The Fonagy review in charts: 13 findings for ADHD and autism services

The Fonagy review (the independent review into mental health conditions, ADHD and autism, published by the Department of Health and Social Care on 9 October 2026) runs to 599 pages. It draws on GP records, national surveys, NHS data and studies it commissioned, and most of its numbers sit inside long chapters.

We picked the 13 findings a clinic lead or clinician can act on: who is now coming forward, how complex their needs are, where people are lost before they are seen, and what happens to them later. Each chart says what it found and what it means for your service.

What this means for you

• The people you assess are changing: more young adults, more women, more inattentive presentations and more people who already see themselves as autistic or as having ADHD.
• Several conditions at once is now the usual case, and the review points to combined pathways and earlier screening for the second condition.
• Independent providers now do most NHS-funded ADHD assessment, and the review wants the same standards and outcome data from every provider.

For the review's recommendations, see the 10 recommendations and what each means for providers. For the full diagnosis data, see ADHD and autism diagnosis trends providers should plan for.

1. Far more young people say they are autistic, and most adults who do have no diagnosis#

In the GP Patient Survey, the share of 16 to 24-year-olds who say they are autistic rose from about 3.4% in 2018 to 8.9% in 2025. In the Adult Psychiatric Morbidity Survey, 3.4% of adults say they are autistic and 1.2% report a professional diagnosis, so nearly three adults identify as autistic for every one diagnosed.

Set expectations before the assessment starts, and plan a proper outcome for people who are not diagnosed. A clinician quoted in the review describes people whose understanding has "solidified" over a long wait: "if you then reach a conclusion that they don't have autism, it can be devastating for a patient." Self-identification is neither a reason to turn a referral away nor a result.

Say they are autistic, GP Patient Survey, EnglandShare of each age group who say they are autistic, 2018 to 2025.20252025: aged 16 to 24, 8.9%, up 5.5 points since 2018; aged 25 and over, 2.0%, up 1.4 points since 2018.Aged 16 to 24, by year: 2018 3.4%, 2019 3.8%, 2020 3.5%, 2021 4.1%, 2022 5.4%, 2023 6.4%, 2024 8.7%, 2025 8.9%. Aged 25 and over, by year: 2018 0.6%, 2019 0.7%, 2020 0.7%, 2021 0.8%, 2022 1.0%, 2023 1.3%, 2024 1.7%, 2025 2.0%. Years before 2025 are approximate.Aged 16 to 248.9%Up 5.5 points since 2018Aged 25 and over2.0%Up 1.4 points since 20180%2%4%6%8%10%8.9%2.0%20182019202020212022202320242025Say they are autistic, against report a diagnosisAdults (APMS): 3.4% say they are autistic, 1.2% report a professional diagnosis. Nearly 3 to 1.Adults (APMS)Nearly 3 to 1Say they are autistic3.4%Report a professional diagnosis1.2%Secondary and FE students (OxWell 2025): 11.4% say they are autistic, 5.5% report a professional diagnosis. About 2 to 1.Secondary and FE students (OxWell 2025)About 2 to 1Say they are autistic11.4%Report a professional diagnosis5.5%Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, Figure 8.8, page 162 (GP Patient Survey; 2025 values are printed on the chart, earlier years read from the chart and approximate), and page 161 (Adult Psychiatric Morbidity Survey; OxWell 2025). The GP Patient Survey question does not separate diagnosed from self-identified autism, and the OxWell figure includes students who are on a waiting list or were identified by others. Contains public sector information licensed under the Open Government Licence v3.0.

2. Adults who identify with ADHD now date their difficulties to age 21, not 10#

Among adults who identify with ADHD, the average age they recall their difficulties starting rose from 9.8 years in 2014 to 21.7 years in 2024. The review warns that recall and changing self-understanding limit how far this can be read.

ADHD is a diagnosis that needs symptoms in childhood. So childhood history and evidence from people who knew the person as a child now carry more weight than ever. Build time and tools for them into adult assessments, and record why the childhood-onset criterion is or is not met.

Average age difficulties started, as recalledAdults in England who identify with ADHDAge 21.7Average recalled age when difficulties started: 9.8 years in 2014 and 21.7 years in 2024, 11.9 years later.20142024ChildhoodTeensAdulthood9.8 years21.7 years11.9 years later051015202530 yearsA long-term study in New Zealand found something similar. In the Dunedin cohort, 1,037 people born in New Zealand in 1972 and followed into adulthood, 9 in 10 adults with ADHD had no history of childhood ADHD. Different country, generation and measure, so it is kept off the scale above.Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, section 7.5.2, page 135 (Adult Psychiatric Morbidity Survey 2014 and 2024), and page 136 (Dunedin cohort, Moffitt and colleagues). The figures are for people who identify with ADHD, not people with a diagnosis, and the review says recall bias and changing self-understanding limit interpretation. Contains public sector information licensed under the Open Government Licence v3.0.

