life-course evidence and gap map
where the evidence
runs out.
Male mental health across the whole UK life course — what happens, when the evidence says it matters, what survives testing, and where it runs out. Built to be argued with by psychologists, not admired by funders. Everything below is the evidence for and against isjamesok?, including the parts that argue against it.
This is a formal, expert-facing document — a different register from the rest of the site, on purpose.
twelve corrections worth arriving with.
Each overturns something the initiative might otherwise have said, and several overturn things the wider sector says routinely. Getting these right before a room full of academics does is worth more than any statistic in a deck. Four are flagged OUR OWN — corrections to what this initiative itself would otherwise have assumed or claimed, not just to the wider field.
01
Male suicide is a rising epidemic.
The male age-standardised rate in England and Wales has fallen from 19.2 per 100,000 in 1981 to 17.6 in 2024 — about 8%. The honest framing is stalled progress and a widening sex ratio in a slowly falling total, not a worsening crisis. The defensible urgency claim is absolute burden — a three-year mean of 3,983 men a year in England — and the near-absence of a tested evidence base, not a trend. Registration-year counts (which rose 2021–23) cannot support a trend statement either way: only 38.7% of 2024 registrations occurred that year.
02
Means restriction works, so fund it.
Means restriction is the best-evidenced category in suicide prevention — earned on firearms, pesticides, bridges and analgesic pack sizes. Hanging, strangulation and suffocation is 61.7% of male suicides in England and Wales, and in the community there is nothing purchasable, licensable or barrier-able to restrict. The field’s single most-evidenced category has its weakest applicability to the way British men actually die — a genuine, fundable research question, not a reason to fund it uncritically.
03
A new organisation should build a way to spot who’s most at risk.
NICE guideline NG225 is unambiguous: “do not use risk assessment tools and scales to predict future suicide.” Applied to men presenting after self-harm, the best real-world positive predictive value is about 6.2% — sixteen men flagged per true case. The instruments do not beat unstructured clinician judgement, and around 80% of people who died in service contact had been rated “low” or “no” risk. The unit of intervention is the population, the moment and the environment — never the individual prediction.
04
78% of men say they’re fine when they’re not — so ask twice.
The 78% figure is a 2018 Censuswide campaign poll, measuring hypothetical self-reported behaviour — not a study, no clinical outcome. “Ask twice” has never been tested at all. The properly evidenced replacement: McCabe et al. found disclosure of suicidal ideation ran at 9.7% after a negatively framed question against 38.1% after a positively framed one — a 3.93× difference; the homepage rounds this to “4×” for the headline figure, and the exact ratio is stated here. The active ingredient is not repetition. It is how the question is put the first time — which is why this site says ask openly, not ask twice.
05
Men don’t seek help.
No longer true once they reach a clinician. The Adult Psychiatric Morbidity Survey 2023/24 found 43.3% of men and 46.7% of women with a common mental health condition received treatment — no difference between the sexes, a change since 2014. Yet men are about a third of Talking Therapies referrals and roughly three-quarters of suicides. The failure is detection and referral, not treatment refusal.
06
Men who die by suicide were invisible to services.
The National Confidential Inquiry examined every suicide by a man aged 40–54 in England, Scotland and Wales in 2017: 91% had contact with at least one frontline service; 82% with primary care or an emergency department. These men were not invisible. They were seen and not detected. (A case series with no control group — read every percentage as what services would have seen, never as a cause.)
07
The album is an intervention against suicide.
The only randomised trial of a music video designed to promote male help-seeking — Boys Do Cry, 476 Australian men — was null on its primary outcome. Meanwhile the harm mechanism is firmly established: media reporting that discloses method carries RR 1.30 for same-method deaths. No study supports the claim that a music release reduces suicide or increases help-seeking. The album is reach and a fundraising mechanism — presented as reach, with an evidenced intervention attached to it, and a Papageno-compliant creative protocol.
08
Occupational risk is about maleness.
Male standardised mortality ratios: low-skilled construction workers 369, roofers and tilers 266 — but male doctors 63, and managers and directors the lowest risk of any group. High-status, high-stress male occupations can be protective. This is the cleanest available inoculation against treating all male stress as mortality-equivalent, and it means the album’s middle-class-pressure story is a story about distress, not about mortality risk.
09
Farming is the highest-risk occupation.
Farmers record SMR 101 — indistinguishable from the national average. What is elevated is the hired end: farm workers 221, gardeners 201 — a class gradient inside agriculture, not an agriculture effect, and neither is the highest in the table. Farming has an unusually dense charity landscape aimed largely at farmers, while the mortality sits with the workers those charities reach least.
10
Male suicide is a rural problem.
A systematic review of 35 UK and Ireland studies found 54% reported higher rates with greater urbanicity against 21% for rurality. On 2024 male data, Wales is the outlier — not Scotland — at 25.0 per 100,000 against England’s 17.1. The defensible rural claim is about occupational risk, access to means and service access, not residence.
