What We Learned From Red Teaming Our Iron Grantmaking

In a nutshell

GiveWell has made ~$24 million in grants to iron supplementation and fortification programs since 2017. We think these programs are among the most promising in our nutrition portfolio, and we’re considering roughly ~$10-20 million in grants in 2025.

Following our red teaming of GiveWell’s top charities, we decided to review our iron grantmaking to understand what were the top research questions we should address as we consider making additional grants in the near future.

Overall, we didn’t find any issues that significantly undermined our view of iron programs. However, we did surface some issues that we think could influence our iron grantmaking in 2025 and beyond. We think further investigation of these issues could cause us to make ~$1 million to ~$10 million in grants we might not make otherwise and cause us not to make ~$1 million to ~$10 million in grants we would’ve otherwise per year over the next five years. We expect to learn more about these issues and follow up in 2026 once we’ve looked into them.

Published: March 2025

Summary

What we did

In June and July 2024, we conducted “red teaming” of our iron supplementation and fortification grantmaking. A team of four researchers spent one week scrutinizing the case for these programs and considering what research we should prioritize over the next year as we consider more grants in these areas. In parallel, we also received feedback from three iron experts. (more)

Main issues we found

Red teaming surfaced the following priority issues to investigate further, ranked roughly in order of potential influence on our grantmaking. We’re unsure whether these are mistakes (that led to important errors or worsened the credibility of our research and that we should have caught already) or just issues that are worth looking into now that we’re considering more grants to iron programs. (These findings are based both on internal analysis and feedback from external experts we spoke to. A summary of external reviewers’ comments is here.)

  • Insufficient attention on more recent research on the effect of supplementation on anemia and publication bias in these estimates. Our analysis relies on a 2013 meta-analysis showing a ~50% reduction in anemia from iron supplementation. However, a more recent meta-analysis from 2023 finds a ~40% reduction. That meta-analysis also raises concerns about publication bias, with some outcomes showing 60-80% smaller effects when adjusted for small-study effects. (more)
  • Insufficient attention on anemia burden. We estimate ~60-70% of children under 5 in India have anemia, based on data from the Institute for Health Metrics and Evaluation’s (IHME) Global Burden of Disease (GBD) study. While this estimate aligns roughly with India's National Family Health Survey (NFHS), we still have some concerns. For example, one expert noted that capillary blood draws might systematically overstate anemia rates if blood samples are diluted through improper collection techniques. There may be other data sources we should triangulate against, too. (more)
  • Insufficient attention on disability from anemia and morbidity in general. Our analysis relies on disability weights from the GBD to value anemia's impact on quality of life. While our previous review found these weights to be broadly reasonable, we think there are ways to probe this further. We also have not yet explored alternative ways to quantify disability from anemia, such as comparing anemia’s effect on well-being measures (such as cognitive scores, fatigue, or subjective well-being) or looking for studies that investigate how individuals value trade-offs between anemia and consumption. More generally, we’ve devoted less time to trying to value improvements in morbidity (compared to valuing reductions in mortality) across our research. (more)
  • Insufficient attention on additional data collection. Our current analysis relies heavily on extrapolations from existing evidence and subjective adjustments. For example, we make numerous adjustments for factors like variation in the effectiveness of different fortification compounds, fortificant concentration, and the effect of inflammation on iron absorption. Rather than continuing to refine these in our model, we think collecting new empirical data could be more valuable. This could include pre-post studies or randomized controlled trials (RCTs) to measure the effect of iron programs on anemia or more targeted research to understand, for example, the effect of consumption of inhibitors on iron absorption in program settings. (more)
  • Insufficient attention on iron supplementation and fortification in Sub-Saharan Africa. Our iron-related grantmaking has focused primarily on India. However, we may be missing promising opportunities in Sub-Saharan Africa, where anemia rates are also high. While there are concerns about iron increasing malaria risk, our initial analysis suggests this does not have a meaningful effect. Two iron experts we consulted noted that this risk might be less concerning for iron fortification programs specifically. We plan to consider opportunities in Sub-Saharan Africa, including pairing iron programs with malaria prevention programs like seasonal malaria chemoprevention (SMC) or insecticide-treated nets (ITNs). (more)

These findings should be viewed primarily as a prioritized research agenda rather than final conclusions. As we investigate further, our views on both the severity and implications of these issues may shift significantly. We'll also assess whether any issues constitute genuine mistakes in our previous analysis.

Implications for grantmaking and next steps

Overall, we continue to think iron supplementation and fortification are among the most promising opportunities within our nutrition portfolio.1 However, we expect that addressing the issues identified during red teaming will lead to some changes in our grant decisions over the next five years.

Our best guess is that these changes will result in:

  • $1 million to $10 million more in grants we wouldn't have made otherwise (e.g., because we discover promising opportunities in Sub-Saharan Africa where anemia burden is high and programs could be paired with our existing malaria prevention work, or we identify high value studies we could fund to learn more about iron programs)
  • $1 million to $10 million less in grants we would've made otherwise (e.g., if recent meta-analyses suggesting smaller effects and publication bias lead us to revise our effectiveness estimates downward, or we find that burden is lower than we’ve assumed)

These estimates on changes in our grantmaking are highly subjective. We haven’t tried to explicitly model out changes but included these best guesses to give a sense of the magnitude of changes we expect. Because we haven’t fully investigated these issues yet, we’re uncertain about their impact on our grantmaking, and their impact could end up being more or less significant than we currently anticipate.

We’ve already begun to look into some of these issues, and we expect to look further into them in 2025. We are currently in the process of writing an iron supplementation and fortification report as an update to our 2020 iron fortification intervention report and our 2019 iron supplementation intervention report, partially in response to the issues raised during red teaming. Due to the fact that we are continuing to update our cost-effectiveness model, the figures in this page may not reflect our most updated thinking. This post is meant to reflect our thinking at the time that we concluded our red teaming.

Our next steps:

  • We plan to continue exploring promising iron supplementation and fortification programs and prioritize investigation of these issues as we evaluate specific grants.
  • We plan to follow up on these issues and report back in 2026 on what we've found, including any significant updates to our analyses or recommendations.
  • We’re also considering how some of these issues apply to other programs beyond iron. For example, we expect insufficient consideration of costs and burden and insufficient consideration of ways to improve our research by funding additional data collection over model tweaks to potentially apply to other grantmaking areas, too.

Main reservations about findings from red teaming and limitations

  • We did not comprehensively vet all of our work on iron supplementation and fortification. As a result, it's likely we've missed some issues.
  • We spent about one week red teaming our iron work, compared to two weeks for each of our top charities, meaning it’s more likely we missed things.
  • This red teaming focused heavily on analyzing our evidence and cost-effectiveness calculations, rather than real-world program considerations. While we tried to look at practical factors (like expanding to Sub-Saharan Africa and collecting new data instead of just refining our models), we probably didn't spend enough time on real-world implementation challenges, actual program costs and adherence rates, feedback from implementing organizations, feedback from local stakeholders about programs, detail on how programmatic activities led to uptake of iron supplementation and fortification, and broader questions (e.g., “how else could we support anemia reduction?” or “should we be fortifying other foods?).
  • We chose expert reviewers based on iron program expertise, potentially skewing us toward favorable views of these programs. We did not actively seek out skeptical perspectives that might have highlighted additional concerns.
  • It's possible that the issues we did identify may end up having a limited impact on our grantmaking or being intractable.

