EA - Responding to recent critiques of iron fortification in India by e19brendan

Link to original articleWelcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Responding to recent critiques of iron fortification in India, published by e19brendan on October 12, 2022 on The Effective Altruism Forum. Two recent EA Forum posts offered critiques on iron fortification in India and particularly the work of Fortify Health, given its relationship with EA: Targeted Treatment of Anemia in Adolescents in India as a Cause Area and Cost-effectiveness of iron fortification in India is lower than GiveWell's estimates, both authored by Akash Kulgod. We heartily welcome the inquiry and critique, and Fortify Health will remain open minded in pursuit of understanding the true impact of its work and in adjusting course accordingly. I am one of the co-founders of Fortify Health, and although I have since stepped back from day-to-day operations, I offered to address the concerns raised so that the active core team can continue to focus on their implementation and partnership work. In this post, I intend to provide initial responses to key points made in Kulgod’s articles: Is iron fortification expected to increase diabetes prevalence in India? Short answer: This should not be inferred from available evidence. Is the proportion of anemia in India attributable to iron deficiency lower than we thought? Short answer: At first glance, more recent estimates of iron deficiency among children cited are on average roughly similar to those incorporated in GiveWell’s 2019 CEA, and state and age specific prevalence could be incorporated into future models. Givewell’s 2019 model largely relies on iron deficiency prevalence rather than iron deficiency anemia prevalence, but I’m uncertain as to whether or for what outcomes use of one or both parameters would lead to the most accurate model. Should hemoglobin cutoff values be changed, and how does this affect the expected impact of fortification efforts? Short answer: The distribution of hemoglobin levels among the Indian sample cited does not support the inference that outcomes associated with anemia are equivalent between populations at different threshold hemoglobin levels. The new information presented by the cited study does not weaken the expected impact of fortification. Is a targeted intervention preferable to widespread fortification? Short answer: Targeted screening and treatment should be available as part of comprehensive primary care and possibly school-based programs, but does not preclude rapidly scaling up fortification efforts. Further exploration of this intervention would be worthwhile. Is micronutrient fortification ethical? Short answer: Extending the benefits of fortification is one part of an ethical imperative for health equity, and absence of fortification does not provide greater autonomy. I recognize that this post only scratches the surface of complex issues, and it does not provide a comprehensive review of all available arguments and evidence that may be relevant. Its scope is substantially more limited, merely contributing to an evolving conversation started by Kulgod’s posts. I’m eager to get feedback from Kulgod and other readers, and hope we can collaboratively advance our own understanding and the EA community’s understanding of fortification efforts. I. Is 10mg/day per capita iron fortification expected to raise prevalence of diabetes by 2-14% as apparently suggested by Ghosh et al. (preprint 2021)? Initial reaction: the cited study’s conclusion does not seem to be supported by its results. Claims of correlation between high fasting blood sugar with discrete elevation in serum ferritin are supported by only a small effect size in a subgroup analysis of the highest wealth quintile, which is sharply discontinuous from the next wealth quintile. Furthermore, their model of marginal increase in fasting blood sugar by serum ferritin levels demonstrates minimal effect size, even in ...