Mental health and educational attainment: Replicating diminishing associations in an England cohort

This study replicates findings from Norway by demonstrating that the educational attainment gap between students with ADHD or internalising disorders and their peers in South East London has narrowed over time, a trend that cannot be explained by an earlier age at first diagnosis.

The Gist

The Big Picture: A Race That's Getting Fairer

Imagine a long-distance race called "School." In this race, some runners have a heavy backpack on (a mental health diagnosis like ADHD or anxiety), while others run with empty hands.

For a long time, we knew that runners with these heavy backpacks tended to finish the race further behind the others. They struggled to keep up, and the gap between them and the "empty-handed" runners was wide.

But here is the twist: A recent study from Norway suggested that this gap is actually getting smaller. The runners with backpacks are catching up!

This new paper asks: "Is this happening in England, too? And if so, why?"

The Investigation: Checking the English Team

The researchers in this study acted like detectives. They looked at the records of nearly 150,000 students in South London who went to school between 2009 and 2019. They focused on two specific types of "backpacks":

1. ADHD (trouble focusing and sitting still).
2. Internalising Disorders (trouble with anxiety, depression, or stress).

They compared the exam scores of students with these diagnoses against those without, looking at how things changed over time.

The Findings: The Gap is Closing

1. The Good News:
Just like in Norway, the researchers found that in England, the gap between students with mental health diagnoses and those without is shrinking.

• The Past: If you were born in the early 90s and had ADHD, your exam scores were likely much lower than your peers.
• The Present: If you were born in the early 2000s and had ADHD, your scores are much closer to your peers' scores. The "backpack" is still there, but it seems to be getting lighter, or the runners are getting better at carrying it.

2. The Mystery: Why is this happening?
The researchers had a hunch about why this might be happening. They thought: "Maybe we are diagnosing kids earlier now? If we spot the problem when they are 6 instead of 12, maybe we can help them fix it before the big exams."

Think of it like a leak in a boat. If you find the leak at the start of the journey, you can patch it easily. If you wait until the end, the boat might sink.

The Result: They checked the data, and this theory was wrong.
Even though kids are being diagnosed at younger ages now, that alone doesn't explain why their exam scores are improving. The "leak" isn't being fixed just because we found it earlier. Something else must be helping.

What Could Be Helping? (The Theories)

Since "earlier diagnosis" isn't the answer, the authors suggest a few other possibilities:

• Better Tools in the Classroom: Schools might be getting better at helping these students. Imagine a teacher who used to just say, "Sit still," but now knows how to give a student with ADHD a fidget toy or extra time on a test. The school environment is becoming more "backpack-friendly."
• Less Stigma: In the past, having a mental health diagnosis was like wearing a scarlet letter. Kids might have been too scared to ask for help. Now, with less shame, more kids get support, and schools are more willing to accommodate them.
• The "Milder" Backpack Theory: It's possible that the definition of what counts as a diagnosis has changed. Maybe today, we are diagnosing kids who have lighter backpacks (milder symptoms) than the kids who were diagnosed 20 years ago. If the backpack is lighter, it's easier to run fast.

The "Primary School" Surprise

The researchers also looked at younger kids (around age 10-11). They found something interesting:

• For kids with anxiety/depression, the gap was already closing by the time they were in primary school.
• For kids with ADHD, the gap hadn't closed yet at that age.

This suggests that the "catching up" happens at different times for different struggles.

The Bottom Line

What does this mean for us?

1. Hope: The situation for students with mental health challenges is improving. They are achieving more academically than they did a generation ago.
2. Global Connection: It's amazing that a study in Norway and a study in London found the exact same thing. It proves that this is a real, worldwide trend, not just a fluke in one country.
3. The Mystery Remains: We know the gap is closing, but we don't know exactly which key unlocked the door. Was it better schools? Better doctors? Or just a change in who gets diagnosed?

In short: Students with mental health struggles are running faster than before. We aren't sure exactly why yet, but the finish line looks much more reachable for them than it used to.

Technical Summary

1. Problem Statement

Recent epidemiological data suggests a rising prevalence of mental health disorders among youth, with established negative correlations between these disorders and educational attainment. However, a 2024 study by Nordmo et al. in Norway presented a counter-intuitive finding: while the prevalence of diagnosed mental health disorders (specifically ADHD and internalising disorders) increased, the strength of their association with educational attainment weakened over time. This suggested a "closing attainment gap" where students with diagnoses in later birth cohorts performed relatively better than those in earlier cohorts.

The authors sought to:

1. Replicate the Norwegian findings in an English cohort to determine if this trend is an international phenomenon or specific to Norway's healthcare/education systems.
2. Investigate the mechanism behind this trend, specifically testing the hypothesis that earlier age at first diagnosis allows for earlier intervention, thereby mitigating long-term educational impacts.

2. Methodology

Study Design:

• Retrospective longitudinal cohort study using linked administrative data.
• Data Sources:
  • Health: Clinical Record Interactive Search (CRIS) database from South London and Maudsley NHS Foundation Trust (SLaM), containing anonymized electronic health records for children/adolescents referred to Child and Adolescent Mental Health Services (CAMHS).
  • Education: Department for Education's National Pupil Database (NPD), containing records for all pupils in England's state-maintained schools.
• Linkage: Individual-level linkage between CRIS and NPD, refreshed up to the 2018/19 academic year.