11
There is a veteran suicide epidemic.
Tri-Service male suicide runs at a 60% decreased risk against the general population. UK veterans overall show an SMR of 94 (88–99) — significantly below the general population. Risk concentrates hard at the young end instead: SMR 305 at ages 16–19 among early service leavers. Repeating a “veteran suicide epidemic” framing is wrong and easily corrected in public.
12
The leverage point is the teenage boy.
No UK cohort measures masculinity-norm endorsement and follows it to adult outcomes. The only major male-specific cohort in the world — Australia’s Ten to Men — finds conformity associated with depression “especially in the oldest age group”, stronger at 51–55 than at 18–25. If that replicates, the leverage point is not the teenage boy. It is the fifty-five-year-old who will not walk into a room — and most of what the sector does, including the instinct this initiative started with, is aimed at twelve.
what this is, and what we want from you.
This evidence map and the method were written by people who are not clinicians and not researchers. The reasoning is ours and so are the mistakes, and it should be read that way. Clinical and academic advisers are coming into the work now and are shaping what happens next; what does not yet exist is formal clinical governance, a named review process with the authority to stop something — building that is a first-year task, not a box already ticked. This started as a concept album about the pressure of an ordinary British male life, and we are building an organisation around it called isjamesok? — aimed at male suicide before the crisis system takes over. Everything behind this page is the evidence we could find for and against that idea, including the parts that argue against it. We would rather you took it apart than approved of it.
What it would actually do in year one
Incorporate, and hand away the power to stop it — an independent chair, an independently chaired steering committee, and a separate data monitoring committee that can halt recruitment. Release the record and run the campaign — reach and funding, not an intervention, and not claimed as one. Start the actual work: trained conversation, in settings men are already in, with someone trained to ask well and a named local route to send him to, agreed before a single conversation happens. Commission two of the eleven missing studies. Run nothing in a school until the safeguarding, consent and harm apparatus is real rather than described.
What we want from you
Tell us where the reading of the evidence is wrong — there are twelve places where the received wisdom does not survive the best studies. Tell us which of the three questions below actually decides the design. Tell us what would have to be true for you to say don’t do this. And if you think it should exist, we need people who will keep arguing with the design, not people who will endorse it.
The three questions we want argued
1. Male depressive symptoms peak around twenty. Male suicide peaks around fifty. What accumulates across those thirty years? 2. 91% of the men who died had contact with a frontline service. Is the problem men who will not speak, or services that did not ask? 3. The best-evidenced interventions in this field are the least male-specific. The male-specific ones have the weakest evidence. How should a new organisation resolve that?
And one decision we would like help with
Do we run our own sites, or evaluate someone else’s? Our own two or three sites cannot detect anything on their own — the power calculation needs 526 per arm — so it is a feasibility study whatever we call it. Evaluating an existing network (Andy’s Man Club reaches 6,500 men a week, UK Men’s Sheds 33,000, neither with a controlled outcome study) would be the largest single contribution available to this field, but it needs a partner willing to be measured by a body with no track record.
Our instinct is that the second is the better use of a first year and the harder thing to secure. We would rather be told we have that backwards now than in three years.
Declaration of interest
We are not a neutral party, and you should not read this as though we were. The record is ours, the idea is ours, and every incentive we have runs towards this existing. We have tried to write against that: the document grades this plan’s own primary activity as untested, grades the music as not supported as an intervention, and marks the claims about the network’s reach as unverified — asserted by us, sourced nowhere. Read it knowing that, and check the places where we have been too kind to ourselves.
How we check ourselves
We’re targeting a behaviour, not a mood.
“Look after your mental health” is not an instruction anyone can follow. “Ask it openly, then wait” is. Everything we make has to survive that test: could a nervous seventeen-year-old change the shape of one sentence on Thursday?
We check our own claims first.
The line we started with — ask twice — turned out to have no study behind it, so we changed it. The figure everyone quotes about men saying they’re fine is a campaign poll, so we stopped using it. This document grades the album as not supported as an intervention. If you find something else that doesn’t hold, tell us and we’ll publish the correction.
We follow the same rules as journalists.
Samaritans media guidelines, everywhere, including social and lyrics. Never method or means. Never glamour. Always a signpost.
We signpost to people who are already brilliant at crisis.
CALM and Samaritans do the hard end. We are not building a helpline and we won’t pretend to. Our job is everything that happens in the eighteen months before anyone needs one.
We’re being marked by people who know more than us.
Being shaped with CALM and with specialists in psychology, behaviour change and education. Nothing on this site is signed off as finished.
And here’s what we’re not claiming.
No impact figures, because we haven’t earned any yet. No charity number, because we don’t have one yet. No proof this works at scale — a design we can defend, and a plan to test it. If you want the version with the footnotes, this document is it.
the shape of a male life.
