Preterm birth: associated risk factors in the tertiary care center

Introduction

A neonate born before 37 completed weeks of gestation requires significant attention and care as they transition into a new environment post-birth. The morbidity associated with preterm birth can persist into later life, leading to physical, psychological, and economic costs. Globally, one in 10 babies is born preterm, and approximately one million babies die annually due to complications from preterm birth.^2–6 Preterm birth is an obstetric emergency and poses a threat to population health, contributing to 75% of infant mortality.^5–10

Preterm birth not only imposes financial and emotional distress on families but can also result in permanent disabilities (physical or neural) in infants. Surviving babies often exhibit periodic disabilities such as learning, visual, and hearing difficulties. The preterm birth rate ranges from 5% to 7% of live births in developed countries compared to developing countries.^11,12 Despite an increased understanding of potential risk factors and their pathological mechanisms, the preterm birth rate has remained unchanged or even increased in most countries over the past two decades.^13,14

The pathway to preterm birth remains unclear, whether it results from the interaction of multiple pathways or an independent pathway. Factors commonly affecting preterm birth include the health condition of the mother or fetus, genetic causes, environmental exposure, infertility treatments, habits, socioeconomic factors, and iatrogenic factors. Preterm birth was the second leading cause of death in children under five years old in 2010. Of the approximately 3.1 million newborn deaths that year, a quarter occurred within the first 24 hours after birth (American College of Obstetricians and Gynecologists, 2020).^15,16

In India, out of 27 million neonates born each year, 3.5 million are delivered prematurely. The antenatal period, labor process, and postnatal period are critical for infant and maternal survival.¹¹ Preterm labor is unpredictable, but cues can be identified, and preventive measures can be taken. Physicians strive to delay delivery to allow the baby to grow as much as possible. Therefore, pregnant women should not omit any essential health details during regular visits to the obstetrics clinic. They should provide a detailed history of their lifestyle, past pregnancies, and any health issues they have experienced and clarify any doubts.¹¹

Given the high prevalence and significant impact of preterm birth, there is a crucial need to assess various risk factors associated with it. The statistics and incidence rates indicate that preterm births contribute to substantial infant mortality and long-term morbidity.¹⁷ However, there is a gap in understanding the specific modifiable and non-modifiable risk factors that contribute to preterm birth in different regions, including Udupi district Karnataka. Identifying these risk factors can help develop targeted interventions to reduce preterm birth rates, improve maternal and neonatal health outcomes, and alleviate the financial and emotional burden on affected families. This research aims to fill this gap by providing comprehensive data on the risk factors specific to the study population, thereby informing healthcare strategies and policies.

The statistics presented above and the supporting data assist obstetric care providers in designing appropriate studies and planning measures to decrease delivery rates before 37 weeks of gestation and improve the health status of women who have delivered. This will help reduce and fill the gaps between study areas, serving as baseline information for other countries. The present study was conducted to identify preventive measures to determine the risk factors of preterm labor. Early identification and extra precautions for risk factors, such as medication intake during pregnancy and previous abortion, can prevent preterm birth. Moderate risk factors like hard physical work and conception age above 30 years can prevent preterm labor if managed.

In the multivariate logistic regression, the following confounders were included: maternal age, parity, gravidity, residence, education, occupation, and past obstetric history (including abortions and stillbirths). These variables were selected based on both existing literature and statistical relevance in univariate analysis. Inclusion was intended to minimize confounding by socioeconomic and biological risk factors. Variables with low frequency or alignment issues were excluded to maintain model stability.

In this table:

Methods

This retrospective case-control study aimed to identify the prevalence and risk factors for preterm birth by examining the case records of women who delivered in tertiary care hospitals in the Udupi district. The cases were women who delivered before 37 completed weeks of gestation, while controls were women who delivered at or after 37 weeks of gestation without any complications. As per the sample size calculation, a purposive sampling technique was employed to select the records of 250 out of 356 women.

Study Timeline and Record Selection

Sample Selection

The tools used included a maternal socio-demographic proforma and a structured Risk Assessment tool for preterm labor. This tool classified items into modifiable and non-modifiable risk factors, with the former further subdivided into social, economic, and environmental factors and the latter into medical, obstetrical, and fetal conditions.

The study utilized a structured Risk Assessment tool

Data Analysis Plan

The study received permission from various authorities related to the data and ethical clearance (IEC 30/2018), and it was registered with the Clinical Trial Registry of India (CTRI/2018/05/014078). Data were analyzed using SPSS version 16, employing descriptive and inferential statistics.²¹

The study identified that the prevalence of preterm labor was 14.82% (356 out of 2402 deliveries). The background characteristics of the cases and controls varied, with the majority residing in rural areas and being homemakers. The mean height and age among the cases were 154.32 cm and 27.44 years, respectively, and the average total weight gain during pregnancy was 7.67 kg.

