Association between maternal multimorbidity and preterm birth: a systematic review and meta-analysis

Abstract

Background

Preterm birth (PTB) remains a leading cause of neonatal morbidity and mortality worldwide. While conditions such as diabetes and hypertension have been linked to PTB, the contribution of multimorbidity, defined as the coexistence of two or more chronic conditions, has not been extensively quantified. This systematic review and meta-analysis evaluated the association between maternal multimorbidity and PTB risk.

Methods

This systematic review and meta-analysis followed PRISMA 2020 guidelines and was prospectively registered in PROSPERO (CRD420251086046). PubMed, Scopus, Web of Science, and EBSCOhost were searched from database inception to November 2025 for observational studies examining the association between multimorbidity and PTB. Eligible studies included pregnant women with two or more chronic conditions compared with those without any chronic conditions. Study quality was assessed using the Newcastle–Ottawa Scale (NOS). Adjusted risk ratios (RRs) with 95% confidence intervals (CIs) were pooled using random-effects models, using retrieved RRs or converted Odds Ratios (ORs) to RRs. Analysis was performed separately for different gestational age thresholds (< 37 weeks and < 28 weeks). Further subgroup analyses were also conducted based on the type of PTB, study design, sample size, geographical region, risk of bias, multimorbidity severity, timing of diagnosis and multimorbidity category. Heterogeneity was evaluated with the I² statistic, and publication bias with funnel plots and Egger’s test. Sensitivity analyses were performed for gestational age cut-offs and risk of bias using leave-one-out procedures and alternative estimators of between-study variance (REML, PM, and DL).

Results

Thirteen cohort studies including over 3 million pregnancies met the inclusion criteria. Maternal multimorbidity was significantly associated with an increased risk of PTB (RR = 2.04, 95% CI 1.61–2.59; p < 0.001). The association remained significant for births before 37 weeks (RR = 2.04, 95% CI 1.61–2.59). Significant associations were also observed for spontaneous PTB (RR = 1.46, 95% CI 1.13–1.89), in both prospective (RR = 1.70, 95% CI 1.29–2.24) and retrospective cohort studies (RR = 2.17, 95% CI 1.44–3.26), and across small (RR = 1.67, 95% CI 1.10–2.54), medium (RR = 1.56, 95% CI 1.29–1.90), and large (RR = 2.69, 95% CI 1.55–4.68) sample sizes. Likewise significant associations were identified in studies defining multimorbidity with higher clinical severity (RR = 2.44, 95% CI 1.33–4.47) and in those with moderate risk of bias (RR = 2.32, 95% CI 1.63–3.31). In addition, studies focusing on mental health-related multimorbidity showed a significant association with preterm birth (RR = 1.92, 95% CI 1.64–2.25), as did studies examining metabolic/cardiovascular multimorbidity (RR = 2.99, 95% CI 1.29–6.91) and those assessing mixed multimorbidity profiles (RR = 1.55, 95% CI 1.12–2.17). Sensitivity analyses, including leave-one-out procedures and alternative estimators (REML, PM, DL), confirmed the robustness of results. Although heterogeneity remained high in the analyses (I² > 80%), there was no evidence of publication bias based on funnel plots or Egger’s test.

Conclusion

Maternal multimorbidity is associated with a significantly increased risk of PTB, highlighting the importance of comprehensive preconception and antenatal care for women with multiple chronic conditions. However, the findings should be interpreted with caution due to substantial heterogeneity, variation in multimorbidity definitions, and the observational design of the included studies. Further large-scale, high-quality prospective research using standardized definitions of multimorbidity and PTB is needed to strengthen the evidence base and clarify causal pathways.

Registration

PROSPERO: CRD420251086046.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12884-026-08734-w.

Background

Preterm birth (PTB), defined by the World Health Organization as delivery before 37 completed weeks of gestation, remains a substantial global public health concern [1]. In 2020, an estimated 13.4 million infants, accounting for more than one in ten live births worldwide, were born preterm, with prevalence ranging from 4% to 16% across countries [2]. PTB is the leading cause of mortality among children under five years, accounting for approximately 740,000 deaths in 2021 [3]. Survivors face increased risks of long-term adverse outcomes, including neurodevelopmental impairments, chronic diseases, and disability [4].

Established risk factors include extremes of maternal age, pre-existing chronic conditions such as hypertension and diabetes, maternal infections, and socioeconomic disadvantage [5]. While these factors have been widely studied individually, increasing attention has been directed toward the combined impact of multiple chronic conditions as a potential driver of adverse pregnancy outcomes, including PTB [6–8]. These findings suggest that maternal health cannot be fully understood by examining single conditions in isolation.

Accordingly, the concept of multimorbidity in pregnancy, defined as the co-occurrence of two or more chronic conditions in which no single condition is considered primarily, often referred to as multiple long-term conditions (MLTC) [9], has emerged as an increasingly important public health concern [10]. Its prevalence appears to be rising globally, paralleling the increase in chronic diseases among women of reproductive age. Reported estimates vary widely across populations, reflecting differences in definitions, study designs, and methods of case ascertainment [10]. Common multimorbidity patterns include combinations of cardiometabolic disorders, mental health conditions, and infectious diseases, each independently associated with adverse pregnancy outcomes [10, 11].

Several biological and clinical mechanisms may underlie the association between multimorbidity and PTB. The coexistence of multiple chronic conditions may exert synergistic detrimental effects on placental development and function, promote systemic and local inflammation, and impair uteroplacental perfusion [12, 13]. Multimorbidity is also associated with an elevated risk of pregnancy complications including hypertensive disorders, gestational diabetes, and maternal infections, each of which is independently linked to increased PTB risk [14, 15]. Furthermore, the management of multiple chronic conditions often necessitates polypharmacy, may fragment continuity of care, and can reduce maternal physiological reserve, thereby amplifying vulnerability to adverse perinatal outcomes [16].

Although several observational studies have investigated the relationship between multimorbidity and PTB, their findings are partly inconsistent due to methodological heterogeneity and differences in case ascertainment and outcome definitions. Definitions of multimorbidity in the primary literature also vary, with some studies applying a broad definition of multimorbidity, encompassing any two or more chronic conditions, while others focus on specific combinations such as cardiometabolic or mental health disorders [7–18]. Similarly, PTB is variably classified as overall (< 37 weeks), by gestational age subcategories (< 34 weeks, < 28 weeks), or by clinical subtype (spontaneous versus medically indicated) [19].