Four instruments, and they disagree with each other. Counts say where a national programme should be aimed; rates say who is most at risk as an individual; deprivation tests whether this is really about class; years of life lost settles the argument between the first two. The sector routinely points at the wrong decades because it reaches for the wrong one.
Suicide is the leading cause of death for men under 50. The male rate has actually fallen since 1981 — what’s stalled is the progress, and what’s widened is the gap between men and women. Source: ONS · England & Wales.
Only one in three people in talking therapies is a man — but once men get there, their outcomes match women’s, and treatment rates are now equal. The failure is being spotted and referred, not being willing. Source: NHS Talking Therapies.
Men aged 50–54 and men aged 90+ have almost identical suicide rates — 22.9 against 23.9 per 100,000 — but nothing like the same number of deaths: 434 a year against 41. Nearly six in ten male suicides in England (57%) occur between 35 and 64; the band 45–64 alone accounts for 1,492 deaths a year, 37.5% of the total.
→ scroll
| England, males | Rate 2021 | Rate 2022 | Rate 2023 | Rate, 3-yr mean | Deaths, 3-yr mean | Share of male deaths |
|---|---|---|---|---|---|---|
| 15–19 | 8.2 | 6.9 | 7.9 | 7.7 | 130 | 3.3% |
| 20–24 | 15.4 | 13.9 | 13.7 | 14.3 | 249 | 6.3% |
| 25–29 | 17.7 | 17.6 | 16.5 | 17.3 | 317 | 8.0% |
| 30–34 | 19.1 | 19.9 | 18.9 | 19.3 | 372 | 9.4% |
| 35–39 | 19.7 | 20.3 | 23.1 | 21.0 | 394 | 9.9% |
| 40–44 | 20.5 | 21.4 | 21.8 | 21.2 | 382 | 9.6% |
| 45–49 | 21.4 | 22.3 | 25.3 | 23.0 | 395 | 9.9% |
| 50–54 | 22.4 | 23.1 | 23.1 | 22.9 | 434 | 10.9% |
| 55–59 | 19.5 | 17.9 | 21.5 | 19.6 | 371 | 9.3% |
| 60–64 | 16.4 | 17.5 | 18.8 | 17.6 | 292 | 7.3% |
| 65–69 | 14.0 | 14.2 | 15.1 | 14.4 | 199 | 5.0% |
| 70–74 | 10.8 | 11.5 | 12.5 | 11.6 | 148 | 3.7% |
| 75–79 | 11.2 | 9.1 | 11.0 | 10.4 | 109 | 2.7% |
| 80–84 | 10.9 | 10.7 | 15.8 | 12.5 | 68 | 1.7% |
| 85–89 | 17.1 | 17.1 | 15.9 | 16.7 | 61 | 1.5% |
| 90+ | 18.8 | 30.4 | 22.5 | 23.9 | 41 | 1.0% |
Source: Samaritans, Suicides in England 2023, Tables 1–2, from ONS death registrations. Unweighted three-year means, not properly pooled rates — adequate for shape, not a headline statistic. Registration delay is severe: only 38.7% of deaths registered in 2024 occurred that year.
Two further instruments — deprivation, and years of life lost
Deprivation is the direct measure of the class argument — occupation, used in corrections 8 and 9, is the weakest instrument available for it, silent on the unemployed by construction and carrying reverse causation. The only male-specific IMD series traced for England ends at 2018 occurrences: most deprived 20.3 per 100,000 against least deprived 11.8 — a ratio of 1.72×. Everything routinely updated since is all-persons, not male.
→ scroll
| Measure and population | Most deprived | Least deprived | Ratio | Period | Grade |
|---|---|---|---|---|---|
| England, males, age-standardised, IMD quintile | 20.3 | 11.8 | 1.72× | 2018 | Moderate |
| England, males aged 43, most vs least deprived | 36.6 | 13.5 | 2.7× | 2010–2019 | Moderate |
| England, all persons 25–44, IMD decile | 14.9 | 10.6 | 1.41× | 2020–2024 | Strong, not sexed |
| Scotland, all persons, SIMD quintile | 18.7 | 7.4 | 2.5× | 2024 | Strong, not sexed |
| Wales / Northern Ireland, males | — | — | — | — | No direct study |
England has no current male suicide rate by deprivation: the only male-specific series is eight years old at the data. Ask for a current figure rather than assert a 2018 one as though it were live.
Years of life lost settles the rates-versus-counts argument — the only measure that weights a death by how much life it removes. Roughly 139,000 years of male life are lost to suicide in England every year, about 35 years per death. The burden is concentrated across 25–39, not at the 45–54 plateau this document’s charts make visually central: ages 15–39 hold 36.8% of deaths and 53.3% of the years lost, while 65+ holds 16.1% of deaths and 5.9% of years. If the question is where the most life is lost, the answer is men in their late twenties and thirties.
