Educational attainment and genetic liability to overweight: Body mass index across the adult life course in females and males

This study of 69,314 adults reveals that higher educational attainment significantly attenuates the effect of genetic liability on BMI in females throughout adulthood, whereas this protective socioeconomic moderation is absent in males.

The Gist

The Big Question: Are We Born with a "Fat Gene" that Can't Be Changed?

Imagine your body's weight is like a car. You have a genetic engine (your DNA) that determines how fast the car could go. Some people have engines built for speed (a high genetic risk for being overweight), while others have engines built for cruising (a low genetic risk).

For a long time, scientists wondered: Does the road you drive on matter? In this study, the "road" is your education level (how much school you finished).

The researchers wanted to know: If you have a "fast engine" (high genetic risk), does going to college (getting a higher education) act like a speed bump that slows you down? Or does the engine just roar regardless of the road?

The Study: A Long Road Trip in Norway

The researchers took a massive look at 69,000 people in Norway over several decades. They checked their DNA, measured their weight (BMI) many times as they aged, and looked at their education levels. They treated men and women separately because, as we'll see, their journeys are very different.

The Findings: Two Different Maps

The study found that the "road" (education) changes the journey, but only for women.

1. The Women's Journey: Education is a Shield

Think of women with high genetic risk for weight gain as having a car with a very powerful, hard-to-control engine.

• Without a good road (Low Education): The powerful engine runs wild. The car speeds up quickly, and the women gain significant weight. The gap between those with "fast engines" and "slow engines" is huge.
• With a good road (High Education): When these women go to college, it's like putting their car on a smooth, well-paved highway with speed limits. The powerful engine is still there, but the environment keeps it in check. The women with high genetic risk don't gain as much weight as they would have otherwise.

The Analogy: Imagine a garden. Some women have seeds that grow into giant, aggressive weeds (high genetic risk).

• If the garden is neglected (low education), the weeds take over completely.
• If the garden is well-tended with a fence and a gardener (high education), the weeds are kept in check. The gardener (education) successfully buffers the aggressive nature of the seeds.

2. The Men's Journey: The Road Doesn't Matter

Now, look at the men. They also have different engines (some fast, some slow).

• The Result: Whether a man has a college degree or only finished high school, his weight follows the same pattern. The "road" doesn't seem to change how his engine performs.
• The Analogy: Imagine men's weight is like a heavy truck. Whether the truck is driving on a dirt road or a super-highway, it moves at roughly the same speed. The truck is so heavy (influenced by other factors like biology or work culture) that the quality of the road (education) doesn't really slow it down or speed it up. The gap between men with "fast engines" and "slow engines" stays the same regardless of their schooling.

How Weight Changes Over a Lifetime

The study also tracked how weight changes as people get older, like watching a movie of a person's life:

• Women: Their weight climbs steadily like a gentle hill until they hit about age 60 (often due to life events like pregnancy or menopause). After 60, the hill goes down, and they start losing a little weight. This pattern was the same whether they were highly educated or not, but the highly educated women started from a lower point on the hill.
• Men: Their weight shoots up quickly in their 20s and 30s, then flattens out like a plateau until age 60, and then slowly declines. Again, education didn't change this shape; it just didn't seem to protect them from their genetic risks the way it did for women.

Why the Difference? (The "Why" Behind the "What")

The researchers suggest a few reasons why education helps women but not men:

1. The "Obesogenic" Trap: Women with lower education often face a "perfect storm" of stress, lack of time for exercise, and unhealthy food options due to caregiving duties and lower-paying jobs. Education gives women the resources to escape this trap.
2. The Male Baseline: Men's weight is often less tied to their education level to begin with. The gap between a working-class man and a college-educated man regarding weight is already small. So, adding more education doesn't create a big "shield" because there wasn't much danger to shield against in the first place.
3. Different Pressures: Society puts different pressures on men and women. A highly educated woman still faces unique societal pressures (like balancing work and family) that education helps her navigate, whereas men's weight might be driven more by different factors like occupational hazards or cultural norms that education doesn't fix.

The Bottom Line

• For Women: Education is a powerful tool. It acts like a protective shield, helping women with a genetic tendency to be overweight stay healthier. It doesn't erase their genes, but it gives them the environment to manage them better.
• For Men: Education doesn't seem to offer this same genetic shield. Their weight is influenced by their genes, but their schooling level doesn't change how those genes play out.

The Takeaway: We can't change our DNA, but we can change our environment. For women, getting an education is a proven way to "tame" the genetic engine. For men, the study suggests we need to look for other factors (like workplace culture or specific health habits) to understand why their weight behaves the way it does.

Technical Summary

1. Problem Statement

The study addresses the complex interplay between socioeconomic status (SES), genetic susceptibility, and body mass index (BMI). While it is well-established that socioeconomic disadvantage correlates with higher BMI, it remains unclear whether SES acts as a moderator for genetic liability to high BMI. Specifically, the authors investigate:

• Does Educational Attainment (EA) moderate the association between genetic liability (measured via a Polygenic Index) and BMI trajectories across adulthood?
• Do these moderating effects differ significantly between females and males?
• How do these gene-environment interactions evolve over the life course, considering distinct biological and social transitions for each sex?