To date, only one systematic review and meta-analysis has quantitively synthesised this association. In essence, this systematic review and meta-analysis reported that maternal multimorbidity was associated with a more than fourfold increased odds of PTB (OR ≈ 4.28, 95% CI: 2.23–6.34) and substantially elevated risks of other adverse birth outcomes [8]. However, this review examined adverse pregnancy outcomes in general, using a broader search strategy that included all birth-related outcomes rather than focusing specifically on PTB. In that review, PTB was analysed as one of several outcomes, without PTB-specific subgroup analyses based on gestational age cut-offs, multimorbidity severity, timing of diagnosis, or other relevant study characteristics. Moreover, multimorbidity was defined broadly to include both pre-existing and gestational conditions, which may have inflated the observed effect estimates and limited comparability across studies. In contrast, the present systematic review focuses specifically on chronic pre-existing conditions applying a more restrictive and clinically relevant definition of multimorbidity. This approach allows for clearer assessment of the underlying chronic disease burden before or during pregnancy, excluding transient gestational complications.

Despite the growing prevalence of multimorbidity among women of reproductive age and its potential impact on adverse pregnancy outcomes, there is currently limited comprehensive synthesis of the evidence quantifying its association with PTB. Previous studies have not systematically examined whether the strength of association varies by type of multimorbidity, nor have they consistently accounted for potential confounding factors. A pooled analysis applying harmonised definitions is needed to provide more precise estimates and to clarify whether certain multimorbidity patterns confer a higher risk of PTB. Such evidence could guide risk stratification, antenatal care planning, and targeted interventions.

The purpose of this systematic review and meta-analysis was to synthesise evidence on the association between maternal multimorbidity and the risk of PTB. We aimed to quantify the overall effect of multimorbidity on PTB < 37 weeks and to assess how this association varies across different clinical and methodological contexts. To achieve this, we explored differences in effect estimates across gestational age cut-offs, with further subgroup analysis based on the type of PTB, study design, sample size, geographical region, risk of bias, timing of multimorbidity ascertainment, multimorbidity severity, and multimorbidity category. By providing a harmonised and focused synthesis of PTB-specific outcomes, this study contributes novel evidence on the impact of chronic pre-existing conditions in pregnancy and addresses gaps not examined in broader reviews of adverse birth outcomes.

Methods

This systematic review and meta-analysis was reported in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) 2020 guidelines [20] for systematic reviews and meta-analyses (Supplemental material 1) and was prospectively registered with PROSPERO (CRD420251086046). It was designed a priori to define methods for searching terms, assessing the quality of included studies, collecting, extracting, and analysing data in the review protocol.

Search strategy and study selection

Two researchers (LM and KG) independently searched PubMed, Scopus, Web of Science, and EBSCOhost databases to identify potentially eligible articles that examine the association between multimorbidity and PTB. Discrepancies were resolved by mediation and discussion with a third author (DL). The literature searches covered all publication years, from databases inception to November 8, 2025, and only articles published in English were considered. The search followed the Population, Exposure, Comparison, Outcome (PECO) framework and included a combination of Medical Subject Headings (MeSH), keywords, and synonyms. Search terms covered concepts related to multimorbidity or Multiple long-term conditions (MLTC), comorbid conditions, and adverse birth outcomes, including PTB. Detailed, line-by-line search strategy for each database is provided in Supplemental material 2. For all articles identified through the search, the titles and abstracts were examined by the two researchers (LM and KG), and then the full texts of potentially eligible articles were reviewed for evaluation. Reference lists of relevant primary studies and reviews were manually screened to identify any additional eligible publications. Grey literature sources, such as dissertations, theses, and unpublished manuscripts, were not included in this study.

Eligibility criteria

We used the PECO approach to identify included studies. Participants were pregnant women of any age or ethnicity, with or without multimorbidity. The exposure of interest was multimorbidity during pregnancy, and the comparator group comprised pregnant women without multimorbidity. Multimorbidity was defined as the coexistence of two or more chronic health conditions in a single individual, none of which took clinical precedence over the others. Studies were included if they applied a broad definition of multimorbidity (i.e., the presence of two or more chronic conditions of any type) or if they examined specific combinations of chronic conditions. This distinction was made to account for conceptual heterogeneity in how multimorbidity is defined across studies. Broad definitions capture the overall burden of chronic disease, whereas analyses based on specific condition combinations allow for exploration of clinically meaningful clusters and shared pathophysiological pathways [21, 22]. Chronic conditions included, but were not limited to, hypertension, diabetes mellitus, obesity, cardiovascular disease, respiratory disease, kidney diseases, mental health disorders such as depression, anxiety, schizophrenia and bipolar disorder, autoimmune disorders, thyroid disease, and chronic infectious diseases (e.g., HIV or hepatitis). The primary outcome was PTB, defined as delivery before 37 completed weeks of gestation. Only observational studies (cohort, case-control, or cross-sectional designs) were eligible for inclusion. Randomized controlled trials were excluded due to their limited suitability for assessing naturally occurring associations between multimorbidity and PTB. Additional exclusion criteria included: non-peer-reviewed literature (e.g., conference abstracts, editorials, letters), studies limited to a single condition unrelated to multimorbidity, studies focusing exclusively on gestational conditions, studies without clearly defined exposure or outcome, and studies not reporting PTB as an outcome. Studies that did not provide effect estimates, comparing women with and without multimorbidity were also excluded from the meta-analysis.

Data extraction

Data were independently extracted by LM and KG using a standardized, pretested form. Any discrepancy was resolved after consultation with a third author (DL). Extracted information included study-level characteristics (author, year, design, country, sample size), participant demographics (age, ethnicity), and exposure details including the definitions of multimorbidity, the conditions examined, the timing of condition ascertainment (pre-conception, during pregnancy), the severity of chronic conditions (severity of multimorbidity) and multimorbidity burden (i.e. number of coexisting chronic conditions). Multimorbidity definitions were categorised as broad, referring to the presence of two or more chronic conditions of any type (physical, mental, infectious, or mixed), or specific referring to particular combinations of chronic conditions.

Each chronic condition identified was classified by severity using a structured approach developed for this systematic review based on its expected impact on maternal and pregnancy health. In particular, chronic conditions were categorized as mild, moderate, or severe according to their clinical impact, management requirements, and potential for pregnancy complications. Mild conditions were those typically managed in primary care, rarely requiring hospitalisation and with minimal impact on pregnancy outcomes. Moderate conditions required specialist follow-up or medication with potential pregnancy risks and may influence pregnancy outcomes. Severe conditions involved organ damage, hospitalisation, or intensive monitoring and were associated with substantial pregnancy risk. Where possible, classification was cross-checked against published comorbidity indices (Obstetric Comorbidity Index) and reviewed independently by the two researchers.