Method: YLL = deaths × remaining period life expectancy at band midpoint, using this document’s own three-year death means and ONS national life tables (UK, 2022–2024). No discounting, no age-weighting, no allowance for competing risks — every figure is an upper estimate of life foregone, not a forecast.
life-course sensitivity windows.
Every entry is a claim about when an exposure matters. Colour is the grade of evidence for the window, not for the exposure — a real risk factor can have a poorly-evidenced window. Two widely asserted windows — the Year 6–7 transition dip and the GCSE-year spike — are graded refuted, because the best-identified studies looked for them and did not find them.
Gestation, infancy and childhood
Adolescence and the transition to adulthood
Established adulthood, midlife and later life
the phases.
What happens, what the evidence says about it, and what the study it rests on cannot tell you. Each phase is a disclosure — open the ones you want to argue with.
Conception – 3Gestation and infancyWhere the strongest-sounding claims are, and where causal designs have done the most damage to them.
3 – 11ChildhoodBoys carry more measured morbidity than girls, are excluded at twice the rate, and almost never reach a clinician.
11 – 18AdolescencePrevalence reaches parity with girls, self-harm doubles, and two of the transitions UK policy is built around turn out not to exist.
18 – 30Early adulthoodSupport structures withdraw at once, psychosis incidence peaks, and the men who are not at university largely stop being studied.
30 – 50Established adulthood and midlifeThe ascent to the peak. Childhood sets susceptibility; concurrent adversity detonates it.
50 – 70The peak and the exit from workThe highest male suicide rates of the life course, and the phase where masculinity norms may matter most — not least.
70 – endOlder ageThe lowest suicide rate of adult male life, and the least visible group in every system designed to find them.
acute windows.
Life-course exposures set susceptibility. Acute events detonate it. These windows — days and weeks, not decades — are the most tractable part of the whole map, and several have no service designed for them at all. Tractable is not the same as large.
→ scroll
| Acute window | Peak duration | Best risk estimate | Provision |
|---|---|---|---|
| Arrest and charge Strong | Days to first hearing | 60 deaths/yr after police custody, 93% male | None. Policing has operationalised the risk; health has not. |
| Psychiatric discharge Strong | Days 3–4 peak | ~133 male deaths/yr within 3 months | 72-hour follow-up standard, 80% target — lands before the peak. |
| Prison reception Strong | First 48 hours; first month | 11% of deaths ≤48h; 41% ≤1 month | ACCT — never evaluated for outcomes. |
| Prison release Strong | Immediate, then ~6 months | Suicide SMR 6.9; drug poisoning SMR 61.0 | Through the Gate judged “negligible” impact. |
| A&E after self-harm Strong | First year | 0.82% of males die by suicide in year one | NICE says everyone gets assessed; median 58% do. |
| Relationship separation Moderate | Active separation phase | Separated men 2.82× odds; >8× for men ≤34 | None. 42,172 new private law cases a year. |
| Widowerhood Moderate | First month | Suicide aRR 13.6 in month one | Cruse’s “Men in Grief” publishes no figure for male client share. |
| Serious illness diagnosis Strong | First 6–12 months | COPD 13.7 vs 5.6 per 100k at 6 months | None. No NICE requirement to ask about risk at diagnosis. |
| The final GP consultation Strong | Days to 3 months before death | 43% consulted within 3 months, 12% within a week | PROMDEP found routine PHQ-9 monitoring null. |
What the ranking does to the argument
The four computable windows together carry roughly 268 to 284 excess male deaths a year, against about 3,983 male suicides in England — about seven per cent of the total. Psychiatric discharge is roughly twice prison release and four times prison reception. And the single largest reachable population is not a risk window at all: at least 520 men a year sitting in a GP consultation within three months of their death.
Source: NCISH Annual Report 2025; IOPC; PPO Learning Lessons Bulletin 22; Slade et al. 2025; Hawton et al. 2015; Mughal et al. 2023; Wilson et al. 2023.
what actually works.
Stated first: the interventions with the strongest evidence are the least male-specific and the least emotionally resonant. The interventions that are genuinely male-specific and culturally powerful have the weakest outcome evidence in the entire field.
Why awareness plateaued, and why this initiative isn’t another campaign
Awareness worked. Ask a 19-year-old whether mental health is real and he’ll look at you like you’ve asked whether Wednesday is real. That fight is largely won, and the people who won it deserve more credit than they get.
able
know what to actually say
an opening
a moment where it isn’t weird
want to
and we’ve had ten years of this one
COM-B — capability, opportunity, motivation. People only change what they do when all three line up. A decade of campaigns bought motivation and left out the other two.
So we stand in the doorway, say “you alright?”, get told “yeah, fine”, feel relieved, and leave. That’s the gap. Not caring. Words, and the moment.
This isn’t a new idea. It’s an unfinished one — and we’ve changed it.
Time to Change asked the country to ask twice. Then the funding stopped and most of us went back to accepting “fine” first time. When we went looking for the evidence behind asking twice, there wasn’t any — no trial, no test, nothing. What there is evidence for is how you ask. So we’ve kept their instinct and changed the instruction.
