Interpregnancy weight change and risks of stillbirth and infant mortality: a protocol of a systematic review and meta-analysis

STRENGTHS AND LIMITATIONS OF THIS STUDY

  • This protocol follows the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols guidelines.

  • A comprehensive search strategy has been developed to include all eligible studies meeting the inclusion criteria.

  • The study screening, selection, data extraction and assessment of the risk of bias will be completed by two independent reviewers.

  • Both the risk of bias and the quality of evidence will be assessed.

  • The exclusion of non-English databases could lead to language bias.

Introduction

Overweight/obesity is usually defined as having a body mass index (BMI, weight in kg/height in m2) greater than or equal to 25/30, affecting approximately 30% of women of reproductive age wordwide.1 2 It is well established that being overweight or obese is a significant risk factor for various adverse perinatal outcomes including stillbirth and infant mortality.3–5

In recent years, a couple of observational studies have suggested that there may be an association between interpregnancy weight change and risks of stillbirth and infant mortality in subsequent pregnancy.6 7 Interpregnancy weight change is defined as the difference between maternal prepregnancy BMI/weight at an index pregnancy and a successive pregnancy.8

As interpregnancy weight change is one of the few modifiable risk factors in prenatal care, synthesised evidence on the effect of interpregnancy weight change on numerous adverse maternal and neonatal outcomes is greatly needed. Several systematic reviews and meta-analyses have summarised the effect of interpregnancy weight change on gestational diabetes),8–10 large-for-gestational-age neonates,8–10 caesarean delivery,8 9 preterm birth,9 10 pre-eclampsia9–11 and pregnancy-induced hypertension.9–11 However, to our knowledge, no systematic reviews and meta-analyses have been conducted to assess the association of interpregnancy weight change with risks of stillbirth and infant mortality. To fill this gap in the literature, we undertake the initiative to synthesise currently available evidence on this association.

Objectives

This systematic review and meta-analysis has three objectives: (1) to estimate the association of interpregnancy weight change with the risks of stillbirth and infant mortality at subsequent pregnancy; (2) to examine the dose–response association between interpregnancy weight change and risks of stillbirth and infant mortality and (3) to examine the association between interpregnancy weight change and risks of stillbirth and infant mortality in women stratified by BMI category of the index pregnancy.

Methods and analysis

The protocol is developed in line with the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-analyses Protocols (PRISMA) 2015.12 The planned date for this study started on 15 October 2023 and anticipated end date is 31 December 2023.

Eligibility criteria

Types of studies

A comprehensive literature search will be conducted on a total of four online databases to gather papers published from inception to 31 October 2023: OVID Embase, Cochrane Libraries, Web of Science and OVID Medline. All study designs will be included to gather as much literature as possible. This will include observational and randomised controlled studies and be limited to human pregnancy. Case reports, book chapters, conference proceedings, abstracts and narrative reviews will be excluded. Furthermore, only English studies will be reviewed due to the challenges associated with reviewing studies written in another language.

Types of participants

The participants will be multiparous women who have given at least two consecutive singleton births within the study period.

Interventions and comparisons

To be eligible, studies must report quantitative measures for interpregnancy weight change interpregnancy weight change is defined as the difference between the prepregnancy BMI/weight of an index pregnancy and a successive pregnancy. This is typically reported as the change in prepregnancy weight (kg) and/or BMI (kg/m2) between the consecutive pregnancies of multiparous women. Both pregnancies may be separated by any time interval. Considering interpregnancy interval might be an independent risk factor for subsequent stillbirth,5 we will only include studies which provided both information of interpregnancy weight change and interpregnancy interval for this review. Weight change of ≥1 BMI unit between consequtive pregnancies will be considered as significant interprengnacy weight change. Women who are stable in their prepregnancy weight/BMI between consecutive pregnancies (−1≤BMI unit change <+1) will be used as the comparison group.

Outcomes and prioritisation

Primary outcomes

Stillbirth and infant mortality at subsequent pregnancy are the primary outcomes of interest for this review. Stillbirth is defined as intrauterine fetal death that typically occurs at the 20th week of gestation or later.13 This outcome must be based on reliable and accurate records. Infant mortality is defined as the death of infants (who are below 1 year of age).14 This outcome must also be based on reliable and accurate records.

Secondary outcomes

  1. Antepartum stillbirth: fetal death occurring after the 16th week of gestation and before labour onset.15

  2. Intrapartum stillbirth: death of a fetus occurring at ≥20 completed weeks of gestation and during labour.16

  3. Neonatal death: deaths among live births during the first 28 days of life.17

  4. Postneonatal death: deaths among live births from 29 days to 1 year of life.18

  5. Perinatal death: fetal death occurring at 20 weeks or more and neaontal death within 28 days following a live birth.19

Search strategy

The search strategy is developed to capture all literature relevant to the risks of stillbirth and infant mortality as a result of interpregnancy weight change. A medical librarian from the Ottawa Hospital Research Institute was consulted to identify the databases of interest and to aid in refining the search strategy. Medical subject headings and search terms related to “interpregnancy”, “between pregnancies” were used in combination with “weight change/gain/loss” and the outcomes of interest. To moderate the sensitivity and broadness of the search, commands including Boolean operators and truncation are used. Sample search strategies are included in online supplemental appendix A.

