Temporal trends in preterm birth rate over the last 30 years in Sweden: a population-based study

This population-based study of over 3.2 million Swedish pregnancies reveals that while preterm birth rates remained stable from 1991 to 2005, they significantly declined from 2006 to 2021 due to a reduction in spontaneous preterm births, particularly among late and moderately preterm categories, despite the absence of a nationwide preventive strategy.

The Gist

Imagine the world of pregnancy as a massive, bustling train station. For decades, trains (babies) have been leaving the station too early, before they were fully packed and ready for the long journey ahead. This "early departure" is called preterm birth, and for a long time, it seemed like the schedule was stuck. Globally, the number of early departures has stayed high or even gotten worse, despite better ticketing systems (medical care) and safer stations.

But in Sweden, something strange and wonderful happened. A new study looked at the station logs for the last 30 years (1991–2021) and found a surprising trend: the number of early departures started to drop significantly after 2005.

Here is the story of that drop, explained simply:

1. The Mystery of the "Spontaneous" Departures

The researchers noticed that the drop wasn't because doctors were stopping early departures that they caused (like scheduling a C-section too early). Those numbers stayed steady.

Instead, the drop happened because fewer trains were leaving on their own, unexpectedly.

• The Analogy: Imagine a train station where, for years, trains would randomly jump the tracks and leave early due to mechanical failures or sudden storms. Suddenly, without anyone changing the schedule or fixing the tracks, these random early departures just started happening less often.
• The Detail: The biggest drop was in "late preterm" births (babies born just a few weeks early, around 34–36 weeks). It's like the station started holding onto the trains just a little bit longer, ensuring they were fully ready before leaving.

2. The "Magic" of the Swedish Data

Sweden is like a library that keeps a perfect record of every single birth in the country. The researchers pulled out 3.2 million records. They saw that:

• Mothers are older: People are having babies later in life (which usually increases risk).
• Smoking is down: Fewer moms are smoking (which usually decreases risk).
• Tech is up: More babies are being made with fertility treatments (which usually increases risk).

The Puzzle: Usually, if you mix "older moms" and "more fertility tech," you'd expect more early births. But in Sweden, the opposite happened. The early birth rate went down.

3. The "Detective Work" (Why did it happen?)

The researchers tried to solve the mystery by asking: "Did we fix the risk factors?"

• They checked if the drop was because moms stopped smoking. No.
• They checked if it was because of changes in fertility treatments. No.
• They checked if it was because of the age of the moms. No.

Even when they adjusted for all these known factors, the drop in early births remained. It's as if the station manager said, "We didn't change the rules, we didn't fix the tracks, and we didn't change the passengers, but suddenly, fewer trains are leaving early."

4. Who Benefited?

The good news wasn't just for one group; it was a nationwide improvement.

• Mothers from Europe, Asia, and Africa all saw a drop in early births.
• The biggest improvement was seen in mothers born in Africa, whose early birth rates dropped the most dramatically.
• This suggests that something in the general environment or society improved for everyone, regardless of where they were born.

5. The "Timing" Question

Could it be that they just got better at measuring when the baby was due?

• The Analogy: Maybe they stopped using a rusty clock (last menstrual period) and started using a super-accurate atomic clock (ultrasound).
• The Result: The researchers tested this. Even when they used the "rusty clock" method to calculate the dates, the drop in early births was still there. So, it wasn't just a measurement error; the babies were genuinely staying in longer.

The Big Takeaway

For the last 30 years, the world has been trying to figure out how to stop preterm births, with little success. Sweden, however, accidentally discovered a "secret sauce" that made preterm births drop by nearly 1% over 15 years.

The most important part? They didn't have a specific national plan to do this. There was no "Stop Preterm Birth" campaign in 2005. It just happened.

The Lesson: If Sweden can do this without a specific plan, there must be something in their society, environment, or healthcare system that is naturally protecting babies. The goal now is to figure out what that secret ingredient is, so other countries can copy it and turn this "accidental success" into a global rule.

In short: The train station in Sweden figured out how to keep the trains on the tracks longer, even though they didn't know exactly how they did it. Now, the whole world is watching to see if they can learn the trick.

Technical Summary

1. Problem Statement

Preterm birth (PTB), defined as delivery before 37 weeks of gestation, is the leading cause of perinatal mortality and a significant contributor to the global disease burden. Despite improvements in living conditions and antenatal care, global PTB rates have remained stagnant or increased in many high-income countries over the last few decades. This trend is often attributed to shifting maternal demographics (e.g., advanced maternal age) and increased use of assisted reproductive technologies (ART).

However, the specific drivers of temporal trends in PTB remain poorly understood. Sweden offers a unique setting to investigate this due to its high-quality, population-based Medical Birth Register (MBR). The study aims to determine the secular trends in overall, spontaneous, and iatrogenic PTB rates in Sweden from 1991 to 2021 and assess whether known risk factors (maternal age, parity, smoking, ART, continent of birth) explain these trends.

