. Further studies have reported similar findings,^15–17 alongside increased risks in GDM¹⁸ and PPH.^{19 20} However, the study design of many of these studies has a high chance of bias.²¹

The association between short IPIs and adverse maternal outcomes has been widely observed, but the precise biological mechanisms are not well understood. One mechanism that has been well-recognised is the maternal nutritional depletion hypothesis, which states that mothers with short interpregnancy intervals do not have sufficient time to replenish critical nutrient stores depleted during the previous pregnancy and lactation period, resulting in a subsequent risk of adverse pregnancy outcomes, such as PROM and infection.¹⁴ Other mechanisms that have been theorised to contribute to adverse maternal outcomes include incomplete healing of uterine scars after a caesarean delivery and abnormal remodelling of endometrial blood vessels, which may increase the risk of haemorrhage from uterine rupture or abnormal placentation.^{4 20}

One leading theory behind the increase in adverse maternal outcomes after a long IPI is the physiological regression hypothesis, which suggests that after a long period without pregnancy, the body may lose the growth-supporting capacities that were acquired during previous gestations, such as increased uterine blood flow and the physiological and anatomical adaptations of the reproductive system.¹² Thus, the maternal characteristics would become similar to that of a primigravid woman, and the incidence of obstetric complications (such as pre-eclampsia and gestational hypertension) would subsequently be comparable between the two groups. This hypothesis has been supported by observations that have shown similarity of maternal outcomes for multiparous women after a long IPI and primigravid women.^{17 19}

The income status of a country may affect the impact of IPI on maternal outcomes as women living in low-to-middle income countries are more likely to have reduced access to adequate healthcare provision,^{22 23} including maternal and child healthcare interventions. Women in LMICs are less likely to receive adequate care, including less access to antenatal care²⁴ and skilled birth attendants.²⁵ Currently, maternal mortality disproportionately affects LMICs, with low-income countries accounting for 43.9% of all maternal deaths.¹⁰ Multiple pathways intersect to contribute to the health disparity that is present in LMICs. The ‘Three Maternal Delays’ framework outlines three main barriers to seeking care for mothers in LMICs.²⁶ The first delay is a ‘delay in seeking care’, which occurs when patients are dissuaded from seeking healthcare due to financial and geographic restrictions, cultural beliefs and poor education. Subsequently there is a ‘delay in reaching care’, in which a patient is restricted geographically from hospitals with the available resources and capacity, due to a combination of a larger distance to a hospital and a variability in access to transport, whether it be private or public. Lastly there is a ‘delay in receiving care’, where a patient may not be able to access the care they require due to a lack of resources, essential equipment and services in first-level hospitals. This and other factors, such as food insecurity, malnutrition^{27 28} and contraceptive access, all contribute to the staggering difference in maternal outcomes in LMICs compared with HICs. The maternal nutritional depletion from a short IPI would be further exacerbated by poor nutrition in LMICs, which may contribute to worse maternal outcomes.²⁹ In addition, IPI has been widely recognised as a modifiable determinant of maternal health that can be directly targeted through adequate antenatal care and counselling around pregnancies,^{26 27}which may have a greater effect in LMICs where access, utilisation and quality of antenatal care is much lower.^{30 31} Addressing the maternal delays to healthcare in LMICs, specifically the delay in reaching care, has been shown to significantly decrease rates of maternal mortality,³² which suggests that further focus towards addressing socioeconomic factors may help decrease rates of maternal morbidity and mortality in LMICs.

More recently, Wang et al²¹ conducted an updated systematic review and meta-analysis evaluating the association between short interpregnancy interval and adverse pregnancy outcomes. While comprehensive, this review primarily focused on overall pooled associations and did not examine whether contextual factors such as country income status modify these relationships. In addition, it did not systematically evaluate long interpregnancy intervals across the same range of maternal outcomes included in the present protocol.²¹

To our knowledge, no studies have comprehensively explored the relationship that may exist between IPI and adverse maternal outcomes and the income status of a country. Here we present our protocol for a systematic review and meta-analysis designed to investigate the impact of short and long IPI on maternal pregnancy outcomes and to examine the differential impact of the country income status on this association. This is important as these findings have the potential to inform clinical practice guidelines and public health recommendations on optimal IPI to reduce maternal morbidity and mortality.

Methods

This systematic review protocol has been published with PROSPERO (CRD420251075283), in accordance with the recommendations of Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocol³³ presented in online supplemental material appendix 1.

Patient and public involvement

This protocol is for the systematic review and meta-analysis of observational studies reporting the impact of IPI on adverse maternal outcomes, hence there is no direct patient or public involvement in this research. However, should the results of this study indicate that recommendations should be made, we will consult with patient and public representatives to disseminate the information via conferences and publication of the results, as well as further consultation to optimally disseminate information globally, especially in LMICs.