What did we do?

Iron deficiency involves an insufficient supply of iron to cells, which can affect their growth and development. Iron deficiency is a common cause of anemia and may lead to a range of adverse physical and cognitive effects.2 Iron deficiency is most commonly addressed by iron supplementation or iron fortification programs. Iron supplementation can be administered orally, intravenously or intramuscularly.3 Iron fortification programs usually involve mandatory, centralized mass fortification of staple foods, such as wheat flour. 4

As of January 2025, we’ve made ~$24 million in grants to iron programs:

Over the next year, our best guess is that we’ll make ~$15 million in grants to programs like wheat flour fortification in India or iron fortified rice in Indonesia, although it’s possible that we make more or less than that.

In anticipation of this, we “red teamed” our grantmaking to iron supplementation and fortification to identify areas we should consider looking into as we make new grants.

During red teaming, we had a team of four GiveWell staff work full-time for one week during June and July 2024. This included staff who were familiar with our iron grantmaking and staff who were less familiar with this work.

The red team relied on an (unpublished) draft of an updated version of our 2020 iron intervention report and 2019 iron supplementation report and an accompanying cost-effectiveness analysis (CEA).7

The goal was to answer the following questions:

  • What do we think are the top issues we should look into across our iron grantmaking?
  • What are the ways we could be over- or underestimating the impact of these programs?
  • Are there ways we could approach our iron grantmaking differently?

In parallel, we asked for feedback on our reports from three external experts. We asked experts:

  • Is our case for why this program is an effective giving opportunity clear? Which parts are unclear?
  • What are the ways you think we could be wrong? Which assumptions or claims seem off to you? Why?

We’ve incorporated findings from these experts into conclusions in the next section on the top issues identified during red teaming. We also provide more detail on the expert reviews in this section.

Where were the top issues we found?

Insufficient attention on more recent research on the effect of supplementation on anemia and publication bias in these estimates

In our analysis of the effect of iron supplementation, we rely on a meta-analysis from Low et al. 2013, which finds a 50% reduction in anemia due to iron supplementation.8

During red teaming, we surfaced some questions that make us think this could be worth looking further into:

We think this could mean we’re overestimating the effect of supplementation.

We also think this may be indicative of our overlooking of publication bias in our research more broadly. For example, we also recently discovered a meta-analysis we rely on for estimating the effect of vitamin A supplementation on mortality may have significant publication bias that led us to overestimate effects.12

There are also reasons to think this is less of a concern.

  • Andersen et al 2023 may use studies that are less directly relevant to our contexts (e.g. more high-income country studies, lower doses of iron).
  • There may not be publication bias. Funnel plots and Egger tests have limitations. For example, smaller studies may be systematically different from larger studies (e.g. use larger doses of iron, treat different populations), leading to different effect sizes. We have not scrutinized the funnel plots, Egger tests, and trim-and-fill adjustments in Andersen et al 2023 in depth.

To address this, we plan to:

  • More closely compare the results in Andersen et al. 2023 to Low et al. 2013.
  • Review publication bias concerns in the iron supplementation literature and decide whether to make an adjustment to account for them.
  • Consider conducting our own meta-analysis of studies from Anderson et al. 2023 but limiting it to low- and middle-income countries.
  • Do an audit of other programs we fund, beyond iron supplementation and fortification, to ensure we’re accounting for publication bias consistently.

Insufficient attention on anemia burden

A major driver of high cost-effectiveness of iron programs is the high prevalence of anemia in countries like India. However, we’ve devoted less attention to this than other parts of our analysis (like efficacy of iron fortification or supplementation at reducing anemia).

We noted in our red teaming of top charities that taking burden of disease estimates at face value may lead us to make bad grant decisions.13 This makes us think there’s a risk of that here as well.

Some specific concerns we have about anemia burden:

  • We briefly looked into whether the IHME’s burden estimates are similar to other sources. Their estimate of anemia in India is roughly in line with NFHS-4 and 5 (~60-70% anemia in children under 5 and 50-60% in women of reproductive age).14 However, a few questions came up during our investigation that made us slightly more concerned about the validity of the “ground truth” numbers from NFHS.
    • In a brief review, it seemed that the estimates from NFHS-5 are controversial due to the method used,15 but we didn’t dig into the merits of the complaints.
    • One of the experts we spoke to during red teaming, Dr. Reina Engle-Stone, mentioned on our call with her that it may be possible that capillary draws could systematically overstate anemia if the people drawing the blood squeeze the finger particularly hard and end up getting a dilute sample of blood.
    • The Indian government is planning to drop anemia measurement from NFHS-6 and collect it instead in a diet and biomarkers (DABS-I) study16 using a different method (the venous method17 ). We aren’t sure if this decision to change survey methodology reflects a distrust of past NFHS anemia estimates that used the capillary method.
  • The Indian government has been funding several interventions aimed at reducing anemia.18 However, anemia levels have risen slightly in the data we’ve seen. Is this because anemia figures are off? Is this a negative update on the effectiveness of these programs?
  • It is surprising that India’s levels of anemia are so much higher than, for example, parts of Sub-Saharan Africa where income levels are lower and other measures of health are worse. This makes us somewhat concerned about whether data quality issues are driving this pattern. However, Dr. Reina Engle-Stone notes that high rates of anemia and iron-deficiency in India could also be consistent with low consumption of animal-sourced foods in some populations.19
  • We’d previously received criticism arguing that hemoglobin “cutoffs” used to define anemia should be lower in India than the current WHO standard and that current cutoffs inflate measures of anemia. We considered this and decided we don’t think we should lower cost-effectiveness to account for this.20 However, this question is complex and makes us think there may be other issues related to measuring anemia burden that make it difficult to assess.
  • There may be other data sources we can compare IHME to, besides NFHS. It’s possible these may be more out of sync with IHME.

To address this, we plan to:

  • Look for alternative measures of anemia and compare them to IHME and NFHS.
  • Talk to experts about alternative ways to measure anemia.
  • Decide whether it’s worth collecting additional data on anemia burden.

Insufficient attention on disability from anemia

In our cost-effectiveness analysis, we try to estimate the disability from anemia morbidity so that we can compare the benefits of reducing anemia to, for example, the benefits of reducing child mortality or increasing income.21 We think this has received comparatively less scrutiny than other parts of our analysis.

Some areas we think it could be worth digging deeper into:

  • Our current approach to valuing anemia is to multiply the disability weight22 for anemia by “units of value” we assign to a disability-adjusted life year.23 How reliable are the GBD disability weights for anemia? How are they calculated? What do they include, and what do they exclude? In a previous analysis of GBD’s weights, we concluded: “Overall, we believe Global Burden of Disease disability weights for anemia are fairly reasonable reflections of the physical impairment it imposes, but they may overestimate the burden of moderate and severe anemia, depending on how severe it is. We are uncertain about this tentative conclusion and would like to discuss it with experts who have experience working with anemic populations and individuals.”24
  • Are there alternative ways to measure the improvement in quality of life that comes from iron supplementation and fortification?25 Do we have any data on effects of fortification on symptoms of anemia (cognitive tests, measures of physical endurance/fatigue) or measures of subjective well-being? Can we use these to help address some questions about anemia burden and whether reductions in this translate to meaningful improvements in well-being? Are there studies that try to elicit how much income an individual would be willing to trade off to avert anemia? We haven’t looked at these issues in depth and think it’s possible this could update our view. In general, we’ve done relatively less work to understand trade-offs between morbidity and consumption than mortality and consumption in our moral weights.

