Hide table of contents

With the AI windfall, mental health may be taken increasingly seriously by mainstream EA. There’s increasingly room to explore the cause area and answer some of the biggest uncertainties; at the same time, costs of the most effective interventions are continuing to fall. StrongMinds is now treating for ~$35, about a third of GiveWell’s best guess from a few years ago, and is expected to fall further.

This presents a unique opportunity to demonstrate that mental health interventions should be considered as part of broader global health portfolios. We can do so by making methodological improvements to our cost-effectiveness analyses.

The Happier Lives Institute have pioneered (and are constantly improving) methods for CEAs of mental health in WELLBY terms. However, their mission doesn’t cover other forms of outcomes, such as DALYs and income doublings. Furthermore, they haven’t yet publicly explored models of psychotherapy that leave lasting impacts on individuals (which may be of interest to HLI). Therefore, with an eye to increasing the estimated impact of mental health interventions in ways that are legible to mainstream EAs, I propose three directions for EA-flavoured mental health research.

About me: I’m Huw, I work at Kaya Guides, an EA mental health charity[1]. This is based on my own experience of building our CEA, and reviewing other people’s CEAs of us. As such, I only consider CEAs for psychotherapy—other interventions are much more neglected and they would also be promising, but I have no expertise in these areas. I don’t have formal psychotherapy training or research experience; I only have a degree in statistics. As such, I am likely missing important facts and many of my estimates may not be high quality.

Converting symptom severity to DALYs

Due to HLI’s excellent focus on the WELLBY, combined with CG/GiveWell’s neglect of mental health, there hasn’t been a lot of work in writing DALY conversions. Prior art is:

  • SoGive: Their CEA of StrongMinds has a unique methodology, which you can read more on in their post. I won’t discuss this in depth, but the methodology isn’t standardised or adapted from something else, so my fear is that it’s too speculative to be trusted more widely.
  • Founders Pledge: They developed a method which takes GBD weights and interpolates them between PHQ-9 states, for their StrongMinds CEA. However, Founders Pledge have since moved to using WELLBYs for mental health CEAs. This CEA is also rather rough.
  • UK Green Book: Converts WELLBYs to DALYs at a naïve ratio of 7:1. (See page 55 in Wellbeing Guidance for Appraisal). Non-EA.
  • Furukawa et al. 2021: Used a mapping dataset to convert PHQ-9 scores to EQ-5D scores, which are the basis of QALYs. We can then naïvely convert QALYs 1:1 to DALYs, although there is little precedent here in EA, to my knowledge. A nice feature of this dataset is that it has different scales for PHQ-9 scores at pre- and post-, to represent the importance of expectations of treatment in quality of life.

I’d like to focus on the Founders Pledge approach, because the Global Burden of Disease Study (GBD) is the generally-accepted source for DALYs within EA. A DALY in EA is calculated by multiplying out the GBD’s disability weight for a condition, by the length of time under which that condition is averted.

As such, if we could determine a single, ‘canonical’ way of estimating the disability weights produced by an intervention, we could get to DALYs very quickly. The challenge with disability weights today is that they are determined under the GBD by asking global survey respondents which of two described health conditions is ‘healthier’. For depression, respondents are given three descriptions:

  • Mild: Feels persistent sadness and has lost interest in usual activities. Sometimes sleeps badly, feels tired, or has trouble concentrating, but still manages to function in daily life with extra effort.
  • Moderate: Has constant sadness and has lost interest in usual activities. Has some difficulty in daily life, sleeps badly, has trouble concentrating, and sometimes thinks about harming themselves.
  • Severe: Has overwhelming, constant sadness and cannot function in daily life. Sometimes loses touch with reality and wants to harm or kill themselves.

The GBD then does some fancy statistical fitting to determine a weight (0–1) for each condition. Mild is 0.145, moderate is 0.396, and severe is 0.658. We also know from Salomon et al. 2015 that the weights are fit in logit-space, so we can reasonably interpolate between them.

However, we have no ability to assign PHQ-9 scores to this scale confidently—PHQ-9 scores weren’t used anywhere in the construction of the GBD. Existing approaches have assumed some kind of linear interpretation, but we must be careful about assuming that a GBD ‘mild’ is the same thing as a PHQ-9 ‘mild’ (to say nothing of the PHQ-9s ‘moderately severe’).