Eligibility criteria

The eligibility criteria are summarised in table 1.

Table 1. Inclusion and exclusion criteria.

Domain	Inclusion criteria	Exclusion criteria
Population	Women with two or more singleton births in low-, middle- or high-income countries	Studies that focus exclusively on women with previous adverse pregnancy outcomes or those who conceived through assisted reproductive technologies
Intervention/exposure	IPI less than 24 months IPI less than 18 months IPI less than 12 months IPI greater than 60 months	Studies that do not specify IPI or use an alternative IPI definition
Comparator	IPI of 24–60 months	N/A
Outcomes	Adverse maternal outcomes pre-specified – preterm pre-labour rupture of membranes, postpartum haemorrhage, pre-eclampsia, gestational diabetes, uterine rupture and maternal mortality	Studies reporting alternate maternal or pregnancy outcomes
Study design	Observational studies ithat is cohort, case-control, cross-sectional	Case series, case reports, editorials, reviews and non-human studies. Systematic reviews unless they present new data
Language	Studies published in English	Non-English studies unless translation is available
Publication date	Nil restrictions	N/A
Setting	Studies conducted in countries defined clearly by World Bank income classification	N/A

IPI, Interpregnancy interval.

Participants

This review will consider all studies which include women with a singleton pregnancy preceding the index pregnancy. Studies that include multifetal pregnancies will be excluded as they are associated with increased risk of adverse obstetric outcomes. Studies that focus exclusively on women with previous adverse pregnancy outcomes (ie, prior obstetric complications or adverse neonatal outcomes) or those who conceived via assisted reproductive technologies will be excluded to reduce confounding from heterogeneity surrounding risk profile and clinical management.

Exposure

This review will include studies that assess the impact on maternal outcomes of short IPI or long IPI.

1. Short IPI will be defined as: (1) less than 24 months; (2) less than 18 months; and (3) less than 12 months, between the birth of a child to the conception of the next pregnancy. Only studies that report the interpregnancy interval as the time between the previous delivery and conception of the index pregnancy (inter-delivery interval minus gestational age of the index pregnancy) will be included in this review.

2. Long IPI will be defined as greater than 60 months.

Comparator

This systematic review and meta-analysis will use an IPI of 24–60 months as the reference, based on the WHO recommendation for shortest IPI³ and other international guidelines for the longest IPI.^{4 5} The upper bound of 60 months is informed by commonly used thresholds in epidemiological studies and international guideline interpretations, although some guidelines and studies have proposed alternative upper thresholds. The 24–60 month category is therefore used as a pragmatic reference group to ensure comparability across heterogenous studies rather than representing a definitive biological optimum. This approach allows inclusion of a broad range of studies while maintaining consistency in exposure classification for meta-analysis. To compare and assess the impact of short and long IPI in countries of different economic status, studies will be divided by the income status of the country of study participants as per the World Bank’s classification, at the time the study was conducted.³⁴ Countries will be grouped into two categories based on gross national income per capita: low- and middle-income countries versus high-income countries. Where sufficient data are available, pre-specified sensitivity analyses will further stratify studies into three categories: low-income, middle-income (including lower-middle and upper-middle) and high-income countries. Studies will not be excluded solely due to inability to contribute to income-stratified analyses. Instead, they will be included in the primary analysis wherever country classification is possible. For studies conducted over extended periods, country income classification will be based on the World Bank income classification at the midpoint year of data collection. Where a country has changed its income classification during the data collection window for an included study, the World Bank income classification representing the country at the midpoint of data collection will be used. For studies conducted prior to the advent of the World Bank Income classification in 1987, the income classification of the country in 1987 will be used. Subnational studies will be classified according to the national World Bank income classification of the country in which the study was conducted. Studies that cannot be classified or focus exclusively on a specific socioeconomic demographic will be excluded from the study, while studies that span classifications or were conducted prior to 1987 will be excluded in the sensitivity analysis. Where studies report results from multinational populations, data will be extracted and classified separately by country or excluded if this is not possible.

Outcomes

Co-primary outcomes are (1) PPROM (rupture of membranes prior to 37 weeks gestation and before onset of uterine contractions), (2) PPH (blood loss of greater than 500 mL and 1000 mL within 24 hours after birth after a vaginal delivery and caesarean section respectively), (3) pre-eclampsia (persistent hypertension with heavy proteinuria or substantial maternal organ dysfunction), (4) gestational hypertension (persistent hypertension >140/90), (5) gestational diabetes, (6) uterine rupture and (7) maternal mortality (both direct and indirect maternal death).

Types of studies

Observational studies, including cohort studies, cross-sectional reviews and case-control studies, will be eligible for inclusion. Previous meta-analyses and systematic reviews that do not include original data will be excluded from the analysis.