What survives testing
What does not survive testing
The album, judged honestly
If a new charity had £1m a year, on the evidence
First: the first controlled evaluation of a male peer model at scale. Second: a trained-conversation programme on the MATES model, in male-dense settings, evaluated from day one. Third: means restriction advocacy and local site work, with the honest caveat that it addresses a minority of male deaths. Fourth: brief contact and safety planning at transition points, delivered with NHS trusts, never as a parallel service. And throughout: the album as measured reach, with the funnel from reach to referral pre-registered and the nulls reported.
the sector, and the white space.
A new organisation is defensible only if it commits to a gap the sector has structurally ignored, and only if it builds evaluation in from day one. The domain gap that justifies that is narrow: no What Works centre owns male mental health as an outcome domain. The 25–55 male peer-support space is genuinely crowded.
→ scroll
| Organisation | Scale | Reach and ages | Evidence |
|---|---|---|---|
| Samaritans | £27.5m | 3.3m contacts a year, ~200 branches. Not male-specific. | Genuine research function — rail suicide prevention is its strongest line. |
| Andy’s Man Club | £1.09m | 350+ weekly groups, 6,500+ men a week | No peer-reviewed outcome evaluation. |
| James’ Place | £3.55m | 1,352 men in suicidal crisis, four cities | The best-evidenced UK organisation — five peer-reviewed papers, six-year LJMU evaluation. |
| UK Men’s Sheds | Unverified | 1,150+ sheds, 33,000 weekly participants | 52 studies, zero RCTs. Ireland’s Sheds for Life is the only controlled evaluation, mean age 69. |
| Movember Europe | £27.4m | The UK’s largest men’s health funder | Major global research funder; Men in Mind has an RCT. |
The National Lottery Community Fund has given £80m+ across nearly 1,500 men’s health projects over thirty years — about £2.7m a year across the whole UK. England’s Men’s Health Strategy (November 2025) commits under £7m over three years to men’s mental health, with no new cohort, no observatory and no numerical target. This is a strategy without a budget.
The five claimed white spaces, retested
Boys aged 7–11
Occupied: the Youth Endowment Fund funded Future Men £951,890 for a 486-boy RCT; Chance UK runs a randomised trial at 5–11. What’s unclaimed is narrow — a mental-health outcome set nobody is funded to measure.
Evidence and evaluation infrastructure
The What Works Network, YEF and EEF already exist and publish negative results. The residual is real but narrower: none of these centres owns male mental health as an outcome domain.
Men 55+, retirement transition and late-life crisis
Men’s Sheds reaches ~33,000 weekly, but the only controlled evaluation had a mean age of 69, 80% retired. Still unoccupied: involuntary work exit, widowerhood and late-life crisis as triggers.
A gender-competency programme for frontline practitioners
Movember’s Men in Mind is piloting in England (York and North Yorkshire, 30 professionals). The incumbent moving in is the most likely funder of anything proposed here.
Men at institutional transition points
Liaison and Diversion reached 100% coverage in March 2020; RECONNECT covers care after custody. What survives: no statutory referral on release without charge, and neither service publishes sex-stratified outcomes.
The second question — and nobody in this sector passes it
Not one organisation in UK men’s mental health has peer-reviewed evidence that it reduces suicide, self-harm or clinical distress in the men it serves. So the honest competitive read is not “the space is taken”. It is that the space is full and the problem is unsolved.
the case for a new organisation.
Stated as it would have to be defended, not as it would be pitched. Not that men need more awareness, more talking, or another place to go. The population layer — everything before the crisis system — is where the failure sits, the men who die are already being seen there and simply not asked well, and the one large effect anyone has measured on that has never been built into a programme.
Part one — why it needs to exist, and the counter-case argued first
1 — Men are three-quarters of the deaths and a tenth of the plans. Only one in ten English local authority action plans has a specific focus on men, against males accounting for around three-quarters of suicides.
2 — No one has calculated what actually causes male suicide in Britain. There is no published UK population attributable fraction for any male suicide risk factor. The ONS linked census-mortality cohort could produce one for the cost of an accredited researcher and a year.
3 — The best-evidenced things in the field do not fit British men, and the things that fit have no evidence. Means restriction is Strong on firearms and pesticides; the therapy trial base is 62% female; Andy’s Man Club has no peer-reviewed outcome evaluation at all.
4 — The men who die are already inside the system, and no one owns the moment. 91% had contact with a frontline service; 60 apparent suicides a year follow police custody with no statutory referral on release.
And the counter-case, argued first: the 25–55 male peer-support space is genuinely crowded; fragmentation is itself a harm; overheads are real; and a research-and-accountability body may not need to be a charity at all. The answer that survives: the function is adversarial — naming an unevaluated national programme or an ignoring local plan is work no delivery organisation can do without damaging its own relationships.