Converting SF-12 scores

One method that hews closely to the GBD would be to assess SF-12 scores. This is because the GBD’s statistical estimates work by comparing the survey preferences to a dataset of SF-12 scores for the same conditions. In Burnstein et al. 2015, the authors map SF-12 scores to GBD weights, and then assess different conditions in terms of SF-12 scores (as well as using third-party datasets for the same).

If we had a useful dataset for mapping PHQ-9s to SF-12s, then we could convert those through to GBD weights by either recreating Burnstein’s mapping function, or by asking Burnstein to produce their function directly.

I was able to identify two usable studies for this purpose:

  • The MEPS panel survey: This study from the U.S. includes paired PHQ-2 and SF-12 summaries. As the PHQ-2 is a subset of the PHQ-9, we can then map to SF-12s. This dataset was also usable by AI to replicate Burnstein’s SF-12 to disability weight mapping function. I don’t have enough confidence in the results here to include them here.
  • Yu et al. 2012: This asked PHQ-9s and SF-12s for 6,000+ people with depression in Hong Kong. My AI was able to use their summaries to produce a mapping between the two; one could also see if they still have the raw dataset.

However, I think this direction is less likely to be useful because (a) the two datasets we have aren’t representative of LMIC populations and (b) it involves an extra conversion step, for which we’re likely to see some loss. The PHQ-2 is particularly suspect, because although it’s useful as a diagnostic tool, people who are merely sad will score highly on it but not be generally depressed—this would potentially underestimate the disability weight of saturating the scale, relative to a depressed population.

Directly mapping PHQ-9s to disability weights

Another direction to explore is how to directly map PHQ-9s to disability weights. One way of doing this is for a researcher to subjectively decide how the three GBD bands correspond to PHQ-9 scores, using their intuition. This is how existing approaches have functioned. However, the choices of bands and mappings here impact the resulting DALY estimates quite meaningfully. It’s my opinion that a small amount of work to standardise this method could pay off fairly substantially.

Another somewhat more objective way would be to use a survey. Respondents could adopt a ‘persona’ as one of the three (+neutral) states, and then fill in a PHQ-9 survey as that person. Respondents could be human, or potentially AI. We could then fit a reasonable mapping between the scores.

The final option is to find populations of depressed people, have them fill in a PHQ-9, and then ask a single vignette:

One last question. Here are three people. Thinking about the past 2 weeks, which person is most like you?

  • Person A: Enjoys usual activities and time with others. Sometimes feels sad or low for a day or two when something goes wrong, but it passes. Sleeps, concentrates, and manages daily life as usual.
  • Person B: Feels persistent sadness and has lost interest in usual activities. Sometimes sleeps badly, feels tired, or has trouble concentrating, but still manages to function in daily life with extra effort.
  • Person C: Has constant sadness and has lost interest in usual activities. Has some difficulty in daily life, sleeps badly, has trouble concentrating, and sometimes thinks about harming themselves.
  • Person D: Has overwhelming, constant sadness and cannot function in daily life. Sometimes loses touch with reality and wants to harm or kill themselves.

I’ve invented Person A’s description as the inverse of the others, while the others are taken from the GBD (I believe a ‘none of these’ option would be too confusing). We could then fit a curve on these.

Because the EA mental health space is fairly well-connected, a coordinated effort here between organisations could produce fairly large datasets for this question across the global south—we could partner with charities in India, Ecuador, South Africa, Kenya, Zambia, Zimbabwe, Uganda, the Phillipines, and so on. Kaya Guides will be asking this question ourselves and I expect to put together a dataset soon (reach out if you’d like to partner on this!).

Understanding long-term effects of psychotherapy

Cuijpers et al. 2026 is an important meta-analysis, which was unable to rule out the hypothesis that depression’s effects may last indefinitely. If this is true, then existing EA CEAs may be dramatically underestimating the impact of psychotherapy, potentially by a factor of 2–15x. However, there isn’t enough existing evidence that can be used to estimate this.

If this were true, CEAs could also start to access the value of suicide prevention, since we could argue that psychotherapy permanently averts suicides, rather than merely postponing them. (In my own BOTECs, suicide prevention wouldn’t provide a meaningful part of a CEA today, because of how rare suicide is, combined with the assumption that suicide prevention only lasts until the next episode of depression).