3. The rise in adult ADHD symptoms is in inattention; hyperactivity has barely changed#

On the adult ADHD screening questions, the average inattention score rose from about 8.0 in 2007 to 8.6 in 2024, while the hyperactivity score stayed at 4.6. The sharpest single rise was in avoiding or delaying tasks that need a lot of thought, from 12.1% to 18.8% of adults.

Inattentive presentations overlap most with anxiety, depression, poor sleep and stress. Resource differential diagnosis as the core of the assessment, not a formality.

Adult ADHD screening questions, EnglandAverage score for each group of questions2007201420242024: inattention score 8.6, hyperactivity score 4.6. Share of adults who often have each difficulty: Avoiding or delaying tasks that need a lot of thought, 18.8% (2007: 12.1%); Difficulty getting things in order, 26.9% (2007: 21.3%); Trouble wrapping up the fine details, 32.3% (2007: 27.5%); Problems remembering appointments, 30.9% (2007: 29.0%); Fidgeting or squirming, 27.6% (2007: 24.8%); Feeling overly active, as if driven by a motor, 13.3% (2007: 12.9%).Inattention score · 4 questions8.6Up about 0.6 since 2007Hyperactivity score · 2 questions4.6About the same as 2007Share of adults who often have each difficulty. Open circle: 2007.Inattention questionsAvoiding or delaying tasks that need a lot of thought18.8% +6.7Difficulty getting things in order26.9% +5.6Trouble wrapping up the fine details32.3% +4.8Problems remembering appointments30.9% +1.9Hyperactivity questionsFidgeting or squirming27.6% +2.8Feeling overly active, as if driven by a motor13.3% +0.40%10%20%30%Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, Figures 7.7 and 7.8, page 134, and text page 133 (Adult Psychiatric Morbidity Survey 2007, 2014 and 2024, adults aged 16 and over; item percentages are printed on the chart; domain scores for 2024 are printed, earlier years read from the chart and approximate). These are screening questions, not diagnoses, and the review says the increase is much smaller than the rise in diagnosis and service demand. Contains public sector information licensed under the Open Government Licence v3.0.

4. From age 12, more girls and women than boys and men are newly recorded as autistic#

In GP records for 2025, the rate of new autism records among 18 to 34-year-olds was about twice as high for females as for males, and females were ahead in every age band from 12 to 64. The review puts this down to better identification of groups that were missed before, rather than a change in how common autism is.

Check that your tools, training and clinicians' experience fit female and later presentations, and plan adolescent and adult capacity for a mostly female caseload.

New autism records on GP systems, England, 2025Approximate rates per 100,000 a year. Each bar splits the male and female rates. Past the dashed line, the female rate is higher.Female against maleActual ratesAgeMaleFemale0 to 51,2705606 to 111,20069012 to 1765586018 to 2421042025 to 3412024535 to 448715745 to 54629355 to 64323665+65All maleEvenAll femaleFrom 12 to 64, the female rate is higher in every age band. Among 18 to 34-year-olds it is about twice the male rate. Under 12, boys are still recorded about twice as often.Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, Figure 8.3, page 157, and text page 169 (CPRD primary care records; values read from the chart and approximate; ratios are our calculation from those readings). Records show sex, not gender, and the review attributes the change mainly to better recognition of previously overlooked groups rather than a change in underlying prevalence. Contains public sector information licensed under the Open Government Licence v3.0.

5. Three or more conditions is now the most common pattern in adults who screen positive for ADHD#

In 2024, 44.6% of adults screening positive for ADHD traits had three or more types of condition, up from about a third in 2007. Fewer than one in four had ADHD traits alone. These are screening results, not diagnoses.

Treat several conditions at once as the usual case. Plan for longer assessments, a wider mix of skills in the team and real time for differential diagnosis.