Outcome measures included PTB and its subcategories either by gestational age cut-offs (< 34 weeks, < 28 weeks) or by type (spontaneous and medically indicated). Confounders adjusted for were also extracted. We considered key maternal and sociodemographic factors, including maternal age, parity, smoking, pre-pregnancy BMI, socioeconomic status or education, and race/ethnicity, to be important confounders of the association between multimorbidity and PTB [1]. When studies reported multiple models, the most fully adjusted estimate was extracted. Variation in the set of adjusted confounders across studies was considered during data synthesis and in the assessment of methodological quality, which accounts for the adequacy of confounder adjustment. Effect estimates (Odds Ratio; OR, Risk ratio RR, 95% Confidence Intervals; CIs) were also extracted from each study.

Assessment of methodological quality

The methodological quality of included studies was evaluated using the Newcastle-Ottawa Scale (NOS) for cohort and case-control studies [23]. Domains assessed included participant selection, group comparability, outcome measurement, and statistical methodology. Each study was categorized as having low, moderate, or high risk of bias based on commonly used thresholds used in the literature [24]. Assessments were conducted independently by LM and KG, with disagreements resolved through discussion.

Statistical analysis

A narrative synthesis was undertaken to describe study characteristics, methodological quality, and key findings across studies. Where studies were sufficiently comparable in terms of design, population, exposure definitions, and outcome measures, a quantitative synthesis was performed through meta-analysis. Reported adjusted effect size estimates on the association between multimorbidity and PTB from each included study were pooled using meta-analysis. Effect estimates were expressed as RR with their corresponding 95% CI, which served as the standard measure of association.

For the primary analysis, only adjusted estimates for overall PTB (< 37 weeks) were included. ORs reported in the original studies were converted to RRs using standard transformation formulas to ensure consistency across effect measures [25] (Supplemental Material 3). When a study reported multiple adjusted effect estimates from the same population, these were treated as correlated repeated measures rather than independent observations to avoid double-counting participants. In particular, a single overall adjusted estimate per study was calculated by combining the results, using inverse-variance weighting, ensuring that each study contributed only one independent effect size to the meta-analysis (Supplemental Material 4). Both aRRs and converted RRs were accepted, and no unadjusted estimates were used. Effect measures were log-transformed, and standard errors were derived from the reported CI.

Heterogeneity across studies was assessed using the χ²-based Cochran Q test, with a p-value < 0.10 considered indicative of statistical heterogeneity, and was quantified using the I² statistic, with values greater than 50% considered to indicate substantial heterogeneity [26]. A random-effects model was applied when there was evidence of clinical and methodological heterogeneity among studies, whereas a fixed-effect model was used when studies were considered sufficiently similar in design, population characteristics, and measurement of exposure and outcome [27]. To explore potential sources of heterogeneity, a series of subgroup analyses were performed. Studies were stratified by gestational age (< 37 weeks and < 28 weeks), according to the type of PTB (medically-indicated PTB and spontaneous PTB), sample size (small: <10 000, medium: 10 000 - <100 000 and large: ≥100 000), study design (prospective cohorts and retrospective cohorts), timing of multimorbidity ascertainment (preconception and during pregnancy), study quality (risk of bias), geographical region, multimorbidity severity (number and type of coexisting chronic conditions), and multimorbidity categories (Mental Health multimorbidity, Metabolic/Cardiovascular multimorbidity and Mixed multimorbidity). Forest plots were generated for each subgroup to illustrate the pooled estimates and to visualise between-study variation. Subgroup analyses based on definition of multimorbidity (specific conditions vs. broad definition) were not performed as most studies applied similar criteria for defining multimorbidity.

Publication bias was assessed visually using funnel plots and statistically using Egger’s regression asymmetry test [28, 29]. Sensitivity analyses were performed to evaluate the robustness of the meta-analysis results. We conducted leave-one-out analyses, in which the meta-analysis was repeated after sequentially removing each study to assess whether the overall effect was disproportionately influenced by any single study. In addition, we applied alternative estimators for between-study variance (τ²), including restricted maximum likelihood (REML), Paule–Mandel (PM), and DerSimonian–Laird (DL) to determine the stability of the pooled estimates across different statistical assumptions [30]. Additional sensitivity analyses were conducted according to the risk of bias level. The meta-analysis was repeated including only studies with low risk of bias and, separately, only those with moderate risk of bias, to examine whether study quality influenced the pooled estimates. Consistency of results across these approaches was taken as evidence of robustness. All statistical analyses were performed in R (version 4.5.1) using the meta package, with statistical significance set at a two-tailed p-value < 0.05.

Assessment of evidence in cumulative evidence

The certainty of evidence for each main outcome (PTB < 37 weeks and < 28 weeks) was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) criteria [31], adapted for non-randomised (observational) studies. Because all included studies were cohorts, the starting level of certainty was rated as “low”. Evidence was downgraded for serious limitations related to the risk of bias, inconsistency, indirectness, imprecision or publication bias, and upgraded when there was evidence of a large magnitude of effect sizes, a dose-response relationship, or when residual confounding was likely to reduce the observed effect [32].

Results

Study selection

A total of 3 622 records were identified through electronic database searching, and additional 7 studies were retrieved from manual reference screening. After removal of duplicates, 3 520 unique records were screened by title and abstract. Of these, 3 487 were excluded for not meeting the eligibility criteria. A further 23 full-text articles were excluded following detailed review. Ultimately, 13 studies [6, 7, 33–43] met the inclusion criteria and were included in the systematic review and meta-analysis. The process of study selection is illustrated in the PRISMA flow diagram (Fig. 1).

Fig. 1.

PRISMA 2020 flow diagram of study selection for the systematic review and meta-analysis

Study characteristics

The 13 included studies were published between 2011 and 2025 and represented diverse geographic regions including Canada, United Kingdom, China and other European countries. Study designs comprised 13 cohorts, of which 5 were prospective and 8 retrospective and enrolled sample sizes ranging from 989 to approximately 3 million participants. Multimorbidity was most frequently defined as the presence of two or more chronic conditions of any type, although several studies examined specific combinations such as diabetes with hypertension or depression with anxiety. PTB was variably defined, with most studies reporting overall PTB before 37 completed weeks of gestation, while others reported PTB before 34 weeks and before 38 weeks or presented separate estimates for spontaneous and medically indicated PTB. Most studies focused exclusively on singleton pregnancies, while those that included multiple pregnancies accounted for them as an adjustment factor in their analyses. A detailed summary of the included studies is presented in Tables 1 and 2.

Table 1.

