Ethnic and Social Health Inequalities in Body Mass Index Trajectories through Childhood and Adolescence: A Longitudinal Population-Based Study in Leicestershire UK

This longitudinal study of over 5,500 children in Leicestershire reveals that BMI trajectories diverge as early as age two, identifying South Asian ethnicity and socioeconomic deprivation as key risk factors for childhood-onset obesity and highlighting the critical importance of early-life interventions, including breastfeeding and physical activity, to address these health inequalities.

The Gist

Imagine a giant, long-term race where thousands of children run from the starting line of birth all the way to the finish line of age 18. Usually, we just look at who is winning or losing at the end. But this study is different: it put on a pair of "time-travel glasses" to watch exactly how every child's body changed along the way.

The researchers in Leicester, UK, followed over 5,500 children (both White and South Asian) to see how their Body Mass Index (BMI)—a simple measure of weight relative to height—changed over time. They wanted to find out: When do kids start gaining too much weight? And who is at risk?

Here is the story of what they found, explained simply.

🗺️ The Five "Body Maps"

Instead of seeing a random mix of weights, the researchers discovered that children naturally fall into five distinct "maps" or paths as they grow up. Think of these like five different hiking trails:

1. The Steady Path (47% of kids): These children stay at a healthy, normal weight all the way from babyhood to adulthood. It's a smooth, flat road.
2. The Low-Altitude Trail (30% of kids): These children stay consistently lighter than average, but they never get dangerously thin. They are just "smaller" hikers.
3. The Early Climb & Descent (8% of kids): These kids were a bit heavy as babies and toddlers, but then they naturally slimmed down as they grew older, returning to a healthy weight by their teens.
4. The Early Detour (4% of kids): This is the dangerous path. These children started gaining weight too fast very early on (between ages 2 and 4). By age 7, they were already obese. The "detour" happened before they even started school.
5. The Late Climb (11% of kids): These kids were fine in early childhood, but around age 4 to 6, they started gaining weight and became overweight by their early teens.

🚩 The Warning Signs: Who is at Risk?

The study found that certain factors act like "headwinds" that push children onto the heavier trails (the Early Detour or Late Climb).

• The Ethnicity Factor: South Asian children were more likely to end up on the heavier trails than White children, even when other factors were the same. It's like the terrain was slightly steeper for them, requiring extra attention.
• The "Wallet" Factor: Children from poorer families (socioeconomically deprived) and those with parents who had less education were more likely to take the "Early Detour" into childhood obesity. Money and education act like a map; without them, it's harder to navigate healthy choices.
• The "Mom's Habits" Factor:
  • Smoking: If a mother smoked while pregnant, the child was more likely to take the "Early Detour."
  • Breastfeeding: If a baby wasn't breastfed, the risk of that early weight gain went up. Breastfeeding is like a "shield" against early obesity.
• The "Movement" Factor: This is the good news! Kids who played actively (at least 4 hours a week) were much less likely to get stuck on the heavy trails. Exercise is the "brake pedal" that keeps weight in check.

⏰ The Critical Windows: When to Act

The most important discovery is timing.

Many people think obesity starts in the teenage years. This study says no, the train leaves the station much earlier.

• The "Early Detour" (Childhood Obesity): The path diverges from the healthy path as early as age 2 to 4.
• The "Late Climb" (Adolescent Overweight): The path starts to tilt upward between ages 4 and 6.

The Analogy: Imagine a garden. If you wait until the weeds (obesity) are tall and thick to pull them out, it's very hard. This study tells us we need to pull the weeds when they are just tiny sprouts (ages 2–6). If we wait until the child is 12, the roots are already too deep.

💡 The Big Takeaway

This research is a wake-up call for parents, doctors, and policymakers.

1. Don't wait: We need to start talking about healthy weights when kids are toddlers, not just teenagers.
2. Help the vulnerable: We need to give extra support to South Asian families and families with less money, as they are fighting an uphill battle.
3. Move more: Getting kids to run and play is one of the most powerful tools we have to stop the "Early Detour."

In short, the study shows that obesity isn't just a "teenage problem" or a "willpower problem." It's a journey that starts very early, and with the right help at the right time, we can keep more children on the healthy, steady path.

Technical Summary

1. Problem Statement

Obesity in the UK is disproportionately distributed across socioeconomic and ethnic lines, with higher prevalence observed in children from ethnic minorities and socioeconomically deprived areas. While cross-sectional data identifies who is affected, it fails to reveal when and how obesity develops over time. Existing longitudinal studies on BMI trajectories are often limited by small sample sizes, lack of ethnic diversity (specifically South Asian populations), or failure to track development beyond early childhood. There is a critical need to identify the specific "critical windows" where BMI trajectories diverge to inform targeted, early-life interventions.

2. Methodology

Study Design and Population

• Cohort: The study utilized data from the Leicester Respiratory Cohorts, a population-based study including 10,350 White and South Asian children born in Leicestershire, UK, between 1985 and 1997.
• Inclusion Criteria: 5,571 participants were included in the final analysis. Inclusion required at least three BMI measurements between ages 0–18 years, with at least one measurement between 0–1 years and one between 3–10 years.
• Data Sources: BMI data was derived from healthcare records (well-child visits), clinical study visits, and parent/self-reported questionnaires.