The scope, in one sentence
This organisation does not deliver crisis services. Samaritans, CALM, James’ Place and NHS crisis care do that, and do it better than a new charity would. The population layer is where it works, the handover is where it advocates and commissions, and neither involves standing up a crisis service.
Inside that boundary the organisation does three things, and it grades its own activities the same way it grades everything else: the record and campaign (reach, not an intervention, graded not supported as one), trained conversation in settings men are already in (the primary work, and untested), and education, upstream (a stated bet, not a claim).
Part two continued — the reach, the objections, and the research it commissions
Why this is not another men’s group. The active ingredient is not the talking. It is the asking, and the route out the other side. Disclosure ran at 9.7% after a negative framing against 38.1% after a positive one — a trainable verbal behaviour, on top of a settings insight (men already there, not venues) and a route (somewhere to send him).
Two things this claim has to survive. First: this is gatekeeper training, graded Contested here, with the instruction never to fund it as a flagship — so the honest label is the most promising untested thing available, not the evidenced work. Second: the setting is occupied — Mates in Mind, the Lighthouse Club and Movember’s SpeakEasy are already there. What survives the test: the evaluation itself, the referral route as an object of study, and male-dense settings the incumbents do not reach.
The reach, and what it is worth. A finished album, routes to recording artists, senior music-industry release experience and a school network are being assembled — stated honestly as warm routes through people who know the individuals, not signed commitments. Reach is not treatment. It may be the thing that makes treatment knowable, by putting men in front of an evaluation that would otherwise never be powered.
Eleven commissioned studies, deliberately unranked — from UK population attributable fractions for male suicide (existing data, ~12–18 months, one analyst) to the first controlled evaluation of a UK male peer model (needs a partner, 2–3 years) to why men die within the same method even once method choice is accounted for. Ordering them is an argument about what the organisation is for, which is why they’re listed by age instead.
Part three — the schools bet, and the trap it has to avoid
England has already run this experiment, at scale, and it failed. MYRIAD, AWARE and INSPIRE — over 40,000 pupils — returned null primary effects with harm signals concentrated in already-vulnerable pupils. A charity arriving to teach boys about mental health in schools would be proposing the thing that has been disproven.
What the evidence does support: structured, relational, activity-anchored, repeated, delivered by men, targeted at boys already in difficulty rather than at everyone. One-to-one mentoring across 70 studies gives d = 0.21, larger where more mentors are male. Asking about suicide is safe — it is universal curricula that showed harm, not asking.
The honest terms of the bet. No schools programme can demonstrate that it reduced suicide, and it never will — the outcome sits thirty to forty years downstream. The chain being bet on has two evidenced links and two unevidenced ones. What gets measured instead: disclosure behaviour, suspension and exclusion rates, and outcomes reported separately for boys already in difficulty, since that is where the last programme did harm.
What it adds up to, and what would have to be true for it to be wrong
One engine, two arms, and an honest label on each. The record and campaign reach grown men now and pay for everything else. The trained-conversation work is the evidenced work rather than the awareness around it. The schools work is a stated bet on boys who would otherwise become the men in the first sentence.
What would have to be true for this to be the wrong idea: that the divergence study finds no male-specific mechanism worth acting on in the 7–11 window; that the programme cannot reach the boys it is designed for; or that a targeted boys’ programme reproduces the harm signal the largest English trials found. Each is a stopping condition, not a risk to be managed — and two of the three currently have no owner.
the apparatus - safeguarding, consent, ethics and harm.
Everything above describes what to do. This describes what makes it lawful, safe and fundable — and where, right now, it is not yet either.
1–7 · The children’s programme — safeguarding, recruitment, consent, ethics, harm monitoring, trial design, equality
Safeguarding. One governing sentence: the programme never holds risk it cannot route. The school’s designated safeguarding lead owns every decision about a pupil; a visiting charity acquires no parallel duty. Detection outruns treatment: 29% of referrals close before any treatment, median wait 224 days — the case for the programme and the strongest argument against it.
Safer recruitment. Two design rules — “men in the room” and “recruit through existing networks” — read literally as a grooming risk unqualified. Rewrite: sourced through networks, selected through safer recruitment, with enhanced DBS and barred-list checks for every facilitator. From 1 September 2026 the supervision exemption for volunteers is removed entirely.
Consent. Boys are selected using SEN status and free-school-meal eligibility — special category data under Article 9 UK GDPR. Opt-out fails three times over. Explicit opt-in parental consent, in writing, stating the selection criteria plainly, plus a mandatory DPIA under Article 35.
Ethics. A founder-led charity should not sponsor the trial of its own programme. An independent steering committee, independently chaired, with authority to halt recruitment; protocol pre-registered; nulls and harms published.
Harm monitoring. The current outcome set contains no symptom scale, no wellbeing scale, no self-harm measure — run on it, MYRIAD’s own harm signal would have reported a null. Two mechanisms the design creates and doesn’t name: labelling, and deviancy training from grouping at-risk boys together.