Characteristics of studies included in the systematic review and meta-analysis on multimorbidity and preterm birth

Author, Year	Country	Study Design	Sample Size	Mean/Median Maternal Age or Age Distribution	Maternal Ethnicity	Ascertainment timing
Al Khalaf et al., ​ 2021 [17, 43]	Sweden	Retrospective cohort	1 420 846 deliveries	Mean 28.8 yrs (overall)	83% Swedish, 2.9% Scandinavian, 14% Other	First antenatal visit
Aubry et al., ​ 2019 [39]	Switzerland	Retrospective cohort	324 664 deliveries	Mean 31.1 yrs (overall)	63.1% Swiss, 28.2% European, 8.7% Non-European	During pregnancy
Azcoaga-Lorenzo et al., ​ 2023 [7]	Scotland	Retrospective cohort	30 557 deliveries (27 771 women)	15–19 yrs: 6.4%; 20–24 yrs: 18.2%; 25–29 yrs: 29.8%; 30–34 yrs: 28.8%; ≥35 yrs: 16.8%	White: 75% ​, Mixed race: 0.2% ​, Black: 0.5% ​, Asian: 2.1%​, Others: 1.5%, Missing: 20.8%	Pre-conception
Bandoli and Chambers., 2017 ​[36]	United States and Canada	Prospective cohort	3 034 pregnant women	Mean: 32.5 yrs (RA), 31.5 yrs (Crohn’s), 32.6 yrs (Psoriasis), 32.1 yrs (none)	RA 77.5% NH White; Crohn’s 91.7% NH White; Psoriasis 83.3% NH White; No conditions 72.8% NH White	Not specified
Berger et al., ​ 2020 [34]	Canada	Retrospective cohort	506 483 pregnancies	NR	NR	Pre-conception
Brown et al., 2016 [44]	Canada	Retrospective cohort	3932 deliveries	< 20 yrs 12.9% and ≥ 35 yrs 11.7% (dual diagnosis); <20 yrs 10.1% and ≥ 35 yrs 13.9% (one disease)	NR	Pre-conception
Cripe et al., 2011 [35]	United States	Prospective cohort	3432 pregnant women	Mean: 32.6 yrs (no mood/migraine), 33.7 yrs (mood only), 32.4 yrs (migraine only), 32.4 yrs (mood + migraine)	No disorder 85.3% NH White; Mood only 94.7% NH White; Migraine only 87.5% NH White; Mood + migraine 95.9% NH White	Not specified
Ibanez et al., ​ 2012 [40]	France	Prospective cohort	2002 pregnant women	Age distribution: <25 yrs 14.5–19.2%; 25–35 yrs 63.4–70.3%; ≥35 yrs 14.8–18.1% (by mental health status)	NR	During pregnancy
Lehloa et al., 2025 [45]	South Africa	Prospective cohort	989 pregnant women	Median: 29 yrs	NR	First antenatal visit
Li et al., ​ 2020 [33]	China	Retrospective cohort	13 198 pregnant women	Mean: 27.6 yrs	Yi 99.1%; Han 0.9%	During pregnancy
Männistö et al., 2016 [42]	United States	Retrospective cohort	223 394 deliveries (203,999 women)​	Age distribution: <20 yrs 9.2–10.7%; 20–24 yrs 25.1–28.4%; 25–29 yrs 27.2–27.9%; 30–34 yrs 19.7–22.7%; ≥35 yrs 13.9–15.1% disorders.	NH White 48.3–64.1%; NH Black 21.0–22.6%; Hispanic 10.4–18.0%; Asian/Pacific Islander 0.7–4.4%; Other 1.3–2.5%; Unknown 2.6–4.2%	During pregnancy
Nakanishi et al., ​ 2023 [6, 46]	Japan	Prospective cohort	86 885 pregnant women	Median: 31 yrs	NR	During pregnancy
Yanit et al., ​ 2012 [37]	United States	Retrospective cohort	532 088 deliveries	Mean: 27.9 yrs (no disease), 32.1 yrs (hypertension), 31.2 yrs (diabetes), 33.9 yrs (both)	No disease: White 33.6%; Hispanic 47.2%; African American 5.0%; Asian 11.5%; Other 2.8%. Hypertension only: White 35.7%; Hispanic 34.1%; African American 14.8%; Asian 11.9%; Other 3.6%. Diabetes only: White 26.6%; Hispanic 53.7%; African American 5.9%; Asian 10.9%; Other 3.0%. Both: White 23.9%; Hispanic 46.9%; African American 12.5%; Asian 13.4%; Other 3.2%	Not specified

Abbreviations: NR Not reported, yrs years, NH Non-Hispanic, RA Rheumatoid arthritis

Table 2.

Characteristics of included studies examining multimorbidity and preterm birth: definitions, adjustments, and main findings