Religion was considered as a factor influencing preterm birth because cultural beliefs and practices, including religious ones, can significantly impact health behaviors and outcomes. For instance, religious beliefs may influence decisions about seeking medical care, adherence to prenatal care recommendations, and the use of traditional medicines or rituals1. These results must be interpreted with caution. For instance, the adjusted odds ratio (AOR) of 62.4 seen for medication consumption is unreasonably high and should be interpreted cautiously. This finding likely reflects confounding by indication meaning that — medications were administered to women who already had high risk of preterm birth rather than causing preterm birth themselves, and hence the very high statistical correlation without a causal effect. Similarly wide confidence internals observed for certain variables (e.g., conception age ≥30 years, AOR 24.8 [2.6–232.9]) suggest limited statistical precision and possible model instability therefore the magnitude of these associations should be interpreted cautiously.³⁶

Univariate analysis was initially performed to identify potential risk factors associated with preterm birth. Variables that were statistically significant in univariate analysis were subsequently entered into the multivariate logistic regression model to control for confounding variables followed by the computation of the adjusted odds ratio to obtain an accurate result. The data showed that risk factors like pregnancy-induced Hypertension Hypertension (p=.001), medication intake (p=.001), and conception age at 30 or above (p=.005) are associated with preterm labor, which is significant at the 0.005 level. Although the analysis suggested that previous abortion and hard physical work appeared ‘protective’ (OR < 1), this is likely due to statistical artifact or residual confounding. In epidemiological terms, these are examples where the observed association may not reflect biological plausibility. For instance, women with a history of abortion may have received closer medical monitoring, leading to earlier detection of complications, while the ‘protective’ effect of physical work may reflect misclassification bias or small subgroup size. Hence, these findings should not be interpreted as clinically protective. Most women in both the case and control groups were administered medications such as Tibolone (a steroid) and Ceftriaxone (an antibiotic). Statistically, these medications were associated with an increased likelihood of preterm labor. The apparent association of steroids and antibiotics with preterm birth reflects confounding by indication. These medications are routinely given to women already at imminent risk of preterm labor (e.g., PROM, PIH). Therefore, the observed statistical link does not imply causation but rather that these drugs are markers of underlying risk. This distinction is essential to avoid misinterpretation of clinical practice.^22,23

Discussion

The data from this study revealed that 14.82% (356 out of 2406) of deliveries were preterm. This finding aligns with a cross-sectional study conducted on the prevalence of preterm labor among young parturient women aged 15 – 24 years attending public hospitals in Brazil, which found a high prevalence of preterm labor (86.3% out of 2400 parturient women). For the world, approximately 15 million babies are born preterm yearly, with rates varying from 5–7% in high-income nations to 11–18% for South Asia and Sub-Saharan Africa (WHO, 2023). Our rate of 14.82% falls within this worldwide high-burden range.^5,6,11,24,37

Similarly, a retrospective study in Jordan identified the existence and reasons associated with preterm delivery, showing that all the preterm deliveries were approximately 647. Most neonates were female (54.9% Vs. 45.1%), and most (75.6%) were the second child. The women who delivered preterm were predominantly between 25 and 35.^4,24–26

This study found associations between preterm labor and several factors, including age, residence, education level, occupation, marital status, gravida, number of deliveries, type of delivery, birth interval, blood group, and religion. These risk determinants reflect international evidence: multicountry research published in Lancet Global Health and BMC Pregnancy & Childbirth repeatedly identify pregnancy-induced hypertension, older maternal age, and hard work as preterm birth predictors. Nevertheless, determinants like religion and culture, which were present in our study, are reported less commonly outside of the region, indicating regional heterogeneity. However, no association was found between preterm labor and height or weight gain during pregnancy.

These findings are supported by a retrospective cross-sectional study on the prevalence of preterm labor in a Labor room, which identified risk factors for early labor such as active relationships during the previous week of labor, multiple pregnancies, small birth intervals between two conceptions, PIH, fetal anomaly, premature rupture of membrane, and Hypertension.^8,13,27

A case-control study by Barbara Luke et al. in the United States on the relation between occupational factors and preterm delivery among nurses found that risk factors related to preterm delivery included working hours per week, different duty timings, standing hours, noisy areas, physical stress, and work-related stress (American College of Obstetricians and Gynecologists, 2020). Similar associations have been found in European and East Asian populations, suggesting that occupational exposures are a global risk factor for preterm birth.^7,28–30