Sample:

• Population: $n=149,841$ pupils born between 1993 and 2003.
• Inclusion Criteria: Resided in the SLaM catchment area (South East London), captured in NPD Key Stage 4 data, and were age 15 at the start of their Key Stage 4 year.
• Exclusion: Pupils with missing Year 11 attainment data ($n=285$, 0.2%).

Variables:

• Outcome: Year 11 attainment (GCSE exams, typically age 15–16). Measured as capped point scores (best 8 grades), standardized as z-scores within each academic year to control for scoring changes over time.
• Exposures:
  • Diagnosis: Presence of ADHD (ICD-10 F90.0, F90.1) or Internalising Disorders (ICD-10 F32, F33, F40, F41, F43; covering depression, anxiety, PTSD) diagnosed before the end of Year 11.
  • Birth Year: Used as a proxy for cohort effects.
  • Age at First Diagnosis: Calculated from the earliest diagnosis date in CRIS.
• Covariates: Gender (binary).

Statistical Analysis:

• Primary Analysis: Linear regression models with Year 11 z-score as the outcome.
  • Tested main effects of diagnosis and birth year.
  • Tested interaction terms (Diagnosis $\times$ Birth Year) to see if the association between diagnosis and attainment changed over time.
  • Stratified analyses by diagnosis status to visualize trends.
• Secondary Analysis (Mechanism): Restricted to diagnosed pupils only. Regressed attainment against birth year and age at first diagnosis to test if earlier diagnosis confounded the birth year effect.
• Post-hoc Exploratory Analysis: Repeated the analysis for Key Stage 2 (Year 6, age 10–11) to see if the trend existed at an earlier educational stage.
• Software: Stata 17.0 (analysis) and R 4.2.1 (visualization).

3. Key Results

Descriptive Statistics:

• Prevalence: ADHD diagnosis prevalence rose from 0.1% (1993 cohort) to 1.1% (2003 cohort). Internalising disorder prevalence rose from 0.7% to 1.9%.
• Diagnosis Age: The average age at first diagnosis declined over time for both disorders (ADHD: mean 11.8 years; Internalising: mean 13.8 years).

Primary Findings (Year 11 Attainment):

• Main Effect: On average, students with ADHD ($b = -1.29$) or internalising disorders ($b = -0.52$) had significantly lower attainment than non-diagnosed peers.
• Interaction Effect: There was a significant interaction between diagnosis and birth year ($p < 0.001$ for both).
  • For students without diagnoses, birth year had no significant association with attainment.
  • For students with diagnoses, birth year was positively associated with attainment ($b = 0.06$ for ADHD; $b = 0.05$ for internalising).
• Conclusion: The attainment gap between diagnosed and non-diagnosed students is narrowing over successive birth cohorts.

Mechanism Analysis (Age at Diagnosis):

• When restricting the sample to diagnosed students and adjusting for age at first diagnosis, the positive association between birth year and attainment remained significant.
• Age at first diagnosis was not a significant predictor of attainment in these models ($p > 0.05$).
• Implication: The narrowing gap is not explained by the trend of diagnosing children at younger ages.

Post-hoc Analysis (Year 6 Attainment):

• ADHD: No significant interaction between birth year and diagnosis at Year 6.
• Internalising Disorders: A significant narrowing gap was observed at Year 6 as well, with a stronger effect size ($b = 0.11$) than at Year 11 ($b = 0.05$).
• Timing: Only 14.5% of students with internalising disorders diagnosed by Year 11 had a diagnosis by Year 6, compared to 41.1% for ADHD.

4. Key Contributions

1. International Replication: Successfully replicated the "diminishing association" finding from Norway in a distinct English cohort, suggesting this is a robust, cross-national trend rather than a country-specific artifact.
2. Ruling Out a Mechanism: Provided empirical evidence that earlier age of diagnosis is not the driver of the improved educational outcomes for diagnosed cohorts, challenging the assumption that earlier intervention automatically translates to better academic recovery in this context.
3. Methodological Demonstration: Demonstrated the feasibility of replicating complex administrative data linkage studies across different countries (Norway vs. UK) despite differences in healthcare systems, data structures, and educational metrics (GPA vs. GCSE).
4. Granular Insight: Highlighted that the trend may manifest differently across diagnostic categories (stronger at Year 6 for internalising disorders) and developmental stages.

5. Significance and Implications

• Public Health & Policy: The findings suggest that despite rising prevalence and earlier diagnosis, the relative educational disadvantage for students with mental health disorders is decreasing. This could be due to:
  • Diagnostic Thresholds: Lower thresholds may be diagnosing less severely impaired students who naturally perform better academically.
  • Systemic Improvements: Increased school-based mental health support, destigmatization, and tailored educational interventions (e.g., Special Educational Needs provision) may be mitigating the impact of disorders.
  • Service Capacity: Increased resources may be reaching more students, though the study notes a lack of data to confirm if treatment efficacy has actually improved.
• Research Direction: The study underscores the need to move beyond simple diagnosis/attainment correlations to investigate why the gap is closing. Future research should explore:
  • The role of school-based interventions.
  • Socioeconomic stratification of these trends.
  • The divergence between administrative data (showing narrowing gaps) and survey data (often showing widening gaps).
• Limitations: The study is limited to state-maintained schools in South London (excluding private school students and those not in education), relies on secondary care diagnoses (potentially missing primary care cases), and cannot establish causality.

Conclusion: The study confirms a positive trend where the educational penalty associated with mental health diagnoses is lessening in England, but explicitly rules out "earlier diagnosis" as the cause, pointing instead toward broader systemic or diagnostic shifts.