To address this, we plan to consider alternative ways to value quality of life benefits from anemia and potentially other programs to address morbidity.

Insufficient attention on additional data collection

Our iron supplementation and fortification analyses require making several extrapolations from existing evidence and subjective adjustments.

In our current, unpublished model, we rely on a meta-analysis by Field et al 2021, as well as several other studies we think are relevant. Then we make several external validity adjustments to account for differences between these studies and programs where we’re considering funding, including differences in fortification compounds used, amount of fortified food consumed, fortification concentration, differences in diet, quality of the fortification process, etc.26

We’ve spent a decent amount of time trying to account for these with various adjustments in the model and still have several uncertainties about them.27

In addition, experts have expressed concerns about using Field et al, and a lack of relevant evidence more broadly.28

Given this uncertainty, we guess the best marginal use of subsequent research hours is on data collection to resolve these uncertainties rather than doing more rough adjustments to try to account for them.

We’re not sure yet how this data collection would look, but it could include:

  • Randomized controlled trials that try to measure the effect of iron fortification or supplementation on anemia or iron deficiency in an area we expect to direct a lot of funding to.
  • Pre-post or other less rigorous studies to measure anemia or iron deficiency before and after introduction of a large fortification or supplementation program.
  • More targeted data collection designed to get at specific uncertainties. For example:
    • A household survey to understand how much iron households consume and how much is consumed by different household members (e.g., child and women who have higher rates of anemia).
    • A survey to understand diets and consumption of possible inhibitors to iron absorption (tea and phytic acid).29

To address this, we plan to decide on upcoming grants whether to pair with this additional data collection. We guess that we won’t do an RCT before making an additional ~$10 million in additional grants, but if we were to make larger grants, we would likely consider an RCT.

Insufficient attention on iron supplementation and fortification in Sub-Saharan Africa

Our iron-related grantmaking has so far been focused on India. We’ve made ~$14 million directed to iron and folic acid (IFA) programs via Evidence Action30 and ~$9 million to Fortify Health for iron fortification.31 It’s possible we’re missing promising opportunities to increase uptake of iron supplementation and fortification in Sub-Saharan Africa.

We think there could be promising iron-related opportunities in Sub-Saharan Africa because:

  • Anemia rates are high in Sub-Saharan Africa. In several countries, over half of women of reproductive age are estimated to have anemia.32
  • We’ve heard some concerns that iron may increase the risk of malaria infection.33 However, the WHO recommends iron supplementation and fortification when paired with malaria control efforts34 (e.g., SMC and nets, which we support in several countries in Sub-Saharan Africa) and has been recommended by other experts as well35 , and an iron expert we spoke to recently told us this issue might be less pressing for iron fortification specifically.36 Our current view on this is that iron supplementation and fortification alone do not have a meaningful effect on malaria risk, although we have some uncertainty.
  • We’ve heard that levels of infection may inhibit iron absorption and are unsure how much this would impact the effectiveness of iron supplementation and fortification in Sub-Saharan African countries, which we guess have high infection prevalence.37 We’re aware of some research on slow-release iron, which may be less likely to exacerbate infections, though we have not reviewed this in any depth.38

To address this, we plan to look further into the cost-effectiveness of iron supplementation and fortification programs in Sub-Saharan Africa. We’ll start by talking to experts about:

  • What potential opportunities there are to support iron supplementation and fortification in Sub-Saharan Africa.
  • Concerns about malaria risk and effect of baseline levels of infection on iron absorption in Sub-Saharan Africa specifically.
  • The extent to which pairing iron fortification or supplementation with other malaria programs we’ve funded (e.g., SMC or ITNs) is feasible and likely to affect cost-effectiveness of iron programs.

What did experts say?

This section provides more detail on feedback from expert reviewers on our iron supplementation and fortification grantmaking. These reviewers were provided with a draft version of our updated iron CEA. We reference their feedback in the previous section but provide a more comprehensive summary, including links to full notes, below.

Three experts provided feedback on our iron supplementation and fortification analysis:

  • Dr. Reina Engle-Stone, Associate Professor in the Department of Nutrition at UC Davis. Dr. Engle-Stone’s written notes are here.
  • Dr. Katherine Adams, Associate Project Scientist, University of California, Davis. Dr. Adams’ written notes are here.
  • Dr. Sue Horton, Professor at the University of Waterloo (partial review). Our rough notes from a conversation with Dr. Horton are here.

None of the experts identified issues that fundamentally undermined the case for iron programs, though they raised several important considerations for further investigation. These experts were consulted because they had expertise on iron supplementation and fortification programs. It’s possible this means we’re skewing toward individuals more likely to say favorable things about these programs. We did not explicitly seek out individuals who are skeptical of iron programs.

The issues raised by these reviewers are listed below ordered by our own assessment of importance.

Development effects

Two experts raised questions about our approach to estimating development effects39 :

  • Dr. Horton expressed skepticism about our reliance on a historical study of the United States (Niemesh 2015) to estimate development effects from iron supplementation. She thinks it's likely that there were other important changes going on during this time period (1940s to the 1970s) that aren't fully controlled for in the model. We agree there are reasons why Niemesh 2015 might overestimate effects. We are accounting for this with a downward adjustment to its effect size, but we’re highly uncertain about the appropriate magnitude of the adjustment.40
  • Dr. Adams suggested considering additional evidence beyond the Niemesh study, particularly looking at studies linking cognitive development to iron status and income. We agree that it’s worth doing more “cluster thinking”41 on the effects of iron on later-life income by bringing in other evidence. During red teaming, we did some quick comparisons to other research. We plan to incorporate these comparisons into our updated iron intervention report.42 An example of using these types of comparisons with other interventions’ development effects is in our analysis of distributing ITNs.43
  • Dr. Adams also questioned whether our assumption that 19 years of coverage are needed to achieve the full effect found in Niemesh 2015 might be overly conservative. If we were to assume effects are concentrated among children under 5, then that would decrease the years of coverage needed to achieve the full effect and increase effects per year (since we divide the effect size by the number of years needed to achieve the full effect). We hadn’t considered this before and think it’s worth looking into further. We model development benefits for other interventions as taking 15 years of coverage for the full effect.44 In an internal analysis of our forthcoming CEA, shifting from 19 to 15 years to achieve full coverage increased the cost-effectiveness of iron supplementation and fortification by ~10%.

Household-level distribution of iron

Dr. Engle-Stone identified household-level food distribution as a potential concern, noting that those most likely to be anemic (children under 5 and women of reproductive age) may consume less than average amounts of fortified foods per capita. In our model, we rely on average estimates of fortified food consumed per capita, but the people most likely to be anemic may also be eating less than average amounts of fortified foods. We roughly guess this would reduce cost-effectiveness by 5-10%, but we plan to look into this further.45 We also think this is an area where more data collection might be more beneficial than substantial desk research (see above).