Search strategy

The electronic databases CINAHL, MEDLINE, Embase via Ovid, Global Health, LILACS and African Index Medicus will be searched without restrictions on date, study type or language. The search strategy will combine Medical Subject Headings (MeSH) and Emtree terms with free-text keywords related to the keywords. These terms will be applied to titles, abstracts and full texts where applicable. All studies will be screened for inclusion by title and abstract by two independent reviewers. The full search strategy used is presented in online supplemental material appendix 2.

Data collection and analysis

Study selection

Titles and abstracts identified through all sources will be exported into Rayyan software³⁵ and duplicates removed. Pilot screening will be performed on 10% of articles to ensure that the eligibility criteria are clear and consistent. If there is discordance between reviewers during the pilot screening, minor modifications will be made to the screening strategy for clarity during this phase before screening all of the results. Two authors will independently screen articles identified by the search strategy for inclusion via title, abstract and keywords screening the articles against the pre-specified inclusion and exclusion criteria. In the case of disagreement, the authors will review the papers together. In the case of remaining disagreement, the article will be discussed with a third reviewer to resolve the situation. Data will also be extracted independently by two authors. The data extraction form will also be tested prior to complete extraction to ensure consistency among the reviewers. Discrepancies will be reviewed by a third and fourth author if required. Inter-rater reliability for categorical items (eg, determining if papers were suitable for study inclusion, categorical variables from studies) will be assessed using Cohen’s kappa while continuous variables from studies will be assessed using Interclass Correlation Coefficient. A third (and fourth, if required) reviewer would review items with a reliability measure less than 0.75 which equates to moderate reliability or less.

Data management

Data will be extracted using a standardised electronic data extraction sheet. Once extracted, after reviewer agreement, data will be transferred into Review Manager V.11.1 data-analysis software.³⁶

The following data will be extracted:

• Study characteristics: authors; publication date; study design; country of study; sample size; confounding factors of participants; study size; source of data; data collection timeframe; funding and risk of bias.

• Exposure characteristics: IPI and number of participants in each exposure group.

• Outcomes: incidence, unadjusted or adjusted OR of PPROM, PPH, pre-eclampsia, gestational hypertension, gestational diabetes, uterine rupture and maternal mortality.

Study characteristics will be examined to identify potential overlap between study populations. Where multiple publications use data from the same cohort with overlapping data collection windows, the study with the most comprehensive dataset for each outcome will be prioritised for data extraction. Where different outcomes are reported from the same cohort, relevant data will be extracted without double counting. In cases of heterogeneity in IPI categorisation across studies, reported interval bands (eg, <6, 6–11, 12–17, 18–23, 24–59, etc) will be pooled into the prespecified exposure (≤12, ≤18 and <24) and comparator (≥24) groups where possible. Studies that are unable to be harmonised to our predefined groups will be excluded from the review. Sensitivity analyses will be conducted with the exclusion of studies with broad or non-standard interval groupings to assess the impact of pooling heterogeneous exposure intervals into the prespecified exposure groups.

The analytical hierarchy of this study will prioritise the extraction of the frequency of the outcomes in relation to our prespecified exposures, followed by unadjusted ORs. Where these are unavailable, adjusted ORs will be extracted provided they have included at least maternal age as a confounder. Unadjusted estimates are prioritised to maximise comparability across studies, given that adjusted estimates vary in selection of confounders used. However, given the importance of confounding in this literature, adjusted estimates will also be analysed in a separate subgroup analysis to assess the consistency of findings. A subgroup analysis will also be conducted to examine adjusted and unadjusted ORs independently.

Outcomes and prioritisation

Measures of exposure effect

A meta-analysis will be performed with Review Manager V.11.1.0. Results will be reported using ORs and 95% CIs, with IPI within 24–60 months as the reference group. Forest plots will be used to present the graphical representation of the results and will include derived from pooled odds or risk ratios. Results will be represented in separate forest plots for each outcome and each IPI interval of <12, <18 and <24 months. Random-effects models will be used as the primary analytical approach for all meta-analyses. The Mantel-Haenszel method, with a fixed-effects model, will be used to calculate pooled ORs with 95% CI. Statistical heterogeneity will be assessed using the I² statistic; prediction intervals and sensitivity analyses will be presented where appropriate.

In the event where quantitative synthesis is not appropriate for an outcome (due to a small number of studies or excessive heterogeneity), a narrative synthesis will be performed, along with justification for the lack of quantitative pooling.

Sensitivity analysis

Sensitivity analyses will be conducted for each maternal outcome where there are sufficient studies by removing low and average quality studies to examine their effect on the effect estimates.

Assessment of risk of bias

Where ten or more studies are examined, publication bias will be visually examined for asymmetry of effect size relative to the size of SE using funnel plots, and quantified using the Peters’ regression test.³⁷ Funnel plots will be plotted with effect sizes on the x-axis and standard errors on the y-axis to allow the visual assessment of funnel plot asymmetry. High-, middle- and low-income countries will be assessed separately.