Author, Year	Chronic Conditions	Multimorbidity definition	Multimorbidity severity a	PTB definition	Confounders adjusted	Comparison Group (Exposure vs. Reference), PTB Subtype	Effect estimates
Al Khalaf et al.,​ 2021 [17, 43]	CKD, CH	Co-existence of CKD + CH	High	PTB < 37 weeks; extreme (< 28 weeks), severe (28–31 weeks), moderate (31–33 weeks), near term (34–36 weeks)	Maternal age; smoking; BMI; parity; country of origin; asthma; diabetes; cardiovascular disease; education; birth year	CKD + CH vs. no chronic conditions, medically indicated PTB < 37 weeks	aOR 9.09 (7.61–10.7)
						CKD + CH vs. no chronic conditions, SPTB < 37 weeks	aOR 1.18 (0.83–1.69)
						CKD + CH vs. no chronic conditions, PTB < 28 weeks	aOR 14.7 (10.1–21.3)
						CKD + CH vs. no chronic conditions, PTB 28–31 weeks	aOR 8.50 (6.20–11.6)
						CKD + CH vs. no chronic conditions, PTB 31–33 weeks	aOR 6.44 (4.72–8.79)
						CKD + CH vs. no chronic conditions, PTN 34–36 weeks	aOR 3.43 (2.84–4.14)
Aubry et al., ​2019 [39]	Diabetes, CH	One or more disorders co-occurring in the same individual	Moderate	PTB < 37 weeks	Maternal age; parity; smoking in pregnancy; ethnicity	Obese + Diabetes + CH vs. no chronic conditions, PTB < 37 weeks	aRR 2.03 (1.88–2.20)
						Non-obese Diabetes + CH vs. no chronic conditions, PTB < 37 weeks	aRR 2.34 (2.26–2.44)
Azcoaga-Lorenzo et al., ​2023 [7]	79 pre-specified long-term conditions (MuM-PreDiCT list)	≥ 2 pre-existing physical/mental conditions; complex ≥ 4	High	PTB 24–<37 weeks; subgroups: <28, 28–<32, 32–<37 weeks	Maternal age; socioeconomic status; ethnicity; previous pregnancies; BMI; smoking;	≥ 2 pre-existing conditions vs. no chronic conditions, PTB < 37 weeks	aOR 1.64 (1.48–1.82)
						≥ 2 pre-existing conditions vs. no chronic conditions, PTB 32–<37 weeks	aOR 1.69 (1.52–1.89)
						≥ 2 pre-existing conditions vs. no chronic conditions, PTB 28–<32 weeks	aOR 1.22 (0.90–1.66)
						≥ 2 pre-existing conditions vs. no chronic conditions, PTB < 28 weeks	aOR 0.86 (0.51–1.44)
Bandoli & Chambers., 2017 ​[36]	Autoimmune condition (RA, Crohn’s disease, psoriasis) ± depression	Co-occurrence of autoimmune condition and depression	Moderate	PTB < 37 weeks	Race/ethnicity; SES; medical comorbidities; pregnancy smoking; pre-pregnancy BMI; maternal age; GA at enrolment	RA + depression vs. no chronic conditions, PTB < 37 weeks	aOR 2.09 (0.71–6.16)
						Crohn’s + depression vs. no chronic conditions, PTB < 37 weeks	aOR 2.89 (0.80–10.44)
						Psoriasis + depression vs. no chronic conditions, PTB < 37 weeks	aOR 2.79 (1.11–7.00)
Berger et al., 2020 [34]	Diabetes Mellitus (D), Obesity (O), Hypertension (H)	Combination of D, O, H	Moderate	PTB 24 + 1/7 to 36 + 6/7; <34 vs. 34–37; provider-initiated vs. spontaneous​​	Age; parity	D + H vs. no chronic conditions, PTB < 37 weeks	aRR 6.34 (5.15–7.81)
						D + H vs. no chronic conditions, PTB < 34 weeks	aRR 10.33 (6.96–15.33)
						D + H vs. no chronic conditions, medically indicated PTB < 37 weeks	aRR 5.09 (4.49–5.77)
						D + H vs. no chronic conditions, SPTB < 37 weeks	aRR 3.05 (1.92–4.85)
						D + O vs. no chronic conditions, PTB < 37 weeks	aRR 3.09 (2.80–3.40)
						D + O vs. no chronic conditions, PTB < 34 weeks	aRR 2.73 (2.16–3.45)
						D + O vs. no chronic conditions, medically indicated PTB < 37 weeks	aRR 44.44 (3.76–5.25)
						D + O vs. no chronic conditions, SPTB < 37 weeks	aRR 1.83 (1.55–2.16)
						O + H vs. no chronic conditions, PTB < 37 weeks	aRR 2.96 (2.70–3.24)
						O + H vs. no chronic conditions, PTB < 34 weeks	aRR 4.44 (3.76–5.25)
						O + H vs. no chronic conditions, medically indicated PTB < 37 weeks	aRR 5.72 (5.13–6.38)
						O + H vs. no chronic conditions, SPTB < 37 weeks	aRR 1.17 (0.96–1.43)
						D + O +H vs. no chronic conditions, PTB < 37 weeks	aRR 5.55 (4.76–6.47)
						D + O +H vs. no chronic conditions, PTB < 34 weeks	aRR 5.04 (3.38–7.49)
						D + O +H vs. no chronic conditions, medically indicated PTB < 37	aRR 11.26 (9.4-13.49)
						D + O +H vs. no chronic conditions, SPTB < 37 weeks	aRR 1.73 (1.14–2.62)
Brown et al., 2016 [44]	Intellectual/developmental disabilities (IDD) + mental illness	Co-occurrence of IDD and mental illness	Moderate	PTB < 37 weeks	Maternal age; parity; neighbourhood income; region; pregnancy health conditions; infant sex	IDD + mental illness vs. no chronic conditions, PTB < 37 weeks	aRR 1.15 (0.94–1.39)
Cripe et al., 2011 [35]	Mood disorders + migraine	Co-occurrence of mood and migraine disorders	Low	PTB 20–<37 weeks; spontaneous vs. medically indicated	Maternal age; race/ethnicity; marital status; parity; smoking; chronic hypertension; pre-existing diabetes; pre-pregnancy BMI	Mood disorder + migraine vs. no chronic conditions, PTB < 37 weeks	aRR 1.87 (1.05–3.34)
						Mood disorder + migraine vs. no chronic conditions, medically indicated PTB < 37 weeks	aRR 2.02 (0.78–5.22)
						Mood disorder + migraine vs. no chronic conditions, SPTB < 37 weeks	aRR 1.94 (0.94–3.98)
Ibanez et al.,​ 2012 [40]	Depression + anxiety	Co-occurrence of depression and anxiety	Low	PTB < 37 weeks; spontaneous vs. medically indicated	Maternal age; education; parity; pre-pregnancy BMI; smoking; hypertension in pregnancy; maternity unit	Depression + anxiety vs. no chronic conditions, PTB < 37 weeks	aOR 1.61 (0.88–2.94)
						Depression + anxiety vs. no chronic conditions, medically indicated PTB < 37 weeks	aOR 0.41 (0.09–1.86)
						Depression + anxiety vs. no chronic conditions, SPTB < 37 weeks	aOR 2.47 (1.27–4.80)
Lehloa et al., 2025 [45]	HIV; Obestiy; CH, Diabetes	Co-occurrence of HIV and one or more non-communicable diseases	High	PTB < 37 weeks	Maternal age; parity; socioeconomic status	HIV + Hypertension vs. no chronic conditions, PTB < 37 weeks	aOR 3.01 (1.01–8.05)
						HIV + Hypertension + Obesity vs. no chronic conditions, PTB < 37 weeks	aOR 2.67 (1.08–6.23)
						Hypertension + Obesity vs. no chronic conditions, PTB < 37 weeks	aOR 3.18 (1.21–7.90)
Li et al., 2020 [33]	HIV and HBV	Co-infection (two infectious diseases)	High	PTB < 37 weeks; subgroups < 34, 34–36	Maternal age; GA; gravidity; parity; mode of delivery	HIV + HBV vs. no chronic conditions, PTB < 37 weeks	aOR 2.55 (0.34-–19.23)
Männistö et al., 2016 [42]	Depression, anxiety, bipolar disorder, schizophrenia, unspecified psychiatric disorder	Comorbid maternal psychiatric disorders	Moderate	PTB < < 37, < 34, <28; spontaneous vs. indicated​​	Maternal age; race/ethnicity; pre-preg BMI; insurance; marital status; smoking; substance/alcohol use; parity; diabetes; hypertension; thyroid disease; clinical site	Depression + anxiety vs. no chronic conditions, PTB < 37 weeks	aOR 2.31 (1.93-–2.78)
						Depression + anxiety vs. no chronic conditions, PTB < 34 weeks	aOR 3.06 (2.36–3.97)
						Depression + anxiety vs. no chronic conditions, PTB < 28 weeks vs. no chronic conditions, PTB	aOR 2.99 (1.83–4.90)
						Bipolar + dep/ax vs. no chronic conditions, PTB < 37 weeks	aOR 1.70 (1.30–2.22)
						Bipolar + dep/ax vs. no chronic conditions, PTB < 34 weeks	aOR 1.51 (0.99–2.31)
						Bipolar + dep/ax vs. no chronic conditions, PTB < 28 weeks	aOR 1.50 (0.66–3.43)
Nakanishi et al.,​ 2023 [6, 46]	Physical/psychological/social chronic conditions	≥ 2 chronic conditions	Low	PTB < 37 weeks; VPTB < 34 weeks	Maternal age; parity; smoking; alcohol; education; household income; neonatal sex	≥ 2 conditions vs. no chronic conditions, PTB < 37 weeks	aOR 1.50 (1.33–1.69)
						≥ 2 conditions vs. no chronic conditions, PTB < 34 weeks	aOR 1.34 (1.03–1.74)
Yanit et al., ​ 2012 [37]	CH + pregestational diabetes	Co-existence of CH + PGDM	High	Overall PTB < 37 weeks; ≤32 weeks	Maternal age; race/ethnicity; insurance; education; parity; prenatal visits; obesity; renal disease	CH + pregestational diabetes vs. no chronic conditions, PTB < 37 weeks	aOR 4.9 (4.0–6.0)
						CH+PGDM vs. no chronic conditions, PTB ≤ 32 weeks	aOR 7.6 (5.1–11.2)