Supplemental material

Information sources

A comprehensive literature search will be conducted on a total of four online databases to gather studies published from inception to October 2023: Embase, Cochrane Libraries, Web of Science and Medline. Initial test searches were conducted on 1 October 2023. Data collection and screening will begin on 31 October 2023. A final updated search will be conducted prior to publication. The reference lists of the captured studies will be reviewed for further potentially relevant papers.

Study records

Selection process

Search results from the electronic databases will be imported into Mendeley, where duplicates will be removed. Two reviewers (YT and RL) will independently screen the titles and abstracts of potentially eligible studies. Eligibility will be assessed based on the predefined inclusion and exclusion criteria. The reviewers will then examine the full texts of the articles to ensure that they meet the inclusion criteria. To manage these records, online Covidence software will be used. Discrepancies will be managed by consensus between reviewers or through meeting with a third reviewer (YG). The selection process will be outlined in a PRISMA diagram.

Data collection process and management

To ensure consistency among reviewers prior to beginning the review, a calibration assessment will be conducted. Two reviewers (YT and RL) will extract data into standardised forms independently on Covidence, each of which will be inspected by one another for completeness and accuracy. Discrepancies between reviewers will be managed by consensus, or through meeting with a third reviewer (YG).

Data extraction

Two reviewers (YT and RL) will independently extract the following data from the identified studies: (1) Publication details: title, authors, publication year, journal, (2) Study details: aim, design, inclusion and exclusion criteria, setting, study period (3) Patient characteristics: sample size and characteristics, number included in analysis (4) Intervention information: interpregnancy BMI/weight change, units used for weight measurement, reference group characterisation; (5) Outcome: primary and secondary outcomes, and diagnostic criteria and (6) Stratified/subgroup analysis: by BMI category at index pregnancy, interpregnancy interval and time point. If needed, the authors of the original papers will be contacted for further information and to clarify ambiguities.

Risk of bias assessment

Two authors (YT and RL) will independently assess the risk of bias for the identified studies. The Cochrane Risk of Bias tool will be used to assess the risk of bias in the randomised controlled studies. Using this tool, we will assess studies based on several criteria: random sequence generation, blinding of outcome assessment, blinding of participants and personnel, selective reporting, incomplete outcome and other bias.20 The Risk of Bias in Non-Randomised Studies of Interventions tool will be referred to for observational studies.21 Disagreements between reviewers will be resolved by consensus or through consultation with a third author (YG). The level of bias (high, low and unclear) will be graded for each important outcome and study and then be presented graphically in a table.

Data synthesis

If sufficient data are available to evaluate a common outcome and the studies are comparable with regard to design, methodology and intervention, a meta-analysis will be conducted to acquire estimates of the pooled effect. If there is a sufficient number of both observational studies and randomised controlled studies, analyses will be done separately for each design. Furthermore, observational studies with different methodologies (eg, BMI unit change or kilogram change or weight percentage change; risk ratio vs HR) will be separately analysed. To express the estimate of the effect for dichotomous outcomes, we will use risk ratios (or HRs) with 95% CIs.

If the statistical heterogeneity is found to be low (p>0.10 or I2<50%), the fixed-effect model will be used to pool the data. If not, a more reasonable estimate of the effect will be attained using the random-effect model. Review Manager V.5.3 will be used to carry out the analyses. In case there are limited data available for an outcome, we will provide a narrative summary of the individual studies’ findings instead.

Subgroup analysis and investigation of heterogeneity

If there ae sufficient data available, the following subgroup analyses will be performed to evaluate heterogeneity: age at follow-up (eg, neonatal death (0–28 days), postneonatal death (1–12 months)), prepregnancy BMI at index pregnancy (BMI<25 and BMI≥25 kg/m2), parity, interpregnancy time interval and various definitions of stillbirth (≥20 weeks or 22 weeks or 24 weeks). The exposure effect will be assessed using the χ2 test and 95% CIs will be used. We will use the I² heterogeneity test to evaluate heterogeneity for the study outcomes. An obtained I² value of greater than 50% or a p<0.10 would indicate there is significant statistical heterogeneity.

Dose-response association analysis

We will perform a restricted cubic spline regression model to assess the dose–response relationship between interpregnancy weight change and risks of stillbirth and infant mortality if sufficient data are available.

Meta-biases

In the case there are 10 or more studies assessing 1 common outcome, funnel plots will be generated to assess publication bias.

Confidence in cumulative evidence

The Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) approach will be applied to evaluate the important outcomes. Two reviewers (YT and RL) will independently evaluate the quality of evidence on five aspects: inconsistency, imprecision, indirectness, risk of bias and other considerations. The gradings for the quality will be identified as high, moderate, low or very low and organised into a table. Disagreements between reviewers will be resolved by consensus or through consultation with a third reviewer (YG).

Patient and public involvement

We did not require patient or public involvement in drafting the protocol.

Ethics and dissemination

Ethics approval is not required for this study as only data from published papers will be used. The findings from this systematic review and meta-analysis will be submitted for publication in a peer-reviewed scientific journal and for presentation at academic conferences.

Ethics statements

Patient consent for publication

Acknowledgments

We would like to thank Risa Shorr for helping to review the search strategy and for providing guidance on appropriate databases to use for the review.

This post was originally published on https://bmjopen.bmj.com