In my opinion, we should model depression as an episodic disorder. If this is true, then a course of psychotherapy may shorten, or lessen the severity of, an existing episode of depression. Under the GBD’s own analysis (Appendix 1), the mean episode of depression lasts around 0.65 years. This means that by the time of 12-month follow-up, most study participants would have completed the episode they started the study with, in both arms. As such, 12-month follow-ups are only catching (1) participants in the midst of very long episodes or (2) participants who have begun a subsequent episode.

Such a study is then almost certainly underpowered to detect effects (if it were powered for post-treatment), as only ~10% of the study population would be in an episode at 12 months (according to a BOTEC I did to simulate a bunch of episodes). The same can be said for later follow-ups. It’s very rare that large enough studies are conducted to detect effects under this measurement paradigm—but this could be a direction for future research.

Another direction could be to develop or standardise methods for determining long-term effects. Under an episodic model, we would need to capture a longitudinal sense of the participant’s trajectory. We could do this in two ways:

  1. Periodic sampling: Every month, send a questionnaire to participants, over many years. We could then calculate the area under the curve.
    1. This method is likely to be very challenging, due to attrition. You would immediately run into sampling bias, as only dedicated people would fill in the survey. It would be expensive, if incentives were to be offered.
    2. An alternative to this method would be to find a source of data that is already measuring the desired outcome periodically, such as a mood tracking app. You could partner with this app, and encourage participants in your intervention to use it (or you could make your own intervention useful after the treatment term).
  2. Retrospective interviewing: There are two main precedents here, but neither is widely used. The first is the LIFE interview, and the second is the NIMH Life Chart.
    1. Measurement will be hard and expensive if using humans for the interviews. However, a team could convert one of these into a structured questionnaire, or use recent advances in AI to piece together the life chart.
    2. The methods don’t have to be particularly accurate to find an effect with reasonable enough statistical power. Some modelling could be done here to determine precise sample sizes under different assumptions of recall.
    3. Such a method could be paired with a more expensive or biased periodic sampling method, to validate its accuracy.

I suspect this is a good research direction for a funded academic study, although perhaps some easy wins could be taken with basic research.

Economic impacts of psychotherapy

Some studies of the economic impacts of depression put the value at a percentage of GDP. Given depression’s prevalence, this implies economic losses of ~20% of income—nothing to sneeze at. Meta-analyses of the effects of psychotherapy on labour-market outcomes find around ~0.15 standard deviations on most outcomes.

More work needs to be done to establish the precise effect sizes, and then convert these to income doublings, to be taken seriously by CG. Similarly, more work needs to be done in establishing appropriate discounts and methodologies for including these effect sizes in CEAs. Many outcomes simply don’t show up as statistically significant.

Similarly, another concern worth researching is that causal impacts of psychotherapy don’t seem to recover a proportionate amount of the burden of depression on GDP. It would be worth understanding why this is, and doing more work to quantify the effects.

Much of the work in this field has involved Johannes Haushofer, who Kaya Guides is collaborating with on our RCT. In the next few weeks, we’ll start our RCT with a very large sample size (around 3,500 participants), which by our count will be one of the largest ever studies of any psychosocial intervention in South Asia. The hope is that at these sample sizes, we’d be able to find effects on measures that other studies have struggled with (such as income at 12 months).

Conclusion

Overall, however, an impact of 0.15sd seems unlikely to move the needle. Improving PHQ-9/DALY conversions improves the credibility and legibility of mental health CEAs for relatively cheap, and studying the long-term effects of mental health interventions could yield substantially larger effects than previously estimated. So economic impacts seem like the least promising direction.

As well as work to improve mental health CEAs, organisations in the space need to communicate better about what has changed since GiveWell’s CEA of StrongMinds. Kaya Guides are guilty of this ourselves, and I’m hoping to share more on this soon.

If you’re interested in pursuing any of these research questions further, please reach out to me!


  1. Sorry, I owe everyone an update on this ↩︎

76

1
0
1

Reactions

1
0
1

More posts like this

Comments7
Sorted by Click to highlight new comments since:

(We've talked about most of this outside of the public view, but I can't restrain myself from commenting none-the-less)

  1. Legibility

I agree with you strategically that people who care about mental health need to make their case in a way that's legible to more traditional metrics like income, DALYs, etc. But the general thing this points to is greater legibility --> greater uptake --> impact blah blah. 