The results of this study and similar ones suggest that eliminating risk factors and reinforcing protective factors could help decrease the rate of preterm labor and its human and social burden.^31,32 However, further studies with better design, such as cohort prospective studies with a proper follow-up period and large study population, are needed to determine these factors accurately. As the Hospital studied is a referral center for these patients, it represents the general population of the country to a great extent.¹¹

The study identified a significant prevalence of preterm labor at 14.82% among deliveries in the tertiary care hospital of Udupi district. Key risk factors such as pregnancy-induced hypertension, medication intake, and conception at or above age 30 were significantly associated with preterm labor. The study’s findings align with other international studies, highlighting the global relevance of understanding and addressing these risk factors. These results also support international public health priorities. The WHO ‘Born Too Soon’ campaign and Sustainable Development Goal 3.2 highlight the imperative for declining preterm-related mortality across the world. Modifiable factors like maternal hypertension, workload, and birth spacing must be tackled not just for India but also for high-burden countries in Sub-Saharan Africa and Latin America.^33,34,38

Limitations of the Study

Several limitations should be considered when interpreting these findings.

The retrospective case-control design limits the ability to establish causal relationships, the study was conducted in a single tertiary care referral hospital, which may limit the generalizability of the findings. Recall bias from participants’ self-reported data can lead to inaccuracies. While several confounders were adjusted for, unmeasured factors might still influence the results. Measurement imprecision, particularly with gestational age, and potential selection bias if the sample is not representative are notable concerns. Loss to follow-up or missing data could also affect the robustness of the findings.

Some important fetal factors such as intrauterine growth restriction (IUGR), congenital anomalies, and abnormal fetal presentations could not be included due to incomplete medical record documentation. Additionally, wide confidence intervals observed in some regression estimates indicate limited statistical precision.

Generalisability of the Study Results

The findings from this study, conducted in a tertiary care hospital in Udupi district, may not be fully representative of other populations or settings due to several factors:

However, they did not find any significant association between these factors and preterm birth. This outcome might be due to various reasons, including sample size limitations, variations in data quality, or the complex interplay of multiple factors influencing preterm birth. It highlights the need for further research with larger and more diverse populations to better understand the role of fetal factors in preterm births. The cross-sectional and retrospective nature of the study limits its generalizability over time. Prospective studies with larger, multi-center populations are needed for robust results.

Thus, while the findings are valuable for understanding preterm birth risks in this hospital context, they may not directly generalize to rural, primary care, or non-referral populations.

Conclusion

The study concludes that preterm birth is commonly seen in pregnant women who are exposed to non-modifiable and modifiable risk factors. Modifiable risk factors can be avoided and thus allow the pregnancy to continue till term. Non-modifiable risk factors need to be supervised very keenly so that there is no risk to the life of the mother and the fetus. It even states that the prevalence of preterm is higher among homemakers. All in all, the study concludes that these risk factors are different for each woman, which will lead to preterm labor, but they need to be identified at the earliest and treated adequately. However, these findings must be interpreted with caution, as they are based on a single-center, referral hospital population using retrospective records, which may limit their applicability to broader community settings.³⁵ Thus, these women need to be released from the level of stress to which they are exposed. All in all, the study concludes that these risk factors are different for each woman, which will lead to preterm labor, but they need to be identified at the earliest and treated adequately.

More information can be found on our data guidelines page. Data is available in the OSF web application: Contributors: Sweety Jousline Fernandes Identifier: https://osf.io/wp8xu/?view_only=da380485c7c74b20befc1e58f95fe431

Ethical statement: Ethical consideration

The data collection for the study commenced only after obtaining prior ethical clearance from the departments, which are as listed below:

Administrative permission from Dean Manipal College of Manipal, to conduct the study.

Permission from IRC of Manipal College of Nursing Manipal for conducting the study (IRC148/2017)

Permission from IEC of Kasturba Hospital Manipal on 10/1/2018 (IEC 30/2018)

CTRI registration (CTRI/2018/05/014078) (its not a clinical trial but during this study duration it was insisted to register under clinical trial)

Permission from the OBG Department of Kasturba Hospital Manipal was obtained for the selection and utilizing the samples Records

Permission from the Medical Superintendent of the Kasturba Hospital Manipal, was obtained to have access to the Medical Record

Consent to participate written consent was obtained from the hospital authority and the MRD section to view the records. Based on the hospital numbers selected every day fifteen records were issued for the study assessment.

Author contributions

Principal Investigator: Mrs. Sweety J Fernandes (role: Recruitment, data collection, rater, intervention provider, concept analysis, training provider, and manuscript)

Guide: Dr Tessy Treesa Jose (Role: Intellectual inputs, mentoring)

Co–guide: Dr Judith A Noronha (Role: Intellectual, mentoring)

Co-investigator: Dr Sushmitha R Karkada (Role: Intellectual analysis)