Costs

Dr. Adams noted that our estimated cost of iron fortification of wheat flour seems high compared to the estimated costs from the literature. We think is driven in large part by our iron fortification costs coming from Fortify Health, which operates at a smaller scale. In the future, we expect costs to fall and become more in line with other estimates as Fortify Health’s program scales. We’ll also be able to see how actual costs per person reached compared to projected costs.

Meta-analysis for iron supplementation estimate

Dr. Engle-Stone recommended using Andersen et al. 2023 instead of Low et al. 2013 as the basis for estimating iron supplementation's impact on anemia, as it's more recent, covers a more relevant age range (0-19 years), and reports both iron deficiency and iron deficiency anemia as outcomes.. Andersen et al. shows a smaller reduction in anemia prevalence (39%)46 compared to Low (50%).47 We agree this is worth looking further into. We discuss this above.

Sense checks on our analysis

Dr. Adams did two sense checks on our analysis:

  • She points out that our fortification model implies that we're expecting a larger anemia reduction effect in at-scale programs (25% reduction) than was found in the RCT meta-analysis (22% reduction), and that this doesn't seem very plausible. Our main explanation for this is that, while we make some downward adjustments, these are balanced out by a ~30% upwards adjustment because we think the bioavailability-adjusted iron dosage received by fortification program participants is considerably higher than what was used on average in the trials.48 However, we plan to do additional work on this adjustment in advance of our next iron grants.
  • Dr. Adams reviewed the magnitude of economic benefits49 we're estimating for iron, relative to a table of other interventions, and concludes, "In the context of other estimates in the literature, the estimated economic benefits used in this CEA seem reasonable and perhaps even a bit conservative." We had not done this type of check before and haven’t vetted her analysis, but this sort of comparison gives us some additional confidence.

Additional benefits we’re not incorporating

All three experts mentioned a potential benefit we’re not incorporating:

  • Dr. Engle-Stone recommends explicitly considering the benefits of other micronutrients in our model. We’ve recently tried to estimate the benefits of folic acid fortification, and our best guess is they would be a small share of overall effects (~1x cash). However, this is from an unpublished initial analysis, and it’s possible we could update substantially with more work.50
  • In our model of iron programs, we include the benefits of increased productivity for workers based on increases in wages. Dr. Adams recommended considering broader benefits of higher productivity (such as spillover effects to the broader economy due to increased productivity of workers receiving iron). We haven’t tried to quantify this benefit.
  • Dr. Horton thinks that there may be maternal mortality reduction benefits of iron programs, which we don't include in our model, and that there may also be additional benefits for infants beyond maternal mortality that we aren't accounting for (such as cognitive gains).
    • We’ve chosen to exclude maternal mortality benefits because we have not found strong evidence that iron reduces maternal mortality.51 We’ve also conducted a brief, back-of-the-envelope calculation to estimate the impact on cost-effectiveness if we assume some impact on maternal mortality.52 We guess the overall impact in India is small because (a) pregnant women are a small proportion of the total population (16 births per 1000 people)53 , (b) maternal mortality is relatively small (97 per 100k births)54 and (c) the effect of iron on maternal mortality is likely to be small (1-2% decrease).55
    • We guess that the additional benefits to infants that Dr. Horton mentions would be incorporated into our development effects estimates.56

Supplementation adherence

Dr. Engle-Stone noted that another way one of these programs could fail to have an impact is if supplementation adherence is actually not very high. We’re accounting for adherence in our cost per supplement estimates, based on the Evidence Action programs we've funded in India, but we haven’t thoroughly interrogated these adherence estimates.

Inflammation adjustment

Dr. Engle-Stone expressed skepticism about how tractable it is to get a quantitative estimate for this but said a “small downward adjustment” seems reasonable. We’re also uncertain about this parameter.57

Funding a new trial

Dr. Horton noted that if we were interested in understanding how much iron deficiency contributes to anemia in a specific population, we could consider funding a new iron supplementation trial in that population (e.g., adolescent girls or young children in a state in India). The idea here is that high-dose iron supplementation might provide an estimate of a cap on the proportion of anemia cases that are responsive to additional iron intake in a particular population. We plan to consider whether additional data collection could help us narrow some of our uncertainties above.

Relationship disclosures

As of February 2025, we have conditionally approved funding for a proposal from a research team at Walter and Eliza Hall Institute, led by Dr. Sant-Rayn Pasricha, who served as an expert consultant on our prior iron grantmaking and is referenced in several places in this report. Details of the proposal will be published if we move ahead with funding.

Sources

Document Source
Andersen et al 2023 Source
Anemia Mukh Bharat (AMB), Homepage Source (archive)
Bleakley 2010 Source
Clark et al 2014 Source
Cost-effectiveness of iron fortification in India is lower than GiveWell's estimates, Kulgod, 2022 Source (archive)
Cutler et al. 2010 Source
DHS Program, Hemoglobin Collection at The DHS Program: Impact of Updated WHO Guidelines on DHS Program Anemia Data, 2024 Source (archive)
Evidence Action, Homepage Source (archive)
Fan 2016 Source
Field et al 2021 Source
Food Fortification Initiative, Homepage Source (archive)
Fortify Health, Homepage Source (archive)
Freeman and Sutton, 2020 Source (archive)
GBD 2021 Anaemia Collaborators, 2023 Source
Givens, Anitha, and Giromini, 2024 Source
GiveWell, A response to a Change Our Mind Contest entry on iron fortification programs in India, 2023 Source (archive)
GiveWell, All iron and folic acid supplementation grants, Airtable Source
GiveWell, All iron fortification grants, Airtable Source
GiveWell, Anemia Prevalence data pull, 2024 Source
GiveWell, BOTEC - Effect of iron supplementation on maternal mortality [January 2025] Source
GiveWell, Examining the cost-effectiveness of iron supplementation Source
GiveWell, Fortify Health – Support for Expansion (December 2021) Source
GiveWell, Fortify Health — General Support (2018) Source
GiveWell, Fortify Health — General Support (2019) Source
GiveWell, GiveWell's moral weights (2020) Source
GiveWell, GiveWell's CEA of insecticide-treated net (ITN) distributions, 2025 Source
GiveWell, GiveWell's Change Our Mind Contest, 2022 Source
GiveWell, Grants Airtable Source
GiveWell, Iron CEA comments from Dr. Katie Adams, 2024 Source
GiveWell, Iron CEA comments from Dr. Reina Engle-Stone, 2024 Source
GiveWell, Iron CEA comments from Dr. Sue Horton, 2024 Source
GiveWell, Iron Fortification, 2020 Source
GiveWell, Iron Supplementation for School-Age Children, 2019 Source
GiveWell, Mass Distribution of Insecticide-Treated Nets (ITNs), 2023 Source
GiveWell, Research Strategy: Nutrition, 2024 Source
GiveWell, Seasonal Malaria Chemoprevention, 2024 Source
GiveWell, Sequence thinking vs. cluster thinking, 2016 Source
GiveWell, Vitamin A Supplementation, 2024 Source
GiveWell, What We Learned From Red Teaming Our Top Charities Source
Government Of Assam, Health & Family Welfare, National Health Mission, National Iron Plus Initiative (NIPI) Source (archive)
Government Of Assam, Health & Family Welfare, National Health Mission, Weekly Iron Folic Acid Supplementation (WIFS) Source (archive)
Hurrell 2015 Source
Hurrell et al. 2010 Source
India's National Family Health Survey (NFHS), Homepage Source (archive)
India's National Family Health Survey (NFHS), National Family Health Survey (NFHS-4)- 2014-15 Source (archive)
India's National Family Health Survey (NFHS), National Family Health Survey (NFHS-5)- 2019-20 Source (archive)
Institute for Health Metrics and Evaluation’s (IHME) Global Burden of Disease (GBD) study Source (archive)
Institute for Health Metrics and Evaluation’s (IHME) Global Burden of Disease Study 2021 (GBD 2021) Disability Weights Source (archive)
Iqbal and Ekmekcioglu 2019 Source
Iron and folic acid supplementation CEA, India [Feb 2023] Source
Karthikappallil and Atkinson 2023 Source
Low et al. 2013 Source
Mavridis and Salanti, 2013 Source
Mohammed et al. 2023 Source
Mwangi et al. 2021 Source
Nair, 2019 Source
Nestel et al. 2004 Source
Niemesh 2015 Source
Peña-Rosas et al. 2015 Source
Petry et al. 2016 Source
Prentice, Verhoef, and Cerami 2013 Source
Project Healthy Children, Homepage Source (archive)
Sachdev et al. 2019 Source
Times of India,Why is anaemia being dropped from National Family Health Survey?, 2023 Source (archive)
UNICEF, Maternal health Source (archive)
University of Cambridge, 'The Iron Lady' and the paradox of treating anaemia, 2017 Source (archive)
Wang et al. 2013 Source
WHO, Daily iron supplementation in children and adolescents 5–12 years of age in malaria-endemic areas, 2023 Source (archive)
WHO, Guideline on haemoglobin cutoffs to define anaemia in individuals and populations, 2024 Source (archive)
World Bank Group, Birth rate, crude (per 1,000 people) Source (archive)
World Bank Group, Prevalence of anemia among women of reproductive age (% of women ages 15-49) Source (archive)
  • 1

