Mothers face immediate, but family-size dependent, costs of sons in preindustrial Finland

Analysis of preindustrial Finnish data reveals that while having more sons does not immediately reduce maternal survival in small families, it significantly increases short-term mortality risks for mothers with larger families, suggesting accumulated physiological costs that are often overlooked in studies focusing on post-reproductive lifespans.

The Gist

Imagine a mother's body as a battery pack that powers her family. In the animal kingdom, there's a long-standing idea called the "Expensive Son Hypothesis." It suggests that raising a son is like plugging in a high-drain appliance (like a gaming console or a heater) that uses up the battery faster than raising a daughter (which is more like a standard lamp). This is because, biologically, male babies are often larger and require more energy during pregnancy and breastfeeding.

For a long time, scientists thought this "battery drain" only showed up later in life, after a woman stopped having children. They assumed the damage was slow and cumulative, only revealing itself in old age.

But this new study from pre-industrial Finland flips the script.

The researchers looked at the life records of over 5,000 mothers from a time before modern medicine, when life was hard and childbirth was dangerous. They asked a simple question: Does having more sons make a mother more likely to die within a year of her last baby being born?

Here is what they found, broken down into a simple story:

1. The "One-and-Done" Mom vs. The "Large Family" Mom

If a mother had only one child, it didn't matter if it was a boy or a girl. Her chances of surviving the next year were almost exactly the same. The "battery drain" of one son wasn't enough to break the system.

However, as the family got bigger, the story changed.

• The Tipping Point: For mothers with around five children (the average for that time), having more sons started to take a toll.
• The Cost: For every extra son in a family of five, the mother's chance of surviving the year after her last birth dropped by about 0.4%.
• The Analogy: Think of it like carrying heavy groceries. Carrying one heavy bag (one son) is fine. But if you have to carry five heavy bags (five children), and three of them are extra-heavy (sons), your back starts to ache, and you're more likely to drop something or get hurt. The stress accumulates.

2. The Mystery Reversal: Why did the risk go away for huge families?

Here is the twist. When the researchers looked at mothers with very large families (10+ children), the pattern flipped. Suddenly, mothers with more sons seemed to have better survival rates than those with daughters.

Why? It's not because sons suddenly became "superheroes" who saved their moms.

The researchers believe this is a case of "Survival of the Fittest" (or rather, the Frailty Filter).

• Imagine a race where the runners are carrying heavy loads.
• The runners who are naturally weaker or sicker (frailer) drop out of the race early if they are carrying the heavy "son" loads.
• By the time you get to the finish line (mothers with 10+ children), the only runners left are the ones who were incredibly strong to begin with.
• So, the group of "mothers with many sons" looks super healthy, not because the sons helped them, but because the weaker mothers with sons had already passed away earlier in life. The data is biased because the "weak" ones are missing from the group.

3. The Big Lesson: Don't Ignore the Short-Term

The most important takeaway from this paper is about how we study human history and biology.

For decades, scientists mostly looked at women who lived to be very old (post-menopause) to see if having sons was "bad" for them. They often found mixed or weak results.

• The Flaw: By only looking at the women who survived to old age, they were ignoring the women who died young.
• The Result: It's like trying to judge how dangerous a car is by only looking at the drivers who survived a crash. You miss the people who didn't make it.

This study shows that the "cost" of having sons happens immediately, during the reproductive years. It's a short-term tax on the mother's health that accumulates with every son, especially in large families. If we only look at the survivors, we miss the real cost.

Summary

• The Hypothesis: Sons are biologically "expensive" to raise.
• The Discovery: In pre-industrial times, having many sons increased a mother's risk of dying shortly after her last birth, but only if she already had a large family.
• The Twist: For the biggest families, the risk seemed to disappear, but that was likely because the weaker mothers had already died, leaving only the strongest ones behind.
• The Takeaway: We need to stop ignoring the short-term dangers of childbirth. To truly understand human evolution and family size, we must look at the whole picture, including the mothers who didn't survive to old age.

Technical Summary

1. Problem Statement and Research Gap

The study addresses the "expensive son hypothesis," which posits that in species with male-biased size dimorphism (where males are larger), mothers incur higher fitness costs raising sons compared to daughters. While this hypothesis is well-supported in non-human mammals, evidence in humans is limited and contradictory.

• Current Limitations: Previous human studies have primarily focused on long-term fitness costs (e.g., reduced post-menopausal lifespan) or reproductive factors (e.g., delayed subsequent births).
• The Gap: There is a lack of understanding regarding short-term survival costs (specifically within one year of birth) during a mother's reproductive years. Existing studies focusing only on mothers who survive to menopause may suffer from selection bias (survivorship bias), potentially underestimating the true physiological costs of bearing sons, as frail mothers may die before reaching menopause.