Sensitivity analyses will also be run with and without non-English papers to assess the introduction of bias from translation error.

Two independent authors will assess the risk of bias using the Newcastle-Ottawa Score (NOS) for cohort and case-control studies³⁸ and a modified version of the cohort NOS for cross-sectional studies (online supplemental material appendix 3). This tool seeks to address bias across three domains: exposure and comparator group selection, group comparability, and assessment of the outcome variable. Any variation in judgement between the authors of this study will be discussed and resolved. The NOS is scored out of nine points; studies scoring ≥8 will be considered high quality. The appropriate NOS version will be used depending on the study design, being cross-sectional, cohort or case-control. No studies will be excluded from the primary analyses based on these scores. Sensitivity analyses will be performed to assess the robustness of the results after excluding studies with a score of seven or less.

Consistency and confidence of cumulative evidence in the primary outcome will be assessed independently by two authors using the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) approach.³⁹ Evidence will be assessed for risk of bias, consistency, directness, precision and publication bias. The criteria for each domain have been specified in online supplemental material appendix 4. Sensitivity analyses will be performed with the exclusion of studies of moderate or high risk of bias to assess the robustness of the results.

Project timeline

It is anticipated that manuscripts will be identified for review by 30 August 2026. Data extraction and analyses will be performed between September 2026 and November 2026, while manuscript preparation will be completed between February and April 2027.

Discussion

This systematic review and meta-analysis aim to evaluate the effect of short and long IPI on maternal pregnancy outcomes and to examine the differential impact of the country income status on this association. Specifically, we will generate ORs and 95% CI for seven pregnancy outcomes by IPI with countries grouped by economic status. This data will allow us to determine the optimal IPI that does not increase the odds of each adverse pregnancy outcome for high and low- and middle-income countries which could be used to revise clinical guidelines for best practice. By identifying maternal risks associated with short and long IPI, this review seeks to inform clinicians in providing evidence-based family-planning and preconception counselling, and may inform clinical practice guidelines and public health recommendations on an optimal IPI to reduce maternal morbidity and mortality.

The study has several strengths: first, the meta-analyses will include multiple publications for each obstetric outcome, many of which have large sample sizes. This is likely to result in relatively narrow CIs for each outcome allowing the direct comparison of IPI durations within country developmental status groups. Further, the large number of studies included are likely to be populations from many countries increasing the generalisability of the results. However, there are limitations to our research paradigm, as it does not allow the examination of confounders or mechanisms as individual patient data is not available for the studies in this meta-analysis. Therefore, we will not be able to assess the biological and social mechanisms that mediate the increased risk of pregnancy outcomes in women with short or long IPIs. Certain outcomes, such as uterine rupture and maternal mortality, may have limited available evidence and may be restricted to specific clinical populations, which may reduce statistical power and increase heterogeneity. Furthermore, maternal mortality disproportionately affects those in LMICs¹⁰ and is much rarer in HICs, which may limit the statistical power when comparing the two settings by IPI. There is also likely to be an under-representation of LMIC studies in major databases. This limitation is mitigated by the inclusion of studies from databases that have a higher portion of papers from LMICs. Additionally, we anticipate heterogeneity regarding outcome definitions. In order to avoid the exclusion of studies from LMICs, we will use all studies that have a desired outcome listed regardless of their definition and perform secondary analyses with studies that do not define outcomes or have varying definitions from those listed in our methodology. A further limitation of this review relates to potential selection bias when comparing IPI effects across country income settings. In low- and middle-income countries, women who achieve an IPI within the reference category (24–60 months) may differ systematically from those with shorter or longer intervals in terms of socioeconomic status, access to contraception, fertility intentions and underlying health status. Similar to ‘healthy worker’ and ‘healthy survivor’ effects described in epidemiological studies, this may result in a non-representative comparison group within and across countries. As a result, observed differences in effect estimates by income classification may reflect underlying population structure and access to reproductive health services, rather than true biological differences in the impact of IPI on maternal outcomes.

Determining the optimal IPI for countries based on economic status may provide important information to support the development of public health strategies aimed at reducing adverse maternal pregnancy outcomes, including PPROM, PPH, pre-eclampsia, gestational hypertension, gestational diabetes, uterine rupture and maternal mortality. By synthesising evidence from multiple studies, this review has the potential to provide evidence to support health authorities when developing country-specific guidance on optimal IPI. Ultimately, the findings may contribute to improved understanding of how interpregnancy interval influences maternal outcomes across different economic settings. We anticipate that the results could be useful for policymakers, healthcare providers, women, families and the broader community by supporting education and family planning decisions in countries of varying income levels.