2. Methodology

Study Design and Population:

• Cohort: The study utilized data from the Trøndelag Health Study (HUNT), a large, population-based cohort in Norway.
• Sample Size: 69,314 participants (genotyped) who participated in at least one of four survey waves (HUNT1–HUNT4, spanning 1984–2019).
• Longitudinal Nature: The design included repeated BMI measurements over a period exceeding 30 years, allowing for the modeling of life-course trajectories.

Data Collection and Variables:

• Phenotypes:
  • BMI: Calculated from measured height and weight (kg/m²).
  • Educational Attainment (EA): Categorized into three levels based on the Norwegian system: Low (≤10 years), Middle (11–12 years), and High (>12 years).
  • Genetic Liability: A BMI Polygenic Index (BMI PGI) was constructed using approximately 1.84 million genetic variants, derived and validated by Khera et al. (2018). The PGI explained 6.54% of the variance in measured BMI.
• Covariates: Analyses were adjusted for the first 20 genetic principal components and genotyping batch.

Statistical Analysis:

• Modeling Approach: Linear mixed-effects models (LMM) with random intercepts and slopes were used to account for repeated measures.
• Non-linearity: Both BMI PGI and Age were modeled as cubic B-splines (4 degrees of freedom) to capture non-linear relationships.
• Interaction Testing:
  • Model 1: Tested associations between BMI PGI, Age, Sex, and EA.
  • Model 2: Tested the two-way interaction: BMI PGI × EA (to see if education modifies genetic risk).
  • Model 3: Tested the three-way interaction: BMI PGI × EA × Age (to see if the moderation effect changes over time).
• Stratification: Analyses were stratified by sex to detect sex-specific patterns.
• Significance Testing: Type III F-tests with Satterthwaite's approximation were used for interaction terms. Post-hoc contrasts were adjusted using Tukey's method.

3. Key Results

Genetic Liability and BMI Trajectories:

• The relationship between BMI PGI and observed BMI was non-linear, with steeper slopes in the upper deciles of genetic risk.
• Sex Differences: The association between PGI and BMI was significantly modified by sex ($p < 0.001$). Females showed a steeper increase in BMI across PGI deciles compared to males.
• Life-Course Patterns:
  • Females: BMI increased steadily from age 20 to 60, followed by a decline.
  • Males: BMI increased rapidly in early adulthood (20–40), stabilized until age 60, and then declined slightly.

Educational Attainment as a Moderator:

• Females: EA significantly moderated the effect of BMI PGI on BMI ($p = 0.003$).
  • Highly educated females showed a weaker association between genetic risk and BMI compared to low-educated females.
  • The difference in BMI between the top and bottom PGI deciles was 0.99 kg/m² smaller in highly educated females compared to those with low education.
  • This protective "buffering" effect was consistent across adulthood (ages 20–80).
• Males: EA did not significantly moderate the genetic effect ($p = 0.089$).
  • The difference in BMI between top and bottom PGI deciles was negligible (-0.16 kg/m²) between high and low-educated males.
  • Genetic influences on BMI trajectories were consistent across all educational groups for males.

Three-Way Interaction (Time):

• There was no significant three-way interaction (BMI PGI × EA × Age) in either sex, indicating that the moderating effect of education (where it exists) remains stable throughout adulthood rather than strengthening or weakening with age.

4. Key Contributions

• Sex-Specific Gene-Environment Interaction: The study provides robust evidence that the protective effect of higher education against genetic obesity risk is exclusive to females. Previous studies often treated sex as a covariate, potentially masking this critical divergence.
• Longitudinal Validation: By utilizing over 30 years of data, the study confirms that this buffering effect is not a transient phenomenon but a stable feature of the female life course.
• Mechanistic Insight: The findings suggest that for women, higher education fundamentally alters exposure to "obesogenic" environments (e.g., labor markets, caregiving burdens, health behaviors) in a way that mitigates genetic risk. For men, the educational gradient in BMI is smaller, suggesting that additional schooling offers less incremental protection against genetic liability.

5. Significance and Implications

• Public Health Strategy: Obesity prevention strategies cannot be "one-size-fits-all." Interventions targeting genetic risk factors may need to be tailored by sex. For women, improving educational access and socioeconomic conditions appears to be a potent mechanism for reducing the expression of genetic obesity risk.
• Understanding Inequality: The results highlight how social stratification (education) interacts with biology (genetics) to produce health disparities. The fact that this interaction is absent in men suggests that the mechanisms linking SES to health are fundamentally different for men and women.
• Future Research: The authors call for further investigation into the specific biological and social mechanisms driving this sex difference, particularly regarding how gendered social roles and structural constraints modulate gene expression.

Limitations Noted:

• The cohort is of European ancestry, limiting generalizability.
• BMI does not distinguish between fat and lean mass.
• Age and period effects (secular trends in obesity) are difficult to fully disentangle in longitudinal data.