On this general point, one thing we've been thinking about and piloting is just trying to make WELLBYs more legible, especially for MH interventions by converting them to years of depression averted (YODAs). 

So I could say that a fictional charity Gaya Kuides generates 1,000 WELLBYs per $1k donated. Or given that the typical case of depression (moderate depression) is associated with a -1.3. lower wellbeing compared to low depression symptoms -- this implies that actually GK averts the equivalent of 1000 / 1.3 =769 years of depression per $1k donated. FWIW we explain the YODAs more here and we're looking for feedback on whether this seems like a useful communication device. 

It just seems way easier to imagine the value of a year of depression averted compared to "WELLBY". 

2. Long-run effects of psychotherapy 

This is very important and understudied. The effects of psychotherapy could last much longer, and we could maybe make these interventions benefit people for much longer than we. But we don't care much about LR effects in general in the research community and we do not have much understanding of the mechanisms that support long-term benefits. I obviously wish that we'd change on both accounts. 

  • I think the Cuijpers stuff is interesting, and I think the results are consistent with the existence of the long-term effects of psychotherapy. I take this from ignoring the splines and just seeing that there are observations of effects years later.
  • Sure it doesn't rule out indefinite effects. But you could say the same thing about finding positive effects after 1 month, so I'm less gung-ho on the language.
  • Some methodological concern with Cuijpers et al. they only present and discuss one sort of unconventional model (why not a linear model? It looks like a linear model would predict decay but I obviously can't tell). They also don't do a lot of (any?) tests like seeing how stable these are when we remove studies or whether there's publication bias where you only do a follow-up when you think you're going to find something (you could interrogate this by seeing if earlier effects are higher than effects in that same time window). Basically I'd be more reassured if they ran and reported a lot of different models and robustness tests.
  • Mechanisms! Why do the benefits of mental health interventions persist and decay? I'd guess that the mechanisms are some mix of A) access to and utilization to new tools for thinking. These require practice and memory so I expect we will see decay without some structure to continue maintenance. B) A change in believes around mental health or in one's agency to affect their MH. This seems like this part is something that could plausibly last much longer as a result of a short intervention. If you make someone believe this is a solvable problem, then that could benefit them for life. But note I expect that the belief mechanism is probably more important in LMICs where stigma and traditional believes around mental health (it's demons or curses, etc) is more common. This would make me think it's likelier that the benefits last longer in LMICs than HICs. But this is the opposite of what our meta-analytic results suggest (effects last 2-6 years in LMICs) versus Cuijpers implied you treat someone in HICs and it benefits them indefinitely.
    • I guess back to Cuijpers... one thing that could explain the LR results is that in HIC after you start getting treatment even if the RCT ends you may just keep getting therapy. That presumably doesn't happen in LMICs and may explain the discrepancy. But I'd need to dig into the Cuijpers data to interrogate that.
  • Mechanisms part 2: treat the young. It seems plausible that you can potentially change lifelong trajectories by treating earlier. This is commonly assumed but there's not too much evidence I've found (e.g., Bhalotra et al., 2025 seems to suggest LR positive effects but I think there are some other econ papers that find null LR results for natural experiments increasing access to MH treatment for youth).

Always welcome Joel!

  • I’m generally pro-YODA (although I think the name is perhaps a little unserious?). I think it’s very legible to people who care about mental health, or wellbeing more generally, and I like that it’s on the same order of magnitude as a WELLBY so the conversions are easy to do in one’s head. I think DALYs are probably still going to be the only way people in EA global health will care about psychotherapy, however.
  • I think if anything, I’ve approached the Cuijpers paper from the mechanisms first. In my personal experience, I’ve used things I’ve learned in therapy over and over, (but I’m not chronically depressed). It seems like basic skills (boundary setting, emotional regulation, behavioural activation) could be like learning to ride a bicycle.
    • My mental model of severe mental illness is that you have little control (without pharmacology) over the weather in your mind, but you do have some ability to regulate your response to those feelings.
    • So to me the Cuijpers paper reads as not disconfirming the belief I already held
  • On treating the young: My best guess is this would make sense after a certain age, but it’s unclear what the age is. There’s decent evidence to suggest conventional psychotherapy doesn’t work on children, last I checked (memory may be completely off here).
  • I read through the paper and although they try some different fits for the cubic spline, they never explain this choice as opposed to other modelling methods
  • They did test regions, and found no effect. But the ‘Other’ region includes the Iran papers HLI have discussed at length before.
  • There’s no moderation for whether people pursued therapy after the trials. I guess, in fairness, this is more of a question of increasing costs in a CEA, since the continued care would be counterfactual.