    From our 2024 blog post on our Nutrition team’s strategy: “Overall, we think iron and vitamin A deficiency are the most promising areas for grantmaking because of their high burden and because there are programs (supplementation and fortification) that offer tractable and cost-effective ways of addressing this burden.”

  • 2

    “Iron deficiency involves an insufficient supply of iron to the cells following depletion of the body’s reserves. Its main causes are a diet poor in absorbable iron, an increased requirement for iron (e.g. during pregnancy) not covered through the diet, a loss of iron due to parasitic infections, particularly hookworm, and other blood losses (Crompton 2002; INACG 2002a)." Peña-Rosas et al. 2015, Pg 7.

  • 3
    • “There are several strategies to reduce and/or treat iron deficiency and iron-deficiency anaemia: dietary modification and diversification that aims to increase the content and bioavailability of iron in the diet (FAO/CAB International 2011); preventive or intermittent iron supplementation through tablets, syrups or drops; blood transfusion, indicated only for very severe anaemia; biofortification through conventional plant breeding or genetic engineering that increases the iron content or its bioavailability in edible plants and vegetables; and fortification with iron compounds of staple foods (typically maize, soy and wheat flour) (WHO/FAO 2006)." Peña-Rosas et al. 2015, Pg 2.
    • "Iron therapy can be administered to children less than three years of age with IDA orally (ferrous sulphate, ferrous fumarate, sodium iron ethylenediaminetetra-acetate and ferrous gluconate), intravenously (iron dextran, iron sucrose and ferric carboxymaltose) or intramuscularly (iron dextran) (Goddard 2011)." Wang et al. 2013, Pg 5.

  • 4

    "There are over 79 countries with legislation to fortify wheat flour countries produced in industrial mills with vitamins and minerals. All the mandatory countries fortify wheat flour with at least iron and folic acid except Australia which does not include iron, and Congo, Venezuela, the United Kingdom, and the Philippines which do not include folic acid (FFI 2014)." Peña-Rosas et al. 2015, Pg 3.

  • 5

    See our list of grants made to Evidence Action’s IFA programs here.

  • 6

    See our list of grants made to other iron fortification programs here.

  • 7

    We expect to publish an updated report on iron supplementation and fortification intervention in early 2025.

  • 8
    • See, for example, our mostly recently published iron supplementation CEA, for Evidence Action’s IFA program in India, where we rely on Low et al. 2013 for the program’s effect.
    • From Low et al. 2013, “We identified 16 501 studies; of these, we evaluated 76 full-text papers and included 32 studies including 7089 children. Of the included studies, 31 were conducted in low- or middle-income settings. Iron supplementation improved global cognitive scores (standardized mean difference 0.50, 95% confidence interval [CI] 0.11 to 0.90, p = 0.01), intelligence quotient among anemic children (mean difference 4.55, 95% CI 0.16 to 8.94, p = 0.04) and measures of attention and concentration. Iron supplementation also improved age-adjusted height among all children and age-adjusted weight among anemic children. Iron supplementation reduced the risk of anemia by 50% and the risk of iron deficiency by 79%. Adherence in the trial settings was generally high.”

  • 9

    See Table 2 of Andersen et al 2023

  • 10

    (39% - 50%)/50% = -22%

  • 11

    From Andersen et al 2023:

    • Trim-and-fill effects are 2.2 g/L on hemoglobin and 4.1 ng/mL on ferritin: "For haematologic outcomes, studies with larger SEs tended to demonstrate more protective effect sizes (Egger’s test p<0.001) (online supplemental appendix 13). Attenuated but still statistically significant benefits were obtained during trim-and-fill sensitivity analyses for haemoglobin (2.2 g/L (1.2, 3.1)) and ferritin (4.1 ng/mL (1.6, 6.5))."
    • Main effects are 6.3 g/L on hemoglobin and 18.5 ng/L on ferritin: "In aggregate, oral iron supplementation versus placebo or control demonstrated clear benefits for haematologic indices (table 2). Haemoglobin levels rose by 6.3 g/L (95% CI 5.5, 7.1) along with serum ferritin increases of 18.5 ng/mL (16.1, 20.9). Iron supplementation reduced the prevalence of overall anaemia by 39% (33%, 45%), and even larger impacts were observed for ID (reduction of 70% (63%, 76%)) and IDA (reduction of 80% (69%, 87%)). Heterogeneity was observed between studies for these haematologic outcomes (I2 ranging from 80% to 100%), which was further explored using meta-regression (see below)."

    For hemoglobin, using the trim-and-fill method implies a (2.2 - 6.3) / 6.3 = -65% effect and for ferritin a (4.1 - 18.5) / 18.5 = -78%. Due to the uncertainty in this analysis, we extend the confidence interval implied by these decreases to -60% - -80%.