Adults screening positive for ADHD, by number of condition typesAdults 16 and over in England, 2007, 2014 and 2024.44.6%44.6% of adults who screened positive for ADHD in 2024 also had at least two other types of condition, the largest group.Three or more typesTwo typesOne type (ADHD traits only)2007 (approximate): three or more types about 34%, two types about 34%, one type (ADHD traits only) about 32%.2014 (approximate): three or more types about 37%, two types about 34%, one type (ADHD traits only) about 29%.2024: three or more types 44.6%, two types 32.6%, one type (ADHD traits only) 22.8%.2007approx.34%34%32%2014approx.37%34%29%202444.6%32.6%22.8%Share of adults screening positive for ADHD traits, by how many of six types of condition they had: common mental disorders such as anxiety and depression, personality disorder, substance dependence, ADHD traits, autistic traits and probable psychosis. Each bar adds to 100%.Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, Figure 9.4, page 198 (Adult Psychiatric Morbidity Survey, adults 16 and over). 2024 values are printed on the chart; 2007 and 2014 values are read from it and approximate. These are screening measures, and the report shows wide confidence intervals for this group. Contains public sector information licensed under the Open Government Licence v3.0.

6. Autistic and ADHD traits rose by more percentage points than any other adult measure#

From 2014 to 2024, the share of adults screening positive for autistic traits rose by 5.9 percentage points and for ADHD traits by 4.2, more than any other measure in the survey. Psychotic disorder and bipolar disorder did not rise. Some measures, such as drug dependence and self-harm, grew faster in relative terms from a lower base.

The growth is not just general distress. Plan neurodevelopmental assessment capacity as a need in its own right, not as a share of wider mental health demand.

Adult mental health measures in England, 2014 and 2024Percentage of adults with each condition or positive screen.20142024Sorted by changePointsAutistic traits (AQ 8+ screen): 14.5% in 2014, 20.4% in 2024, up 5.9 percentage points.Autistic traitsAQ 8+ screen14.5% to 20.4%14.520.4+5.9ADHD traits (ASRS 4+ screen): 9.7% in 2014, 13.9% in 2024, up 4.2 percentage points.ADHD traitsASRS 4+ screen9.7% to 13.9%9.713.9+4.2Non-suicidal self-harm (ever): 6.4% in 2014, 10.3% in 2024, up 3.9 percentage points.Non-suicidal self-harmever6.4% to 10.3%+3.9Drug dependence: 3.8% in 2014, 6.7% in 2024, up 2.9 percentage points.Drug dependence3.8% to 6.7%+2.9Post-traumatic stress disorder (PCL-C screen): 4.4% in 2014, 5.7% in 2024, up 1.3 percentage points.Post-traumatic stress disorderPCL-C screen4.4% to 5.7%+1.3Personality disorder (SAPAS screen): 13.7% in 2014, 14.9% in 2024, up 1.2 percentage points.Personality disorderSAPAS screen13.7% to 14.9%+1.2Obsessive compulsive disorder: 1.6% in 2014, 2.6% in 2024, up 1.0 percentage points.Obsessive compulsive disorder1.6% to 2.6%+1.0Suicide attempt (past year): 0.7% in 2014, 1.0% in 2024, up 0.3 percentage points.Suicide attemptpast year0.7% to 1.0%+0.3Borderline personality disorder: 2.4% in 2014, 2.5% in 2024, up 0.1 percentage points.Borderline personality disorder2.4% to 2.5%+0.1Bipolar disorder (MDQ screen), a severe mental illness: 2.0% in 2014, 1.9% in 2024, down 0.1 percentage points.Bipolar disorderSevere mental illnessMDQ screen2.0% to 1.9%−0.1Psychotic disorder (SCAN interview), a severe mental illness: 0.7% in 2014, 0.4% in 2024, down 0.3 percentage points.Psychotic disorderSevere mental illnessSCAN interview0.7% to 0.4%−0.3Alcohol dependence: 3.4% in 2014, 2.9% in 2024, down 0.5 percentage points.Alcohol dependence3.4% to 2.9%−0.5Antisocial personality disorder: 3.3% in 2014, 2.4% in 2024, down 0.9 percentage points.Antisocial personality disorder3.3% to 2.4%−0.90%5%10%15%20%Change in percentage points. Drug dependence, self-harm and OCD grew faster in relative terms, from much lower starting points.Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, Figure 6.13, pages 96 to 97 (Adult Psychiatric Morbidity Survey 2014 and 2024; values printed on the chart). The chart mixes screening tools, diagnostic interviews and lifetime measures, and age ranges differ by row; a positive screen is not a diagnosis. Contains public sector information licensed under the Open Government Licence v3.0.

