War Zones Paradox: Newborns in Conflict Areas Show Heavier Births in Landmark Meta-Analysis

One of the most counterintuitive findings in conflict health research has just been published. A team of researchers from Qatar, Palestine, and Egypt conducted what they describe as the first quantitative synthesis of its kind, pooling data from seven studies across Ukraine, Ethiopia, Poland, Serbia, and the former Yugoslavia to ask a deceptively simple question: does exposure to armed conflict change the odds of preterm birth, stillbirth, neonatal death, or the physical measurements of newborns? The answer, published in the International Journal of Public Health, is stranger than anyone expected. After statistical corrections, the analysis found no significant association between conflict exposure and preterm birth, stillbirth, or neonatal mortality. Even more strikingly, babies born in conflict-exposed populations appeared to weigh more and have larger head circumferences than their non-exposed counterparts.

The research team, led by Amani N. Alansari of Hamad Medical Corporation in Doha, followed PRISMA guidelines and registered the review prospectively. They searched PubMed, EMBASE, Web of Science, and SCOPUS from database inception through January 2026, initially retrieving 7,902 records. After removing 1,885 duplicates and screening 6,017 titles and abstracts, nineteen full-text reports were assessed for eligibility. Twelve were excluded for reasons including lack of a comparison group, irrelevance to war, or conference abstract status. Seven studies ultimately made the cut, spanning retrospective cohorts, historical cohorts, cross-sectional comparisons, and national registry analyses. Sample sizes ranged from a mere 39 participants in a small Kharkiv hospital comparison to over 500,000 live births in Ukrainian national registry data.

The methodological machinery behind the findings is worth understanding. Dichotomous outcomes such as preterm birth, stillbirth, and neonatal death were pooled as odds ratios using Review Manager 5.4 software, while continuous measures like birth weight, birth length, and head circumference were pooled as mean differences. The researchers started with fixed-effects models and switched to random-effects models when heterogeneity, quantified by the I² statistic, proved substantial. For preterm birth, the initial fixed-effects estimate across four studies was essentially null (OR 0.96, 95% CI 0.83–1.10), but heterogeneity was extreme at 85%. After switching to a random-effects model and excluding one influential study, heterogeneity dropped to a moderate 46% and the pooled estimate remained non-significant at OR 0.80 (95% CI 0.61–1.05).

Stillbirth told a similar story. The fixed-effects model across four studies yielded an odds ratio of 1.00 (95% CI 0.94–1.07) with 80% heterogeneity. After applying random effects and removing the Polish World War II study by Liczbinska and colleagues, heterogeneity vanished entirely (I² = 0%) and the estimate stayed flat at OR 0.98 (95% CI 0.93–1.05). Neonatal mortality followed the same pattern: the fixed-effects model initially suggested a significant 22% reduction in death among conflict-exposed infants (OR 0.78, 95% CI 0.71–0.85), but with 89% heterogeneity that finding dissolved under random-effects modeling into a non-significant OR of 1.07 (95% CI 0.73–1.56). In other words, what looked like protection was statistical noise.

Then came the anthropometric paradox. Pooling two studies, the random-effects model showed that conflict-exposed neonates had significantly higher birth weight, with a mean difference of 1.56 kg (95% CI 0.55–2.56), though heterogeneity was severe at 92%. Head circumference also came out significantly larger in exposed infants, with a mean difference of 0.60 cm (95% CI 0.47–0.74) and only modest heterogeneity at 35%. Birth length showed no significant difference. These are not small effects; a 1.56 kg mean difference in birth weight is enormous by any clinical standard, and the authors themselves flag it as biologically implausible as a true protective effect of war.

So what explains it? The researchers propose three mechanisms. First, survival bias: infants who survive to delivery in conflict zones may represent hardier pregnancies, while compromised fetuses may end in early pregnancy losses or stillbirths that go unrecorded in settings with collapsed surveillance systems. Second, displacement patterns may concentrate healthier, wealthier families near functioning hospitals, while those unable to flee face outcomes invisible to facility-based data collection. Third, the included studies varied enormously in how they defined exposure, timed it, and selected comparison groups, from the 1999 Belgrade bombing cohort to Ukrainian national registry trends comparing 2012 with 2022. The cross-sectional Kharkiv study comparing residents with internally displaced persons, rated only fair quality, reported shorter birth length and lower Apgar scores among displaced infants, hinting at what facility-based data may be missing.

The findings clash sharply with prior literature. Le and Nguyen’s analysis of 53 developing countries over three decades found that intrauterine conflict exposure reduced birth weight by 2.8% and raised low birth weight incidence by 3.2 percentage points. Mansour and Rees documented increased low birth weight risk among Palestinian infants during the al-Aqsa Intifada. Keasley and colleagues’ systematic review identified consistent evidence of increased low birth weight across nine studies from Libya, Bosnia, Palestine, and Afghanistan. Jawad and colleagues, analyzing 181 countries from 2000 to 2019, found that wars increased maternal and infant mortality, though notably not neonatal mortality, a result consistent with the current review. Behboudi-Gandevani and colleagues found increased odds of small-for-gestational-age infants and neonatal mortality among immigrants from conflict zones, suggesting that war’s effects travel with displaced populations long after the guns fall silent.

The authors are emphatic that the absence of adverse associations should not be read as evidence of protection. Most included studies relied on hospital-based data, systematically undercounting the most vulnerable women who cannot reach healthcare at all. Several studies scored only moderate quality on the Newcastle-Ottawa Scale, primarily because they lacked multivariable adjustment for maternal age, parity, socioeconomic status, and pre-existing conditions. The review also could not examine long-running conflicts in Syria, Yemen, Sudan, or Palestine, and the authors note that outcomes in these settings are entangled with economic collapse, food insecurity, and healthcare disruption, making it nearly impossible to isolate war’s direct effects. Exposure to hazardous munitions, including reported use of incendiary substances in populated areas, adds another layer of risk that routine hospital datasets cannot capture.

The practical implications are stark. The authors call for community-based prospective cohorts with proper adjustment for nutrition, socioeconomic indicators, and healthcare access, alongside mobile health initiatives and parallel surveillance systems operating outside hospital walls. Clinicians caring for conflict-affected populations should maintain heightened vigilance regardless of what pooled estimates suggest, and humanitarian organizations should invest in data collection mechanisms that reach women delivering at home, in shelters, or in transit. The paradoxical heavier babies in this meta-analysis are, in effect, a statistical signature of everything the data cannot see. As the authors conclude, the most effective intervention for protecting maternal and neonatal health globally remains the prevention of armed conflict itself, a reminder that some public health problems cannot be solved within the health system at all.

Subject of Research: The association between armed conflict exposure and neonatal outcomes including preterm birth, stillbirth, mortality, and anthropometric measures

Article Title: The impact of armed conflict exposure on preterm birth, stillbirth, and neonatal anthropometric measures: a systematic review and meta-analysis

Article References: Alansari, A. N., Sabbah, A. F., Salim, A., Varghese, B., Singh, K., & Elshanbary, A. A. (2026). The impact of armed conflict exposure on preterm birth, stillbirth, and neonatal anthropometric measures: a systematic review and meta-analysis. International Journal of Public Health, 71, Article 1609784. https://doi.org/10.3389/ijph.2026.1609784

Image Credits: AI Generated

DOI: 10.3389/ijph.2026.1609784

Keywords: armed conflict, neonatal outcomes, preterm birth, stillbirth, birth weight, meta-analysis, systematic review, survival bias, humanitarian health, Ukraine, maternal health, public health