2. Methodology

• Study Design: A population-based register study.
• Data Source: The Swedish Medical Birth Register (MBR), covering nearly all births in Sweden since 1973.
• Study Population:
  • Inclusion: Pregnancies resulting in live births between 154 and 308 days (22–44 weeks) from January 1, 1991, to December 31, 2021.
  • Exclusions: Stillbirths, missing gestational duration, mothers without pseudonymized identifiers, duplicate records, and outliers based on birth weight/gestational age correlation (using GAMLSS).
  • Final Sample: 3,210,294 pregnancies (3,163,389 singletons; 46,905 multiples).
• Variable Definitions:
  • Gestational Age: Primarily based on the "best estimate" (ultrasound in 85% of cases or Last Menstrual Period [LMP]). Sensitivity analyses were conducted using LMP-based estimates.
  • PTB Categories:
    • Spontaneous: Onset of contractions or prelabor rupture of membranes.
    • Iatrogenic: Induction of labor or cesarean section.
    • Gestational Sub-types: Extremely (<28w), Very (28–31w), Moderate (32–33w), and Late (34–36w).
  • Covariates: Maternal age, parity, smoking, ART usage, and maternal continent of birth.
• Statistical Analysis:
  • Linear regression models were fitted to annual PTB rates, split into two periods (1991–2005 and 2006–2021) based on observed trend shifts.
  • Logistic regression models were used to estimate Odds Ratios (OR) for PTB across birth years, adjusting for maternal age, parity, smoking, ART, and continent of birth.
  • Analyses were stratified by singleton/multiple births and maternal origin.
  • Software: R (version 4.4.2).

3. Key Results

• Overall Trends:
  • 1991–2005: The overall PTB rate was stable at approximately 5.50%.
  • 2006–2021: A significant decline occurred, dropping to 4.78% in 2021.
  • Driver of Decline: The reduction was entirely driven by spontaneous preterm births. Rates of iatrogenic preterm births remained stable or increased slightly throughout the study period.
• Gestational Age Stratification:
  • Late Preterm (34–36 weeks): Showed the most significant decline (from 3.92% in 2005 to 3.52% in 2021).
  • Moderate Preterm (32–33 weeks): Also declined significantly.
  • Very/Extremely Preterm (<32 weeks): Rates remained stable or showed non-significant increases; no decline was observed in the most severe categories.
• Maternal Origin:
  • Declines in spontaneous PTB were observed across all major groups (European, Asian, and African origins).
  • The largest relative decline was observed in mothers of African origin (from 2.65% in 1991 to 1.72% in 2021).
• Risk Factor Analysis:
  • Adjusting for maternal age, parity, smoking, ART, and continent of birth did not explain the observed decline.
  • While smoking rates dropped significantly (24.25% to 3.25%) and ART usage increased (3.70% to 12.33%), these opposing trends did not account for the reduction in spontaneous PTB.
• Sensitivity Analysis (Dating Method):
  • When using LMP-based gestational age estimates instead of the "best estimate," the decline in spontaneous PTB persisted, confirming the trend is robust to dating methodology changes.

4. Key Contributions

• Identification of a Unique National Trend: The study documents a sustained, nation-wide decline in spontaneous preterm birth in Sweden from 2006 to 2021, contrasting with global trends of stagnation or increase.
• Decoupling from Known Risk Factors: It provides strong evidence that the decline is not attributable to changes in established risk factors (smoking, maternal age, ART, or parity), suggesting the influence of unmeasured population-level determinants.
• Timing Relative to Policy: The decline began in 2006, predating Sweden's first national guidelines for the medical prevention of preterm birth (introduced in 2023), indicating that the trend is likely driven by factors other than specific clinical interventions.
• Granular Stratification: The study offers detailed breakdowns by gestational age sub-types and maternal continent of birth, highlighting that the improvement is specific to late/moderate spontaneous births rather than extreme preterm births.

5. Significance and Implications

• Public Health Insight: The findings challenge the assumption that PTB rates are inevitably rising or stable in high-income nations. The Swedish experience suggests that spontaneous PTB can be reduced at a population level without a specific, targeted national prevention strategy.
• Research Direction: Since traditional risk factors do not explain the decline, future research must investigate "unmeasured" determinants. Potential candidates include:
  • Improvements in general maternal health and nutrition.
  • Reductions in environmental exposures (e.g., passive smoking bans implemented in 2005).
  • Changes in antenatal care practices or broader social determinants of health.
  • Birth cohort effects (differences in early-life exposures for mothers born in the 1960s/70s vs. later decades).
• Global Applicability: Understanding the mechanisms behind the Swedish decline could provide a blueprint for other nations to transform isolated reductions in PTB into a reproducible global pattern.

Limitations: The study relies on self-reported smoking data (subject to misclassification) and lacks data on specific neonatal complications or the exact timing of ultrasound dating, which limits the ability to fully assess clinical outcomes. However, the use of a near-complete national registry (96% coverage) and rigorous statistical adjustments strengthens the validity of the observed trends.