7. Children already diagnosed with anxiety, depression or an intellectual disability waited years longer for an autism diagnosis#

In GP records, autism diagnosis in children came around three years later after an earlier anxiety diagnosis, five years after depression and seven years after an intellectual disability. In one city's linked records, the second of autism and ADHD was recorded four to five years after the first.

When a child is referred with anxiety or depression, or already has one neurodevelopmental diagnosis, screen for autism and ADHD at that point. Do not wait for the first diagnosis to run its course.

Delay to autism diagnosis after an earlier diagnosisChildren diagnosed with autism before 18. One block is one year.Year 7After an earlier anxiety diagnosisaround 3 yearsAfter an earlier depression diagnosisaround 5 yearsAfter an earlier intellectual disability diagnosisaround 7 years12345678The same pattern in one city’s linked recordsChildren with both: second of autism and ADHD recorded4 to 5 years after the firstAnxiety or depression recorded laterChildren with a neurodevelopmental condition19.5%Other children8.7%Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, page 204 (CPRD GP records, children diagnosed with autism before 18) and page 299 (linked records from one medium-sized, diverse city). The review gives the delays as rounded figures and does not state the comparison group, so no baseline is drawn; the city figures come from one area. Contains public sector information licensed under the Open Government Licence v3.0.

8. In one city, a combined ADHD and autism pathway had about a third of the wait#

By 2024, the median wait for assessment in one city was 7 months on a combined ADHD and autism pathway, against around 21 to 24 months on the autism-only pathway. Time to diagnosis was shorter too, including against the city's ADHD-only pathway (16 months against 25).

Consider assessing ADHD and autism together in one pathway, not one condition after another. This is one observational comparison in one city: the review says it does not establish cause, and a second combined pathway saw waits rise.

Median waits on combined and single-condition pathways in one cityMonths, by 2024.Combined ADHD and autism pathwaySingle-condition pathwayMonth 50Wait for assessment. Autism-only pathway: median around 21 to 24 months. Combined pathway: median 7 months. Combined took about a third of the wait.Wait for assessmentCombined took about a third of the waitAutism-only21 to 24 monthsCombined7 monthsReferral to diagnosis. Autism-only pathway: median around 50 months. Combined pathway: median 25 months. Combined took half the time.Referral to diagnosisCombined took half the timeAutism-onlyabout 50 monthsCombined25 monthsTime to diagnosis, per patient. ADHD-only pathway: median 25 months. Combined pathway: median 16 months. Combined took 9 months less.Time to diagnosis, per patientCombined took 9 months lessADHD-only25 monthsCombined16 months1 year1 yr2 years2 yrs3 years3 yrs4 years4 yrsMedians, by 2024. A second combined pathway, in another area, saw waits to assessment increase; its median time to diagnosis was 4.9 months overall, rising to around 14 months for adolescents in 2023.Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, Box 14.1, page 299 (review analyses of regional data; medians). The report says "these observational comparisons do not establish causality" and that service models differed; a second combined pathway saw waits to assessment increase. Contains public sector information licensed under the Open Government Licence v3.0.

9. A third of children's referrals were rejected unseen, and three in ten of those came back within a year#

Across ten children's mental health services, 32.3% of 202,244 referrals were rejected without the child being seen, and around three in ten of those children were referred again within a year. Half of all referrals ended in recorded treatment.

A rejection does not end the need. A front door should point each family to support or advice that fits, not only say no.

Referrals to ten children's mental health services in EnglandUnder-18s, financial years ending 2016 to 2019. Shares are of all referrals received.All referrals received202,244202,24467.7%67.7%Seen and assessed136,893 referrals32.3%32.3%Rejected without being seen65,351 referralsTreated50.3% · 101,812Discharged with no recorded treatment14.5% · 29,243Treatment not known2.9% · 5,838About 3 in 10of the children turned away were referred again within a yearThe hatched block is illustrative. The review gives "around three in ten" children turned away, not a count of referrals, so it is drawn as three tenths of the rejected bar.Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, Figure 14.1, page 303, and text, pages 302 to 303 (referrals for under-18s from ten services in England, financial years ending 2016 to 2019; values printed on the chart). These are children's mental health services in general, not ADHD or autism pathways, and referrals still open when the data were extracted are excluded. Contains public sector information licensed under the Open Government Licence v3.0.

10. The deprivation gap in new ADHD diagnoses has closed#

For much of the 2000s, people in the most deprived fifth of areas were newly diagnosed with ADHD at around 2.3 times the rate of the least deprived. By 2025 the ratio was 1.01. The review gives several possible reasons, one of them unequal access to independent assessment.