Variables

• Outcome: Body Mass Index (BMI) trajectories from age 0 to 18. Raw BMI was used rather than z-scores to better capture adiposity changes.
• Exposures:
  • Demographic: Sex, ethnicity (White vs. South Asian), birth year.
  • Socioeconomic: Parental education, area-based urbanization (inner city vs. suburbs/rural), and the Townsend Index of deprivation.
  • Perinatal: Maternal age, birth order, gestational age, and birth weight.
  • Lifestyle: Maternal smoking during pregnancy, breastfeeding duration, secondhand smoke exposure, physical activity levels, and dietary intake (fruit, vegetables, crisps, sweets).

Statistical Analysis

• Trajectory Modelling: The authors employed Group-Based Trajectory Modelling (GBTM) using the lcmm package in R. This method identifies discrete latent classes of BMI development.
  • A grid search determined the optimal parametric function (a 2nd-degree spline with knots at 0.7 and 9.8 years).
  • Models were tested for 1 to 7 classes; a 5-class model was selected based on fit indices (AIC, BIC), entropy, class size, and biological plausibility.
• Risk Factor Analysis: Multinomial logistic regression was used to identify associations between exposures and trajectory classes.
  • Inverse Probability Weighting (IPW): Applied to correct for selection bias arising from missing data and non-random exclusion of participants.
  • Class Membership Uncertainty: Participants were weighted by their posterior probability of belonging to a specific trajectory class.
  • Adjustments: Models were adjusted for demographic, socioeconomic, perinatal, and lifestyle factors.

3. Key Contributions

• Granular Longitudinal Data: Provides one of the few large-scale, longitudinal analyses of BMI trajectories specifically comparing White and South Asian children from birth to late adolescence (18 years).
• Identification of Critical Windows: Pinpoints the exact ages where BMI trajectories begin to diverge, offering specific targets for public health intervention.
• Methodological Rigor: Utilizes GBTM combined with IPW to handle missing data and selection bias, providing more robust estimates of risk factors compared to standard cross-sectional or simple longitudinal analyses.
• Ethnic Specificity: Explicitly models South Asian children, a group often underrepresented in obesity trajectory research, revealing distinct risk patterns that differ from the general population averages.

4. Key Results

BMI Trajectory Classes
Five distinct trajectories were identified across both sexes and ethnicities:

1. Stable Normal BMI (47%): Followed the 50th centile of UK-WHO growth charts.
2. Persistent Low BMI (30%): Consistently below the 50th centile but above the 2nd centile (not underweight).
3. Early Overweight Resolving (8%): Above the 91st centile in infancy/childhood but resolved toward the 50th centile in adolescence.
4. Childhood Onset Obesity (4%): Diverged from normal BMI at 2–4 years, becoming overweight at age 5 and obese at age 7.
5. Adolescent Onset Overweight (11%): Began increasing at 4–6 years, becoming overweight at age 11.

Risk Factors and Inequalities

• Ethnicity: South Asian children had a significantly higher risk of:
  • Childhood onset obesity (aOR: 1.66; 95% CI 1.08–2.53).
  • Adolescent onset overweight (aOR: 1.29; 95% CI 0.98–1.71).
  • Persistent low BMI (aOR: 1.69).
• Socioeconomic Status: Children from deprived backgrounds (highest vs. lowest Townsend quintile) and those with less educated parents (compulsory vs. higher education) were at increased risk of childhood onset obesity (aOR: 1.66 and 1.67, respectively).
• Perinatal Factors:
  • Maternal Smoking: Increased risk of childhood onset obesity (aOR: 1.50) but was associated with a lower risk of persistent low BMI (likely due to lower birth weight followed by rapid catch-up growth).
  • Breastfeeding: Absence of breastfeeding increased the risk of childhood onset obesity (aOR: 1.56).
  • Birth Weight: Higher birth weight was associated with early overweight resolving and childhood onset obesity.
• Lifestyle Factors:
  • Physical Activity: High levels of vigorous activity (≥4 hours/week) significantly reduced the risk of childhood onset obesity (aOR: 0.64) and adolescent onset overweight (aOR: 0.75).
  • Diet: Lower fresh fruit intake was weakly associated with adolescent onset overweight.

5. Significance and Implications

• Timing of Intervention: The study demonstrates that BMI divergence occurs much earlier than previously thought. Interventions must target the preschool period (2–4 years) for obesity prevention and the early school years (4–6 years) for overweight prevention.
• Targeted Public Health Policy: Current UK obesity strategies often lack specific focus on ethnic and socioeconomic disparities. The findings necessitate policies that prioritize South Asian children and those from socioeconomically deprived backgrounds, as they face compounded risks.
• Modifiable Factors: The identification of maternal smoking, lack of breastfeeding, and low physical activity as modifiable risk factors provides clear actionable targets for clinicians and policymakers.
• Limitations: The authors note potential underreporting of BMI in adolescence (self-reported data) and the possibility of residual confounding due to unmeasured genetic factors. However, the use of IPW and large sample size strengthens the validity of the findings.

In conclusion, this study provides robust evidence that ethnic and social health inequalities in BMI are established very early in life. It calls for a shift in public health strategy toward early-life, targeted interventions that address the specific needs of high-risk ethnic and socioeconomic groups.