Yeah, anyway, I think the paper basically raises the need to investigate further.

I am inspired by your and @huws attention to detail in the appropriate measurement choices for this. 

As a clinical psychologist since more than 15 years i can imagine quite a few ways short therapies can have effects lasting a long time. If you find generating more possible explanations interesting it would be fun to expand on this. 

I think i could use your critical and experienced eyes on a post i just wrote called "Closing the wisdom gap: Learning from the failure of universal school-based mental health interventions"
https://forum.effectivealtruism.org/posts/dEm7h7vDW3iwLnsZi/closing-the-wisdom-gap-learning-from-the-failure-of

In that post i expand on the idea of early and ambitious efforts not only for wellbeing but also, wisdom, happiness, reduced criminality and more. I have a short TL;DR in the beginning if you are curious. 

I would truly appreciate help on the measurement, impact calculation side of things. I used Claude as a help in making some simple scenario calculations. But since the post is mostly written after the kids have fallen asleep i did not have time to truly dig deep and develop this part of the post. I focused more on the other parts of this post. Trying to share my vision and understanding of the science behind bringing change during kids development.

There are so many interesting questions out there to be answered :)

I am glad mental health is also being seriously considered within the EA community. It was one of the causes I missed seeing in the forum when I entered, and this post was spot on. 

As a psychologist myself who has been treating several years with clients regarding self-esteem issues, I really hope your charity and this cause as a whole has a deep impact!

Thanks for this, really useful post.

We’re preparing a pilot on one specific source of error in PHQ-9 data. The instrument isn't anchored to menstrual cycle phase, leaving hormonal state as an unmeasured confound. In programmes where most participants are women, this could be shifting the very numbers you’re trying to make credible, and right now nobody is accounting for it.

It seems especially relevant to your periodic-sampling idea: a 'regular' cycle is about the length of a month, so sampling on a fixed date each month could drift across cycle phase and surface as a slow trend rather than noise. Methods to detect this already exist, and we have a pilot proposal built to measure the effect directly using an existing programme’s PHQ-9 data.

I’ll follow up directly, as I would value your eyes on it.

Thank you! I’ll get back to your email.

My sense is that cyclical data like this should come out in the wash, but just require more statistical power. It would depend on the amplitude of the menstrual cycles, if it’s only a couple of PHQ-9 points per person then it should be fine. I mean, there isn’t any evidence showing men and women have differential improvements from psychotherapy, which you’d expect if this was having a large impact.

Thanks, Huw! You’re right that in a randomised trial, cycle phase is noise rather than bias, and that whether it matters depends on amplitude. I should have been clearer about that.

The best prospective estimates I’ve found are small on average. Klusmann et al. (2024), three ratings a day across one cycle on an adapted PHQ, found an average amplitude of 0.84 on a 0–5 item scale; that’s only a rough proportional conversion, not a validated mapping, but it suggests movement on the order of one PHQ-9 point peak-to-trough, with a heavier tail of larger individual amplitudes (31.2% of participants were above a 0.25 amplitude for depressed mood, the highest share of any symptom). 

Delray et al. (2025), 352 women with depression tracked over two or more cycles through a mood app, found a perimenstrual dip of 0.13 SD below each person’s own average, present in about half the sample. That is the “couple of points” scenario you describe. Where it stops washing out is in non-randomised programme data: if people enrol disproportionately in a symptomatic phase, intake scores are elevated, and later scores are at random phases, which inflates pre-post change. That’s a selection mechanism, not a variance one, and nobody has tested it in a programme that measures weekly. Premenstrual exacerbation in depression itself is documented in a systematic review of 35 prospective studies across several disorders, depression among them (Nolan & Hughes, 2022).

For context: I’m not a statistician. I run a fairly new nonprofit on mental health education and advocacy focused on women and the question of cycle phase came from that work. This is my first contribution here, and I used AI tools to help me read and check the papers; any errors are mine. I’ve sent you the sources by email so the thread doesn’t fill with citations.

 I am also happy to share the sources with anyone interested in this line of research. 

More from huw
Curated and popular this week
Relevant opportunities