Parity in new diagnoses is not proof of fair access. Check who your own service reaches, by area deprivation.

New ADHD diagnoses, most against least deprived areas, England, 2000 to 2025GP records. Most deprived fifth of areas against least deprived fifth.Rate of new ADHD diagnoses in the most deprived fifth of areas as a multiple of the rate in the least deprived fifth, by year (values before 2023 read from the chart and approximate): 2000: 1.8×; 2001: 1.5×; 2002: 1.9×; 2003: 2.3×; 2004: 2.7×; 2005: 2.4×; 2006: 2.6×; 2007: 2.5×; 2008: 2.3×; 2009: 2.6×; 2010: 2.4×; 2011: 2.4×; 2012: 2.1×; 2013: 2.0×; 2014: 2.0×; 2015: 2.1×; 2016: 1.9×; 2017: 1.6×; 2018: 1.3×; 2019: 1.3×; 2020: 1.2×; 2021: 1.1×; 2022: 1.0×; 2023: 0.93×; 2024: 0.93×; 2025: 1.01×. The line for everyone with a diagnosis rises from about 1.5× in 2000 to about 2.2× around 2010, then falls to about 1.0× by 2025.1.01×the rate in the least deprived areas, so no gap20251.5×2.0×2.5×3.0×1.0× · no gapNew diagnosesThe gapThe gapAll diagnosesEveryone with a diagnosis0.931.01200020052010201520202025Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, Figure 7.5, page 132, and text pages 131 to 132 (CPRD primary care records; 0.93 for 2023 to 2024 and 1.01 for 2025 are stated in the text, other years read from the chart and approximate). Uncertainty is wide before about 2012, and the review says recorded prevalence remains higher in more deprived populations. Contains public sector information licensed under the Open Government Licence v3.0.

11. ADHD and autism are recorded far less often in South Asian and Black groups, and in one city the gap began at referral#

In GP records for 2025, 0.4% of South Asian people had a recorded ADHD diagnosis, against 1.7% of White people. In one city, South Asian children made up 21% of referrals to assessment pathways but about a third of all children. Once referred, waits showed no consistent difference by ethnicity. The review asks for caution in reading these figures.

Check equity in who reaches your front door and in your referral mix, not only in how long each group waits.

Referrals to assessment in one city, and recorded ADHD and autism by ethnic groupOne city · children referred to assessment pathwaysShare of all childrenShare of referralsWhite British children67% of referrals · about half of childrenSouth Asian children21% of referrals · about a third of children0%25%50%75%100%Once on the pathway, no consistent gap. In this city and a second one, waiting times showed no consistent differences by ethnicity or deprivation.Across England · recorded prevalence in GP records, 2025, % of peopleADHDADHD, Mixed: 2.0%Mixed2.0%ADHD, White: 1.7%White1.7%ADHD, Black: 0.7%Black0.7%ADHD, South Asian: 0.4%South Asian0.4%AutismAutism, Mixed: 2.2%Mixed2.2%Autism, White: 1.9%White1.9%Autism, Black: 1.1%Black1.1%Autism, South Asian: 0.6%South Asian0.6%Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, Box 14.4, page 311 (regional analysis, one city); Figure 7.6 and text, pages 132 to 133, and text, page 172 (CPRD GP records, recorded prevalence 2025, values as stated in the text). The report says the prevalence differences should be interpreted cautiously, and that small subgroups and incomplete ethnicity recording limit the one-city findings. Contains public sector information licensed under the Open Government Licence v3.0.

12. Independent providers now do more than half of NHS-funded ADHD assessments#

Provider estimates put independent providers at more than half of NHS-funded ADHD assessments and more than a third of autism assessments in 2025/26. The providers that responded to a survey reported that 66% to 93% of their autism assessments ended in a diagnosis. There is no comparable NHS figure.

The review wants the same standards, minimum dataset and outcome measures for every provider. Expect to report your outcomes, and be ready to publish them. Our article on the national standards the review proposes covers what your records should show.