Trial design. Allocate at the school, not the boy. The comparator is not a waitlist — this map already shows what waitlist comparisons do. An illustrative power calculation needs roughly 108–170 schools, and a trial powered for benefit is underpowered for harm in the exact subgroup it exists to protect.
Equality. The selection rule (boy + SEN + FSM + Year 5–6 + suspension history) is patterned by ethnicity, most sharply on permanent exclusion. A public sector equality duty analysis is required before the first school is approached, not retrofitted. Boys-only provision in a mixed school needs a Section 158 positive-action case that, as of this map, has not been made.
Why this trial, as specified, would not yet pass a panel
The eligible population is far smaller than the calculation assumes — perhaps three eligible boys per school, not twelve. The detectable effect (0.20 SD) is the most optimistic number available. And the trial is underpowered for the harm it was built to detect. The fundable object is a feasibility and pilot study, not this trial.
8–16 · The adult arm — and why none of the children’s apparatus transfers
The primary activity is now a trained peer-conversation intervention for adult men in male-dense settings. Every instrument in sections 1–7 governs children. Searched here, Care Act, Mental Capacity, adult at risk, lone working, insurance and vicarious return zero occurrences.
Adult safeguarding. Section 42 of the Care Act requires all three limbs conjunctively — a suicidal, drinking, indebted groundworker fails two of them: no statutory enquiry, no designated safeguarding lead, no agency under a duty to receive the concern.
The disclosure problem, at volume. MATES’s case-management demand up 265% is the size of the load the route must absorb, not evidence the route absorbed it. No defensible UK forecast base exists — the first eight sites are the forecast.
Who the asker is. The connector is almost certainly not in regulated activity. DBS states befriending “isn’t one of the activities listed” — the check available is Basic, and the adults’ barred list plays no part. Connectors come from the population this map grades most lethal — the design trains the highest-risk group to absorb the disclosures of the highest-risk group.
Lone working. “No unobserved one-to-one work of any kind” — the children’s rule — cannot be carried across, because the adult intervention is unobserved one-to-one contact by design. The replacement is weaker: pre-agreed contact intervals, a tested lost-contact procedure, every contact logged.
Occupational consequence of disclosure. A man who discloses suicidal ideation in a safety-critical trade risks a site pass withdrawn, an occupational health referral, or his job. This harm appears nowhere in the standard literature. A data firewall from the employer is a signed term of the site agreement.
Andy’s Man Club may work partly because it keeps no records — first names, no forms, no follow-up. Imposing consent, identifiers and follow-up changes the intervention being measured.
Postvention. At this map’s background rate and construction SMRs of 234–369, a programme reaching several thousand men over three years should expect deaths among them. There is a media protocol for the album and none for a death on site.
Equality. A men-only offer on a mixed site excludes the women on it. The duty-holder changes from a school to a private contractor, and the argument — though the evidence for it exists in this map — has never been made.
gap register.
These are the holes, ranked by how much they cost the argument. Many could be closed by re-analysing data that already exists — cheap, specific asks for an academic partner rather than vague calls for more research.
All twenty-four gaps
No UK attributable fraction exists for any male suicide risk factor
The field has hundreds of relative risks and not one published population attributable fraction for male suicide in the UK.
847 coroner reports on preventable suicide deaths, never analysed for men
Prevention of Future Deaths reports are free and public on judiciary.uk. No male-focused analysis of the corpus exists.
Right Care Right Person was implemented nationally and never evaluated
The largest change to UK crisis response in a decade, with no outcome evaluation, no mortality analysis, and no sex-stratified assessment.
Hanging is 61.7% of male deaths and has no community prevention study
Community hanging prevention has never been studied anywhere — not in the UK, not in any comparable country.
The residual of the gender paradox is unexplained
Within identical methods men still die at OR 1.64 (p = 0.000005). Intent, physiology, interruption and care received have never been separated.
More screening is the obvious answer and the trial says it does not work
PROMDEP, the UK trial of routine PHQ-9 monitoring, was null. Adding a questionnaire is not the fix.
Men are 75% of suicides and appear in about 1 in 10 local plans
Across 145 English local authorities, only one in ten published action plans had a specific focus on men.
Masculinity norms are never measured in a UK cohort
ALSPAC, the Millennium Cohort and Next Steps carry no masculinity instrument at all.
A national commitment to assess fathers exists and has never been audited
NHS England committed that partners would be offered a mental health assessment by 2023/24. No named instrument, no data return, no coverage figure.
Non-resident fatherhood is asserted everywhere and studied nowhere
No UK cohort, register or linkage study estimates suicide or mental-illness risk in non-resident fathers. ONS stopped collecting age at divorce in 2020.
ACCT has never been evaluated against outcomes
The national care-planning process for every at-risk prisoner has been studied only qualitatively, with 31 staff and 17 prisoners in 2015.