Abbreviations: PTB Preterm birth, aOR adjusted odds ratio, aRR adjusted risk ratio, CI Confidence interval, BMI Body mass index, SES Socioeconomic status, SPTB Spontaneous preterm birth, GA Gestational age, RA Rheumatoid arthritis, IDD Intellectual and developmental disabilities, CH Chronic hypertension, CKD chronic kidney disease, PGDM Pregestational diabetes mellitus, HBV Hepatitis B virus, HIV Human immunodeficiency virus, NR Not reported, DH Diabetes and Hypertension, DO Diabetes and Obesity, OH Obesity and Hypertension, DOH Diabetes, Obesity and Hypertension

^aThe criteria and procedures used for severity categorisation are described in detail in the Methods section

Risk of bias

The quality of the included studies was assessed using the NOS. Overall, 5 studies were judged to have a low risk of bias (9 stars), 7 were rated as moderate (7–8 stars, ), and one was judged to be at high risk (≤ 6 stars) [24]. The most common sources of bias related to the representativeness of the sample to the total population. A full quality assessment is provided in Supplemental material 5.

Meta-analysis

Among studies included in the meta-analysis, maternal multimorbidity was significantly associated with an increased risk of PTB. The RR was 2.04, 95% CI of 1.61–2.09; p < 0.001 for PTB < 37 weeks, while the RR for PTB < 28 weeks was 3.23, 95% CI: 0.10-108.07; p < 0.001 (Fig. 2).

Fig. 2.

Forest plot of the pooled adjusted relative risks (aRR) with 95% confidence intervals for the association between maternal multimorbidity and preterm birth (A) <37 weeks and (B) <28 weeks

Subgroup analysis

To further assess heterogeneity additional subgroup analyses were performed based on sample size, multimorbidity severity, type of PTB, study design, geographical region, ascertainment time of diagnosis, risk of bias, and multimorbidity category. Analysis stratified by sample size showed a pooled RR of 2.04; 95% CI: 1.61–2.59; p < 0.001. Small sample size (< 10 000 participants) analysis resulted in an RR of 1.67; 95% CI: 1.10–2.54; p = 0.031, medium sample size (10 000 - <100 000 participants) RR: 1.56; 95% CI: 1.29–1.90, p = 0.294 and for large sample size (≥ 100 000 participants) the RR was 2.69; 95% CI: 1.55–4.68, p < 0.001. For small, medium and large sample size the I² values for heterogeneity were 62.2%, 18.2%, and 98.3%, respectively, and 97.4% for the overall (Fig. 3). Analysis based on multimorbidity severity demonstrated an overall pooled RR of 2.04; 95% CI: 1.61–2.59; p < 0.001. Subgroup analyses showed a RR of 1.49; 95% CI: 1.29–1.70; p = 0.714 for studies defining multimorbidity with low severity. Studies defining multimorbidity with moderate severity illustrated an RR of 2.08; 95% CI: 1.21–3.55; p < 0.001 and a RR of 2.44; 95% CI: 1.33–4.47; p < 0.001 for studies defining multimorbidity with high severity. The I² values for heterogeneity in the overall, low, moderate and high severity were 97.4%, 0%, 98.3% and 96.3%, respectively (Fig. 4).

Fig. 3.

Forest plot of the pooled adjusted relative risks (aRR) with 95% confidence intervals for the association between maternal multimorbidity and preterm birth (<37 weeks) by sample size

Fig. 4.

Forest plot of the pooled adjusted relative risks (aRR) with 95% confidence intervals for the association between maternal multimorbidity and preterm birth (<37 weeks) by multimorbidity severity

Further analyses stratified by type of PTB (spontaneous and medically indicated PTB), study design (prospective and retrospective cohort studies), regional analysis (North America, Europe and Asia), analysis based on the level of risk of bias (low and moderate risk of bias), analysis based on diagnosis ascertainment time (pre-conception and during pregnancy), and analysis based on the multimorbidity category (Mental Health multimorbidity, Metabolic/Cardiovascular multimorbidity and Mixed multimorbidity) are presented in Supplemental Material 6.

Assessment of publication bias was conducted using funnel plots and Egger’s regression asymmetry test. Visual inspection of the funnel for the overall analysis (PTB < 37 weeks) and subgroup analyses by gestational age (< 28 weeks), type of PTB, study design, sample size, geographical region, risk of bias, definition of multimorbidity, ascertainment time of diagnosis, and multimorbidity category indicated an approximately symmetric distribution of studies around the pooled effect estimate, with no marked asymmetry (Supplemental material 7). Egger’s test was not statistically significant across all models (p > 0.05), suggesting no evidence of small-study effects or publication bias. Taken together, these findings suggest that publication bias is unlikely to have materially influenced the results of this meta-analysis, although the limited number of studies in certain subgroups reduces the power to detect asymmetry. Despite the generally high methodological quality of the included studies, substantial heterogeneity was observed across analyses, resulting in a very low overall certainty of evidence according to the GRADE rating system. Summary results of the meta-analysis and the corresponding GRADE assessments are presented in Table 3.