  • 12

    From our current intervention report on vitamin A supplementation: “We have received some expert feedback that the meta-analysis shows evidence of substantial publication bias, implying a smaller effect size on mortality than reported. Our internal validity adjustment intends to account for some possibility of publication bias, but we don’t use any statistical methods to estimate this and we have not systematically investigated this question. It’s possible that doing so would lead to a larger downward adjustment. (More)”

  • 13

    We discuss this issue further in our What We Learned From Red Teaming Our Top Charities page: “Failure to sense check raw data in burden calculations. We’ve found surprising differences in all-cause mortality estimates in different states in Nigeria between the Institute for Health Metrics and Evaluation’s (IHME) Global Burden of Disease data (our preferred source for mortality data) and other sources, which could impact which states we decide to fund New Incentives in, and surprisingly low malaria-specific mortality in some countries. We think we’ve taken this burden of disease data too much at face value, which makes us vulnerable to letting noisy data drive what grants we make.”

  • 14

    See this lightly vetted spreadsheet for our figures.

  • 15

    Two sources that we identified during our review brought up concerns with capillary measurements for anemia, the method used in the NFHS 4 and 5:

    • The Demographic and Health Surveys (DHS) website notes concerns with capillary measurement including, “The unpredictable variability in single-drop capillary hemoglobin measurements can lead to inaccurate anemia prevalence estimates and trends.”

    An article from the Times of India also notes concerns stemming from the increase in anemia prevalence in the NFHS 5 stating, “In the last survey, NFHS-5 (2019-21), the prevalence of anaemia showed a huge jump among almost all sections with the most dramatic increase (from 59% to 67%) among children aged 6 months to five years. This was an embarrassment to the central government, which had launched Anaemia Mukt Bharat in 2018 and had been carrying out an iron and folic acid supplementation programme for several decades. Several experts said that the prevalence was ‘inexplicably’ high and blamed it on the method of using capillary samples and said that the high estimates were being used by many, including international organisations, to push fortification of food.”

  • 16

    The decision to drop anemia testing from the NFHS and move monitoring to the DABS is discussed in this Times of India article.

  • 17

    A 2024 paper by Givens, Anitha, and Giromini discusses the fact that the DABS will use the venous method, “The current Diet and Biomarkers Survey in India (DABS-1) will use venous blood to measure [hemoglobin]“

  • 18

    For example, the Indian government has been implementing several interventions aimed at reducing anemia under its Anemia Mukh Bharat (AMB) program such as:

    • “Prophylactic Iron and Folic Acid supplementation
    • Deworming
    • Intensified year-round Behaviour Change Communication Campaign focusing on four key behaviors:
      • Improving compliance to Iron Folic Acid supplementation and deworming
      • Appropriate infant and young child feeding practices
      • Increase in intake of iron-rich food through diet diversity/quantity/frequency and/or fortified foods with focus on harnessing locally available resources
      • Ensuring delayed cord clamping after delivery (by 3 minutes) in health facilities
    • Testing and treatment of anemia, using digital methods and point of care treatment, with special focus on pregnant women and school-going adolescents
    • Mandatory provision of Iron and Folic Acid fortified foods in government-funded public health programs

    Intensifying awareness, screening, and treatment of non-nutritional causes of anemia in endemic pockets, with special focus on malaria, haemoglobinopathies, and fluorosis”

  • 19

    Dr. Engle-Stone commented that while data quality issues cannot be ruled out, low consumption of animal-sourced foods that are rich in bioavailable iron is a major risk factor for iron deficiency and anemia which should be considered when comparing the anemia burdens in India and sub-Saharan Africa, and that the prevalence of anemia in India may be consistent with dietary patterns in some regions, though possibilities such as genetic contributors should also be considered.

  • 20

    See this post we made on the Effective Altruism Forum responding to criticism that we received from Akash Kulgod during the 2022 Change Our Mind contest. Akash’s primary criticism is based on analysis by Sachdev et al. 2019 which argues that the WHO definitions for anemia, developed based on predominantly white populations, overestimate anemia rates in India due to physiological and body-composition factors. Based on our own research and discussions with outside experts, we have decided against adjusting for country-specific hemoglobin cutoffs at this time. However, we will continue to investigate external validity concerns for iron supplementation in India and elsewhere.

  • 21

    We have to make adjustments to compare the impact of iron supplementation to interventions such as seasonal malaria chemoprevention (SMC) due to the fact that we believe the main impact of iron supplementation programs is reduced disability from anemia (~60% of benefits in several locations in India in our most recently published CEA) compared to SMC where we model ~84% of benefits coming from reduced mortality (80% from reduced mortality for children under 5 and 4% from reduced mortality for other ages).

  • 22

    The definition of disability weights from GBD 2021, “Disability weights, which represent the magnitude of health loss associated with specific health outcomes, are used to calculate years lived with disability (YLD) for these outcomes in a given population. The weights are measured on a scale from 0 to 1, where 0 equals a state of full health and 1 equals death.”
    While we do not rely solely on the IHME GBD for disability weights in our CEAs, we commonly rely on their estimates.

  • 23
    • We use “units of value” as a way to compare programs with different impacts, such as reduction in morbidity, increases in consumption, and mortality averted. For more on why and how we use units of value to compare programs, see this post on GiveWell’s moral weights.
    • For anemia, the GBD 2021 disability weights are 0.15 for severe anemia, 0.05 for moderate anemia, and 0.004 for mild anemia. We currently assign 2.3 units of value per disability-adjusted life year, as discussed in our post on moral weights.For anemia, the burden is X for severe anemia, X for moderate anemia, X for mild anemia. We discuss these calculations and our uncertainties in the ‘Key questions for further investigation’ section of our iron supplementation intervention report.

  • 24

    See this section of our previous report on the cost-effectiveness of iron supplementation for further discussion.

  • 25

    We discuss this alternative way to calculate the benefits in YLDs of iron supplementation in the ‘Key questions for further investigation’ section of our iron supplementation intervention report.

  • 26

    At the time of writing, our latest CEA was unpublished and subject to change. Other adjustments in the CEA included:

    • Variation in compound effectiveness - external validity adjustment
    • Variation in amount of fortified food consumed - external validity adjustment
    • Variation in fortificant concentration - external validity adjustment
    • Whether the effect of iron intake on anemia burden is linear - we assume it is
    • Baseline population iron-deficiency anemia - external validity adjustment
    • The effect of inflammation on iron absorption - external validity adjustment
    • Population diet - especially the amount of fortified food consumed and inhibitors - external validity adjustment
    • Quality of fortification process (is the food adequately fortified, is the fortificant good quality, is it still effective after the food has been stored) - external validity adjustment
    • Amount of food consumed (wastage and intra-household iron consumption) - external validity adjustment

  • 27

    At the time of writing, our latest CEA was unpublished and subject to change. Our primary uncertainties regarding our adjustments in the CEA included:

    • Variation in compound effectiveness -
      • We are pretty uncertain about some of the comparisons between different iron-containing compounds . In an unpublished exchange with GiveWell, Dr. Richard Hurrel mentioned one of the studies we include in our analysis might be using “reduced” iron, which is not recommended by WHO. See also Nestel et al. 2004 which finds, “Neither electrolytic nor reduced iron had an effect on Hb concentration among preschoolers.”
    • Variation in amount of fortified food consumed
      • We are using estimates from Fortify Health, a nonprofit specializing in iron fortification in India that GiveWell has previously supported, but those don’t account for inter-state variation, and are based on a rough best guess of food wastage
    • Variation in fortificant concentration
      • We assume this has a linear effect on anemia burden, but in an unpublished exchange with GiveWell, Dr. Chris Somerville suggested that might not be the case. He stated that a study in Thailand showed that doubling iron concentration in fortified wheat had an outsized effect and that we should not expect the effect to be linear because iron has to saturate a chelator in wheat (phytic acid) to become bioavailable.
    • Whether the effect of iron intake on anemia burden is linear.
      • We assume it is. However, comments from Dr. Richard Hurrell, in an unpublished exchange with GiveWell suggest that might not be the case. Dr. Hurrell suggested measuring the effect of iron fortification by looking only at studies providing at least the minimum daily amount recommended by the WHO. This is because his own review of the literature (Hurrell et al. 2010) found that studies supplementing over the minimum recommendation found moderate to good iron efficacy while those supplementing less than the recommendation had no effect on iron status.
      • This concern might be compounded by uncertainty about intra-household distribution of fortified food, where we’re concerned that a disproportionate amount of fortified foods will be consumed by adult men, who do not account for much of the anemia burden, reducing cost-effectiveness. There are reasons to believe the impact will be attenuated because most evidence is already collected from families in this setting. We have not deeply researched this topic..
    • The effect of inflammation on iron absorption
      • We have researched the issue of inflammation decreasing an individual’s ability to absorb iron, including in this article Hurrell 2015. However, we are very uncertain about the magnitude of the effect and how it varies across contexts.
    • Population diet - especially the amount of fortified food consumed and inhibitors
      • There is some discussion in the literature of foods and beverages that can inhibit iron absorption, such as this article by Fan (2016) discussing iron deficiency due to green tea consumption.
      • We believe this could cause a significant negative adjustment l, since most trials target children (who are unlikely to drink tea), while most of the burden is in women (who are more likely to drink tea)

  • 28

    In unpublished exchanges with GiveWell, Dr. Richard Hurrell and Dr. Sant-Rayn Pasricha expressed concern the studies included in Field et al may not be generalisable to the interventions of interest.

    • Dr. Hurrell referenced the small number of studies and the fact that several studies did not follow the WHO recommendations for wheat flour fortifications, either in relation to fortification levels or type of iron compound used.
    • Dr. Pasricha noted differences in design and population groups between reviewed studies that limit generalisability of this evidence to the interventions and populations of interest.

  • 29

    We discussed other potential research questions to address on this page on a grant made to Fortify Health in 2021.

  • 30

    Our list of grants made to Evidence Actions IFA programs, totalling ~14 million as of January 2025, can be seen here.

  • 31

    Our pages for grants made to Fortify Health, as of January 2025, can be seen here, here, and here.

  • 32

    Data from the World Health Organization, Global Health Observatory Data Repository in 2019 indicate that in Nigeria (55.1%), Mali (59%), and Senegal (52.7%) the prevalence of anemia in women of reproductive age (15-49) is over 50%. For reference, the figure in India is 53%.

  • 33

    For example, Prentice, Verhoef, and Cerami 2013, Pg 914: “To minimize the risk of infection, the physiology of these children has adapted to actively exclude iron much of the time, which is validated by the fact that rare genetic variants in the hepcidin-ferroportin pathway lead to iron overload even in people on low-iron diets. Second, because so much evolutionary experience has been distilled into extensive genomic investment in mechanisms mediating the host-pathogen competition for iron, it may be hazardous to intentionally override such processes by administration of excess iron… If iron deficiency is protective, reversing such deficiency would be expected to enhance malaria susceptibility as previously demonstrated.”.

  • 34

    From the WHO page on daily iron supplementation, “Daily iron supplementation is recommended as a public health intervention in school-age children aged 60 months (i.e. 5 years old) and older, living in settings where the prevalence of anaemia in infants and young children is 40% or higher, for preventing iron deficiency and anaemia.
    In malaria–endemic areas, the provision of iron supplementation in infants and children should be done in conjunction with public health measures to prevent, diagnose and treat malaria.”

  • 35
    • From Karthikappallil and Atkinson 2023 p 8 “Current evidence suggests that universal iron supplementation is safe in malaria-endemic areas when accompanied by appropriate malaria prevention and diagnosis. In low-resource settings where malaria control is inadequate, there is some evidence that iron supplementation may increase the risk of malaria and gram-negative bacterial infection in children. On the balance of risks, it is likely that the benefits of universal iron supplementation in reducing population levels of iron deficiency and anaemia outweigh any potential increases in the risk of infection”
    • From Mwangi et al. 2021 p e741 “In our view, administering malaria preventive measures to individual recipients of iron interventions is ethically required, in place of public malaria control measures that might not reach these individuals.”

  • 36

    Dr. Sant-Rayn Pasricha, in an unpublished exchange with GiveWell in 2023 regarding a previous version of our CEA for Fortify Health, noted that while iron fortification could theoretically increase malaria risk, this intervention is likely too gradual and modest to trigger the rapid red blood cell production (reticulocytosis) that enables increased parasite invasion, as described in Clark et al 2014.

  • 37

    Our understanding, informed by conversations with experts and studies such as Hurrell 2015, is that iron absorption may be inhibited by the body's inflammatory response to infections. This may reduce the effectiveness of iron supplementation and fortification programs in areas with high infection prevalence. We plan to adjust for this concern using an "inflammation-exposure index" developed by Petry et al. 2016. We estimate that this was a larger concern in trial settings than for a hypothetical program in India today due to higher levels of infection in the past.

  • 38

    A 2017 article from the University of Cambridge discussed the potential advantages of a slow-releasing iron supplement that would alleviate iron-deficiency while making iron less available to pathogens in the body. In a 2023 phase-II trial, a slow-release form of iron, iron hydroxide adipate tartrate (IHAT), “showed sufficient non-inferiority compared to standard-of-care FeSO4 [a standard formulation of iron].”

  • 39

    Our best guess is that averting childhood illness, such as treatment of anemia with iron, results in small income gains in adulthood. We refer to these long-term income increases as “development effects” and in other programs, such as distributions of insecticide-treated bed nets (ITNs) or seasonal malaria chemoprevention (SMC), they account for roughly 15-40% of the benefits of those programs. For a longer discussion of long-term income increases and how we model them, see this section of our intervention report on distribution of ITNs.

  • 40

    As a result of our red-teaming we applied a -70% downwards replicability adjustment to the effect size of Niemesh 2015 in our forthcoming CEA. This figure accounted for the following:

    • A 40% (i.e., -60%) subjective validity adjustment to account for the evidence being non-randomized and outdated.
    • A 75% (i.e. -25%) adjustment to account for gender disparities in income effects. Niemesh 2015 only finds statistically significant effects (at the 1% level) when the analysis is restricted to men. We think it is plausible that averting malaria cases today would result in economic gains for women that Niemesh 2015 did not find (e.g., because of women’s higher labor force participation today than when the campaigns were conducted), but there may still be reasons why women would not see economic gains as large as men. To account for this, we apply a further adjustment of 75% to our estimate based on our guess that the effect for women might be around half as large as for men, and women comprise ~50% of the population.
    • Combining these two components, the implied adjustment factor is 40% x 75% = 30%, so our final replicability adjustment is a 70% reduction in the Niemesh effect size.