2. Methodology

Study Population and Data:

• Source: Historical life-history records from preindustrial Finland (before 1890).
• Sample: 5,456 mothers whose last birth occurred between 1750 and 1889.
• Data Origin: Lutheran church records (births, marriages, deaths, religious attendance) allowing for the reconstruction of complete individual life histories.
• Context: A high-mortality, low-standard-of-living environment where infectious diseases were common, and maternal mortality was significant.

Variables:

• Response Variable: Binary survival status (survived vs. died) within one year of the last birth.
• Predictors:
  • Sex of the last birth (Male vs. Female).
  • Proportion of total children who were sons.
  • Total number of children (Parity).
  • Socioeconomic status (Lower, Intermediate, Upper).
  • Covariates: Mother's age at last birth, multiple birth status (twins/triplets), region, and birth year.

Statistical Analysis:

• Model: Bayesian Generalised Linear Mixed Models (GLMMs) with a logit link function.
• Software: R (v4.3.2) using brms and Rstan.
• Approach:
  • Tested for interactions between the proportion of sons, total number of children, and socioeconomic status.
  • Included random effects for birth year and family (in supplementary analyses) to account for unobserved heterogeneity.
  • Used Probability of Direction (Pd) to assess significance (Pd > 0.975 considered significant).
  • Conducted posterior predictive checks and convergence diagnostics (Rhat, trace plots).

3. Key Contributions

• Temporal Focus: Shifts the focus from long-term post-menopausal survival to immediate, short-term survival costs (within 12 months of birth) during the reproductive window.
• Non-Linear Interaction: Demonstrates that the cost of sons is not static but is family-size dependent, revealing a complex, non-linear relationship between the number of sons and maternal survival.
• Methodological Correction: Highlights how excluding mothers who die before menopause introduces collider bias, leading to underestimates of reproductive costs and potentially reversing observed trends in older cohorts.

4. Key Results

• Overall Mortality: 8% of mothers died within one year of their last birth; half of these deaths occurred within the first 42 days (post-natal period).
• Last Birth Sex: The sex of the last child alone did not significantly predict survival (mothers with a last-born son actually showed a non-significant trend toward higher survival).
• Accumulated Costs (The Core Finding):
  • Low Parity: For mothers with few children (e.g., 1 child), the proportion of sons had no significant effect on survival.
  • Medium Parity (Peak Cost): As family size increased, the cost of sons became apparent. Among mothers with 5 children (the median), having more sons significantly reduced survival.
    • Magnitude: Approximately 0.4% lower survival per son.
    • Comparison: Mothers with 4 children who had only sons had a 2.1% lower survival probability compared to those with only daughters.
  • High Parity (Reversal): Among mothers with very large families (e.g., 10+ children), the trend reversed. Mothers with more sons showed higher (though statistically non-significant in some models) survival probabilities than those with more daughters.
• Socioeconomic Status: No significant interaction was found between socioeconomic status and the cost of sons; the survival costs were consistent across all classes.
• Other Factors:
  • Older age at last birth correlated with higher survival.
  • Multiple births (twins/triplets) significantly reduced survival.
  • Mothers in Northern Finland had lower survival rates than those in Central/Southwestern regions.

5. Significance and Discussion

• Physiological Accumulation: The results suggest that the cost of sons is not due to a single event (like a difficult birth) but rather an accumulation of physiological costs (e.g., immune suppression from testosterone exposure, chronic inflammation) that manifests as increased mortality risk in the year following birth.
• Selective Disappearance (Frailty Hypothesis): The reversal of costs at high family sizes (where sons appear "beneficial" for survival) is likely due to selective disappearance. Frailer mothers who bear sons are more likely to die earlier, leaving a surviving cohort of mothers with many sons who are inherently more robust ("survivors"). This creates a statistical artifact where high-parity mothers with sons appear to have better survival.
• Implications for Evolutionary Theory:
  • The study validates the expensive son hypothesis in humans but emphasizes that costs are time-dependent and cumulative.
  • It argues that studies focusing solely on post-menopausal women are biased. By excluding mothers who died during reproductive years, previous research may have missed the most severe fitness costs and misinterpreted the relationship between offspring sex and maternal longevity.
• Future Directions: The authors call for future research to explicitly model frailty and non-linear survival costs to better understand life-history trade-offs and the evolutionary drivers of human family size and sex ratios.