Table 3.

Summary of meta-analysis results and GRADE assessment of the certainty of evidence for the association between multimorbidity and preterm birth

Outcome	Number of studies	Relative effect (95% CI)	Egger test p-value	Risk of bias (NOS)	Certainty of evidence (GRADE)
Preterm birth (< 37 weeks)	12	RR 2.04 (1.61–2.59)	> 0.05	Not serious	⨁◯◯◯ Very low a
Preterm birth (< 28 weeks)	3	RR 3.23 (0.10- 108.07)	> 0.05	Not serious	⨁◯◯◯ Very low a

Abbreviations: RR Relative Risk, NOS Newcastle–Ottawa Scale, CI Confidence Interval, GRADE Grading of Recommendations, Assessment, Development, and Evaluation

^aHigh heterogeneity observed across studies although the direction of effect was consistent. Downgraded by two levels for very serious inconsistency

Sensitivity analysis

Sensitivity analyses demonstrated that the findings were robust. In the RR models, the overall pooled effect size remained consistent, with pooled estimates ranging from 2.04 to 2.09 across leave-one-out analyses. None of the individual studies unduly influenced the overall results while the CI consistently excluded the null value. When alternative heterogeneity estimators were applied, including PM and DL, the pooled estimates remained virtually unchanged (≈ 2.04–2.05), indicating that the results were stable across statistical models and not dependent on the choice of variance estimator. Similarly, subgroup sensitivity analyses by risk of bias showed consistent findings. Among studies with low risk of bias, the pooled effect sizes remained stables across leave-one-out analyses, and no single study materially affected the results. When alternative estimators were applied (PM and DL), the pooled RR was 1.63 (95% CI: 0.99–2.69; p = 0.053), again indicating stable estimates across statistical approaches. For studies with moderate risk of bias, the leave-one-out analyses also demonstrated consistency, with pooled estimates ranging narrowly across iterations. The pooled RR was 2.32 (95% CI: 1.63–3.31; p = 0.001), confirming a statistically significant and robust association. Despite this robustness, the level of heterogeneity remained very high in most of the models, with I² values exceeding 90%, suggesting that true differences across study populations, multimorbidity definitions, and PTB classifications contributed substantially to between-study variability.

Discussion

This systematic review and meta-analysis investigated the association between maternal multimorbidity and the risk of PTB. Across the included studies, multimorbidity was associated with an increased risk of PTB, while also the increased association was consistent across different subgroups analysed. Sensitivity analyses confirmed robustness, as no single study unduly influenced the pooled estimates and findings remained stable across different statistical models. These results provide strong evidence that maternal multimorbidity is an independent risk factor for PTB.

Our findings are consistent with previous large-scale studies and a meta-analysis demonstrating that maternal multimorbidity substantially increases the risk of PTB. Evidence synthesized in a prior systematic review and meta-analysis of over 6.5 million pregnancies indicated more than a fourfold higher odd of PTB among women with multimorbidity compared with those without. Importantly, our study extends this evidence by drawing upon a different set of included studies and by employing a more restrictive definition of multimorbidity, limited to the coexistence of two or more chronic conditions. In contrast, the previous review also incorporated gestational diseases within its definition, which may partly account for differences in observed effect estimates [8]. Large population-based prospective cohort studies across diverse settings have reported more modest but still significant associations, with adjusted OR ranging from 1.50 to 1.64 [7, 46], and some demonstrating a clear dose–response pattern in which the risk increased with the number of chronic conditions [46]. Epidemiological data suggest that over one in five pregnant women in high-income countries meet criteria for multimorbidity [10], highlighting the substantial population burden. In addition to elevated PTB risk, women with multiple chronic conditions have been shown to experience higher rates of other adverse obstetric outcomes and increased delivery-related healthcare utilization and costs [47], underscoring the broader clinical and economic implications of managing multimorbidity during pregnancy. Importantly, our study extends previous evidence by focusing exclusively on adjusted estimates and conducting subgroup analyses by PTB definition (< 28, < 37 weeks) and by study characteristics or multimorbidity severity. This approach provides a more precise assessment of the independent contribution of multimorbidity to PTB risk, while addressing heterogeneity in outcome definitions and statistical reporting across studies.

Several interconnected biological and clinical mechanisms may explain the strong association between maternal multimorbidity and PTB. The coexistence of multiple chronic conditions can generate cumulative and synergistic pathophysiological effects that extend beyond the sum of individual disease risks [12, 13]. Chronic low-grade systemic inflammation and immune dysregulation, which are common features across many noncommunicable and infectious diseases, can amplify pro-inflammatory cytokine activity and prostaglandin synthesis, established triggers of cervical ripening and myometrial contractility [12, 48, 49]. Vascular compromise, resulting from endothelial dysfunction, impaired angiogenesis, and oxidative stress, may further reduce uteroplacental perfusion and oxygen delivery, promoting placental ischemia and dysfunction [13, 50, 51]. These placental insults are well-recognized contributors to both spontaneous PTB and medically indicated early delivery [5].

In addition to direct biological pathways, multimorbidity increases vulnerability to pregnancy complications such as pre-eclampsia, gestational diabetes, and maternal infections, each independently linked to elevated PTB risk [14–43, 46–53]. Management of multiple conditions often necessitates complex medication regimens (polypharmacy) [16], frequent monitoring, and multi-specialty input, which can fragment antenatal care and reduce continuity [54]. In high-risk pregnancies, clinicians may opt for iatrogenic PTB through induction of labor or planned caesarean section when maternal or fetal compromise is suspected, particularly in the context of worsening comorbidities or placental insufficiency [5, 55]. Psychosocial and behavioural factors, including increased stress, mental health disorders, reduced physical activity, and suboptimal nutrition, may also interact with biological pathways to heighten PTB risk [56, 57]. Collectively, these overlapping mechanisms underscore the multifactorial nature of the multimorbidity–PTB relationship and point to the need for integrated, multidisciplinary antenatal care models to optimize maternal health before and during pregnancy [58].