  • 41

    “Cluster thinking … involves approaching a decision from multiple perspectives (which might also be called “mental models”), observing which decision would be implied by each perspective, and weighing the perspectives in order to arrive at a final decision.” See our post for more

  • 42

    These checks include:

    • How biologically plausible is the mechanism?
    • Are there other ways to get a quantitative estimate? In a previous intervention report, we’d relied on evidence for the effect of IQ on income and the effect of iron on IQ. How does that approach compare? This was also recommended by Dr. Katherine Adams, one of the expert reviewers.
    • Can we make more straightforward comparisons between malaria programs and iron programs? For malaria, we estimate ~0.6% gain in income per malaria case. Can we translate iron development effects into per episode of anemia?
    • Can we compare to deworming programs as well? This is another program where we’ve spent a long time estimating development effects.
    • At a macro level, how much does increased uptake of iron fortification or supplementation explain changes in GDP per capita over time? If we were to find iron programs account for ~20% or more of increases in GDP per capita, that might make us skeptical that we’ve overestimated the effects.
    • How do our estimates compare to others in the literature? Dr. Katherine Adams, one of the expert reviewers we commissioned, compared our estimates to others used in the literature and found "In the context of other estimates in the literature, the estimated economic benefits used in this CEA seem reasonable and perhaps even a bit conservative." We think this type of check is useful and potentially worth including in our analysis.

  • 43

    From our intervention report on distributing ITNs: “To create our best guess on the effect of reductions in malaria on later-life income, we rely on evidence from two historical quasi-experiments, Bleakley 2010 and Cutler et al. 2010, that measure the effect of eradicating malaria on later-life income in the Americas and India. Based on this evidence, we estimate that each case of childhood malaria averted increases adult income by 0.6%, and that these benefits persist over 40 years of an individual’s working life.”
    “Intuitively, we think it is plausible that long-term income increases constitute an important share of the benefits of ITNs because:
    It seems plausible that averting malaria during a sensitive period of childhood development could improve long-run outcomes. Though we haven't deeply investigated potential mechanisms and the chain of causation, we know that malaria can cause fever, anemia, enlarged liver or spleen, kidney injury, and severe neurological symptoms associated with cerebral malaria, among other morbidities.247 Avoiding these symptoms could benefit physical and/or cognitive development. (More)
    Several additional quasi-experimental studies beyond Bleakley 2010 and Cutler et al. 2010 support direct effects of eradicating malaria on later-life income. (More)
    Our bottom-line long-term income effect size is broadly consistent with our findings for other programs for which we analyze long-run income effects, as well as macroeconomic trends in the relationship between malaria burden and economic growth. (More)
    While we value a small increase in adult income much less than averting a death, nets avert many more childhood malaria cases than deaths, adding up to a large total benefit.”

  • 44

    For example, see this section of our CEA on distribution of ITNs.

  • 45

    While we think that household-level food distribution may reduce the effectiveness of iron fortification programs, this effect might be smaller than expected since most evidence from iron fortification programs measures benefits for children, meaning differential consumption is already partially accounted for in the evidence. We are most concerned about decreased effectiveness of fortification for women of reproductive age, who have higher iron needs than children but may not consume sufficient amounts of fortified foods to meet those needs. This is partially offset by adult men, who consume more food than children but have lower iron needs. Taken together, our best guess is that this results in a small decrease in cost effectiveness.

  • 46

    See Table 2 of Andersen et al 2023

  • 47

    From Low et al. 2013, “We identified 16 501 studies; of these, we evaluated 76 full-text papers and included 32 studies including 7089 children. Of the included studies, 31 were conducted in low- or middle-income settings. Iron supplementation improved global cognitive scores (standardized mean difference 0.50, 95% confidence interval [CI] 0.11 to 0.90, p = 0.01), intelligence quotient among anemic children (mean difference 4.55, 95% CI 0.16 to 8.94, p = 0.04) and measures of attention and concentration. Iron supplementation also improved age-adjusted height among all children and age-adjusted weight among anemic children. Iron supplementation reduced the risk of anemia by 50% and the risk of iron deficiency by 79%. Adherence in the trial settings was generally high.”

  • 48

    These figures reflect our iron supplementation and fortification CEA at the time that it was reviewed by Dr. Adams. It may not reflect updated thinking or changes we’ve made since her review.

  • 49

    The economic benefits that Dr. Adams references here include both development benefits and anticipated increases in income and work capacity for adults benefiting from iron fortification.

  • 50

    One of the main benefits of folic acid fortification or supplementation is the prevention of neural tube birth defects (which is what our ~1x cash estimate is based on). Dr. Engle-Stone noted that this benefit only occurs for women who receive additional folic acid at or before the time of conception, as folic acid received during pregnancy may not come soon enough to prevent neural tube defects, making fortification or preventative supplementation more likely to be beneficial for the purpose of preventing neural tube defects.

  • 51

    To date, we have not identified RCT evidence that iron supplementation impacts maternal mortality. We are aware of observational evidence, however we have chosen not to review it due to the limitations of observational evidence in establishing causality. We may update our beliefs if we find RCT evidence, particularly strong observational evidence, or mechanistic evidence of the efficacy of iron supplementation in reducing maternal mortality.

  • 52

    See here a rough, unvetted calculation of the impact that including benefits from maternal mortality would have on our iron fortification CEA. Given the low impact on estimated cost-effectiveness, we have not prioritized researching this further.

  • 53

    Source: United Nations Population Division. World Population Prospects: 2022 Revision, estimates 16 births per 1,000 people in India in 2022

  • 54

    From the WHO page on maternal mortality, “Maternal Mortality Ratio(MMR) of India for the period 2018-20, as per the latest report of the national Sample Registration system (SRS) data is 97/100,000 live births, declining by 33 points, from 130/ 100,000 live births in 2014-16.”

  • 55

    This is our estimate based on:

    • A 0.97 risk ratio for maternal mortality in Iqbal and Ekmekcioglu 2019 (implying a 3% reduction in maternal mortality)
    • Our estimate that fortification is ~43% as effective as supplementation, calculated in our forthcoming CEA
    • 3% * 0.43 ~= 1.29%

  • 56
    • We assume that benefits to infants would be incorporated into our development effect estimates because iron deficiency at this age may have long-run effects on child development.
    • Dr. Horton noted that iron deficiency in pregnancy may lead to iron deficiency in infants for up to six months, and that this could have long-run effects on development. She also noted that it’s generally not possible to supplement iron in infants under six months as there is some evidence that it causes harm.

  • 57

    Our understanding, informed by conversations with experts and studies such as Hurrell 2015, is that iron absorption may be inhibited by the body's inflammatory response to infections. This may reduce the effectiveness of iron supplementation and fortification programs in areas with high infection prevalence. We plan to adjust for this concern using an "inflammation-exposure index" developed by Petry et al. 2016. We estimate that this was a larger concern in trial settings than for a hypothetical program in India today due to higher levels of infection in the past.