Independent providers: share of NHS-funded assessments and autism diagnosis ratesShare of NHS-funded assessments done by independent providers, 2025/26ADHD assessments: independent providers did more than half of NHS-funded assessments in 2025/26. The true share is higher, but not stated.ADHD assessmentsMore than half0halfallAutism assessments: independent providers did more than a third of NHS-funded assessments in 2025/26. The true share is higher, but not stated.Autism assessmentsMore than a third0a thirdallThe bars end at the threshold the review gives and fade out beyond it: the true share is higher, but not stated. Around 90% of this work came through non-contracted Right to Choose, outside local contracts.Autism assessments that ended in a diagnosisIndependent providers that responded: 66% to 93% of autism assessments ended in a diagnosis, range across providers, with averages of 79% to 86%.Independent providers that respondedrange across providersaverages 79% to 86%66%93%NHS servicesNo comparable NHS data0%50%100%Providers varied from fewer than 100 to more than 75,000 assessments a year, and prices for apparently equivalent assessments varied about fourfold between integrated care boards.Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, pages 306 and 328 to 329 (provider estimates and evidence gathered by the review) and page 165 (survey of responding independent providers, Independent Healthcare Providers Network). The shares are thresholds, not exact values. The review says the diagnosis figures "require careful interpretation, as comparable NHS data on case mix and outcomes are not routinely available". Contains public sector information licensed under the Open Government Licence v3.0.

13. Autism and ADHD at 17 were the strongest single health predictors of being out of education or work at 23#

In the Millennium Cohort Study, young people with recorded autism at 17 were 3.6 times as likely to be NEET (not in education, employment or training) at 23, and those with ADHD 3.2 times. That compares with 2.4 times for a long-term mental health condition, and less for every physical health factor and health behaviour measured.

Post-diagnostic support should reach into the move to college, training and work, not stop at the diagnostic report. Our article on the post-diagnostic offer the review expects sets out what that support can include.

How much more likely to be NEET at 23, by factor at 17Against young people without that factor. Overall, about 13% of more than 8,300 were NEET at 23.Main factorsAll 31 factorsFor comparisonAlready NEET at 17~4.2×Already NEET at 17about 4.2 times as likely to be NEET at 2395% interval about 3.2 to 5.5 · Stated as “more than four times”; point read from chartHealth factors at 17Autism3.6×Autism3.6 times as likely to be NEET at 2395% interval about 2.7 to 4.8 · Stated in the textADHD3.2×ADHD3.2 times as likely to be NEET at 2395% interval about 2.4 to 4.5 · Stated in the textLong-term mental health condition2.4×Long-term mental health condition2.4 times as likely to be NEET at 2395% interval about 1.85 to 3.1 · Stated in the textPsychological distress~1.8×Psychological distressabout 1.8 times as likely to be NEET at 2395% interval about 1.4 to 2.3 · Read from the chart, approximateHyperactivity symptoms~1.5×Hyperactivity symptomsabout 1.5 times as likely to be NEET at 2395% interval about 1.2 to 1.9 · Read from the chart, approximateObesity~1.5×Obesityabout 1.5 times as likely to be NEET at 2395% interval about 1.2 to 2 · Read from the chart, approximateLong-term physical health condition~1.3×Long-term physical health conditionabout 1.3 times as likely to be NEET at 2395% interval about 0.95 to 1.7 · Read from the chart, approximateSmoking~1.2×Smokingabout 1.2 times as likely to be NEET at 2395% interval about 0.95 to 1.6 · Read from the chart, approximateCannabis in past year~1.2×Cannabis in past yearabout 1.2 times as likely to be NEET at 2395% interval about 0.93 to 1.5 · Read from the chart, approximateHigh social media use~1.0×High social media useabout 1.0 times as likely to be NEET at 2395% interval about 0.77 to 1.2 · Read from the chart, approximatetimes as likely1×2×4×Lines show 95% confidence intervals, read from the chart. A dot left of the 1× line means less likely than peers. ~ marks a value read from the chart. Three or more mental health difficulties together carried a risk similar to autism (about 3.6×, read from chart).Source: Independent review into mental health conditions, ADHD and autism, DHSC, 2026, page 358 (text) and page 360, Figure 16.2 (Millennium Cohort Study). Values for autism, ADHD and long-term mental health condition are stated in the text; all other values and every confidence interval are read from the chart, approximate. The review notes the cohort data "cannot reliably distinguish" young people who could work with better support from those who will need lifelong provision of other kinds. Contains public sector information licensed under the Open Government Licence v3.0.

Where the numbers come from#

Every value comes from the review's final report, with the page it is on in the note under each chart. Values the report prints are given as stated. Values we read from the report's charts are marked approximate, and where we did a sum ourselves, such as a gap between two figures, the note says so.

Many of these numbers come from screening questionnaires in national surveys, which show an increased likelihood of a condition, not a diagnosis. Several come from one city or a small group of services, which the notes also say. The rest of our Fonagy review coverage explains the recommendations, the move to CQC regulation and the new data rules for Right to Choose providers.