In our meta-analysis, the overall RR for the association between multimorbidity and PTB (< 37 weeks) was 2.04 (95% CI: 1.61–2.59). These findings indicate that multimorbidity confers a substantial increase in PTB risk. When compared with single-condition studies, the magnitude of risk is comparable to or exceeds that reported for some individual chronic diseases. For instance, pre-gestational diabetes has been associated with a pooled OR of 3.46 (95% CI: 3.06–3.91) across 81 observational studies [59], closely aligning with the risk observed in our pooled RR analysis. Chronic hypertension also shows a significant association with PTB (RR = 2.7, 95% CI: 1.9–3.6) [60], similar to our pooled RR estimate. In contrast, maternal asthma (RR ≈ 1.41, 95% CI: 1.22–1.61) [61] and maternal HIV infection (aOR = 1.72, 95% CI: 1.49–1.95) [52] are associated with more modest increases in risk. Antenatal depression has been linked to pooled RR ranging from 1.4 to 2.4, depending on context, with a pooled RR of 2.41 (95% CI: 1.47–3.56) in low- and middle-income countries [62]. Taken together, these comparisons suggest that multimorbidity is at least as strong a predictor of PTB as some of the most established single-condition risk factors, underscoring the importance of considering multiple chronic conditions collectively rather than in isolation.

To date, most evidence syntheses on PTB have focused on single risk factors rather than on the cumulative effect of multiple chronic conditions. For instance, a recent umbrella review comprehensively evaluated non-genetic predictors of PTB, including infections, environmental exposures, and psychosocial risk factors, but did not address multimorbidity as a composite exposure [63]. This underscores the novelty and importance of the present analysis, which specifically quantifies the impact of co-existing chronic conditions on PTB risk. By moving beyond isolated exposures, our study contributes to a broader understanding of the complex interplay between maternal health and adverse birth outcomes.

This systematic review and meta-analysis has several notable strengths. A comprehensive search strategy across multiple major databases was employed, and study selection, data extraction, and risk of bias assessment were performed independently by two reviewers, reducing the potential for selection and extraction bias. Importantly, only adjusted effect estimates were included in the quantitative synthesis, thereby minimizing the influence of confounding factors and enhancing the validity of the pooled results. Furthermore, we conducted subgroup analyses based on gestational age cut-offs, sample size, study design, geographical region, risk of bias, ascertainment time and multimorbidity severity, and robustness was assessed through multiple sensitivity analyses, including leave-one-out analyses and alternative estimators of between-study variance for the overall outcome and for the quality of the studies.

Nevertheless, several limitations should be acknowledged. The overall heterogeneity across studies was substantial, reflecting variations in study design, population characteristics, definitions of multimorbidity, and classification of PTB. Although subgroup analyses helped to partially address this issue, residual heterogeneity remains and should be interpreted with caution. The number of studies available for some subgroup analyses, particularly < 28 weeks, was limited, which reduces statistical power and precision of estimates. Furthermore, a dose–response analysis according to the number of chronic conditions could not be performed because most included studies defined multimorbidity as ≥ 2 conditions and rarely reported stratified data for higher counts. A further limitation relates to the inconsistent handling of hypertensive disorders across studies. While pre-eclampsia is a well-known cause of medically indicated PTB, most included studies did not clearly handle pre-eclampsia in association with chronic hypertension, nor did they adjust for it in a standardized manner. This lack of distinction may have influenced the estimated association between multimorbidity and PTB. In addition, reliance on observational studies means that residual confounding cannot be excluded, even though adjusted estimates were prioritized. Another limitation is the restriction to peer-reviewed, English-language publications, which may have introduced language and publication bias despite the use of funnel plots and Egger’s test to assess for small-study effects. Finally, several of the included studies were based on large national or regional cohorts, which may disproportionately influence pooled estimates, and differences in healthcare systems and maternal care practices across settings may limit the generalizability of findings.

The growing burden of multimorbidity among women of reproductive age carries significant consequences for both clinical practice and public health, with our results and prior evidence confirming its role as a major independent risk factor for PTB [7, 8, 64]. Early identification, ideally before conception or during the initial stages of pregnancy, can strengthen antenatal risk assessment, allowing healthcare providers to pinpoint high-risk patients and tailor monitoring, preventive strategies, and delivery planning to their specific needs [10]. Coordinated, multidisciplinary care models that bring together obstetric, medical, mental health, midwifery, and social care services are particularly well-positioned to address the complexity of multimorbid pregnancies [58]. These integrated approaches have the potential to reduce both spontaneous and medically indicated PTB by optimizing management of chronic conditions, maintaining continuity of care, and addressing social and behavioural determinants of health [5, 58]. Given that more than one in five pregnancies in certain high-income settings are affected by multimorbidity [10, 11], embedding routine screening and comprehensive care pathways within maternal health systems is essential, particularly in areas with high prevalence. Such measures could play a pivotal role in lowering PTB rates and improving long-term outcomes for both mothers and infants [47].

Future research should prioritize standardizing definitions and outcomes for maternal multimorbidity and PTB subtypes to improve comparability across studies. Establishing a core outcome set that captures key maternal outcomes (e.g., mortality, severe morbidity, changes in long-term conditions, quality of care, and new mental health diagnoses) and child outcomes (e.g., survival, gestational age, birth weight, neurodevelopment, and separation from the mother) would ensure consistent measurement and strengthen evidence synthesis [65]. Large, multicentre studies with diverse populations are needed to capture variations in risk across different demographic, socioeconomic, and geographic contexts. Further work should investigate effect modification, examining how factors such as socioeconomic status, healthcare access, and regional differences influence the association between multimorbidity and PTB. There is also a pressing need for intervention studies to evaluate targeted strategies, such as integrated and multidisciplinary care models, to reduce adverse outcomes in this high-risk population. Finally, individual participant data meta-analyses could refine risk estimates, assess heterogeneity, and provide patient-level insights to inform tailored care and risk stratification.

Conclusion

Our study shows that maternal multimorbidity is significantly associated with an increased risk of PTB. The findings remained robust across multiple definitions of PTB and consistent between effect measures, although substantial heterogeneity across studies warrants careful interpretation. Recognizing and addressing multimorbidity in pregnancy is crucial for improving maternal and neonatal outcomes, underscoring the need for integrated clinical management and further research into preventive strategies.

Authors’ contributions

LM contributed to the conceptualization of the review, development of the search strategy, database searches, screening of titles and abstracts, data extraction, statistical analysis, data visualization, and drafting of the original manuscript. DL contributed to methodological guidance, verification of data extraction, and writing – review and editing. KG contributed to the conceptualization, full-text screening, data extraction, risk of bias assessment, validation of analyses, methodological oversight, project administration, supervision, and drafting of the original manuscript.

Funding

This research received no funding.

Data availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.