The male prevalence drop at 17 is unexplained
Measured probable-disorder prevalence in males falls from 22.3% to 15.4% while female prevalence rises to 31.6%.
Older men’s access to therapy is unmeasured by sex
We know 65+ receive 5.6% of referrals against 12% expected. We do not know the male share.
The men at highest risk are absent from the survey that sets the baseline
APMS excludes prisoners, rough sleepers, hostel residents and inpatients — every population with the most extreme male mortality on this map.
IICSA never analysed the male stratum of its own data
The Truth Project heard from 5,862 people, 29% male, and published no sex-disaggregated breakdown of age at abuse or mental health impact.
People with learning disabilities are absent from the suicide strategy
Men with a learning disability die 16.5 years earlier than the general population. Learning disability is not among the strategy’s seven priority groups.
The waiting lists manufacture the risk marker
ADHD mortality rises with age at diagnosis — MRR 4.25 if diagnosed at 18 or later. England has 294,792 open autism referrals, 86.8% waiting 13 weeks or more.
No instrument has ever been validated against male suicide
A systematic review found no study had assessed suicide victims or attempters for male-typical depression symptoms.
Nobody knows what is spent on male-specific mental health research
UK mental health research runs at about £124m a year, categorised by condition, funder and age — but not by sex.
Half a million men, and no mental health offer
An estimated 447,000 men use anabolic steroids; 77% obtain equipment from needle exchange services with no psychological offer attached.
Nobody has separated the age effect from the cohort effect
The male suicide peak oscillates between bands. Drug-misuse deaths peak as an ONS-stated Generation X cohort effect.
The largest male network in the country has no outcome evaluation
Andy’s Man Club and UK Men’s Sheds reach nearly forty thousand men a week between them with no randomised trial.
Two national series have gone dark
The ONS deaths of homeless people series has not been updated since November 2022, and ONS stopped collecting age at divorce in 2020.
questions for experts.
Written to be asked out loud. Each is a genuine open question, chosen because the answer would change how the organisation is designed, not just what it says.
The question the whole map turns on
For a developmental psychologist
For a psychiatrist or clinical epidemiologist
On founding an organisation
method, and what this map cannot do.
The value of a document like this is entirely in its accuracy. What follows is how it was built and where it is weakest, so anyone reading it knows which parts to lean on.
Twenty-three parallel deep-research passes — one per life phase, plus cross-cutting factors, acute transition points, the intervention evidence base, the sector landscape, and research infrastructure. Sources were prioritised: UK cohorts first, then UK administrative and population data, then UK quasi-experimental evaluation, and only then international evidence where UK data was absent. Findings were graded down, not up. Where a source could not be retrieved, the figure was left out and flagged rather than recalled from memory.
Seven structural limitations to state before anyone else does
1. Registration artefacts. The July 2018 change to the civil standard of proof breaks time-series comparability. Only 38.7% of deaths registered in 2024 occurred that year.
2. Household surveys miss the highest-risk men. APMS excludes prisoners, rough sleepers and inpatients — every group with the most extreme male mortality on this map.
3. Cohort attrition is socially patterned. The men who drop out are disproportionately the men the studies most need — and this map applies that caution to nulls it would like to keep, not only to ones it would like to discard.
4. Male-stratified analysis is the exception. Only 46.4% of neuroscience and psychiatry papers analyse results by sex. Men are not excluded from mental health research — they are included and then not analysed separately.
5. Instruments are not neutral. The SDQ is not gender-invariant at age 3; the male:female domestic abuse ratio moved purely as a function of ONS question wording.
6. Rates and counts answer different questions, and this map carries both. Men aged 50–54 and 90+ have near-identical rates and wildly different counts. Any figure quoted without saying which it is should be treated as unusable.
7. Suicide is statistically rare, so almost nothing measures it directly. A UK birth cohort of ~17,000 accumulates only around fifty suicide deaths by age 50. Every adequately powered cohort finding here is on a proxy.
Figures excluded, and figures still needing a primary check
Excluded for want of a traceable primary source: the claim that men take twenty-plus years to disclose sexual abuse; “3–5 men a week die by suicide because of domestic abuse”; the “one million UK steroid users” figure (the defensible estimate is 447,000); a fatherhood report’s “47 times” suicide-risk claim; an HGV driver suicide figure “20% above the national average”.
Still needing a primary check: the 55% distress rate in deselected academy footballers (paywalled); the “85% of academy entrants released” figure; the AWARE/INSPIRE effect sizes to two decimal places, obtained via secondary synthesis.
Where The Evidence Runs Out · life-course evidence and gap map for isjamesok? · September 2026. This document discusses suicide, self-harm and mental illness in research terms. It is a research instrument, not clinical guidance, and nothing in it should be used to assess risk in an individual. In the UK, Samaritans can be reached free on 116 123, and CALM on 0800 58 58 58.