Developmental Cognitive Neuroepidemiology
In Superintelligence Bostrom writes:
'Our individual cognitive capacities can be strengthened in various ways, including by such traditional methods as education and training. Neurological development can be promoted by low-tech interventions such as optimizing maternal and infant nutrition, removing lead and other neurotoxic pollutants from the environment, eradicating parasites, ensuring adequate sleep and exercise, and preventing diseases that affect the brain. [...] 'According to the World Health Organization in 2007, nearly 2 billion individuals have insufficient iodine intake (The Lancet 2008). Severe iodine deficiency hinders neurological development and leads to cretinism, which involves an average loss of about 12.5 IQ points (Qian et al. 2005). The condition can be easily and inexpensively prevented though salt fortification (Horton et al. 2008).'[1]
Here, I propose launching a new philanthropic cause area and academic field: 'Developmental Cognitive Neuroepidemiology' (DCN).
DCN combines the following two fields:
And so, in essence, DCN studies what causes brain disease—widely construed to also include suboptimal brain development and cognitive function—in children and young adults (eventually also across the whole life span), and how many are affected and how badly.
The field's aim might be to elevate optimal neurodevelopment to a similar rank on the policy agenda as physical development. I motivate this by examining the importance and scale of the problem. I focus on reviewing some previous work trying to quantify the scale of what might cause suboptimal cognitive development. I start by giving some concrete examples
Sometimes IQ loss during development is very cost-effective to prevent, e.g. preventing iodine deficiency [2],[3] (supplementation or fortification of staple foods 100 additional years per $100 spent on supplements, and an increase in ~1k-2k IQ points (7-10 IQ points for ~185 children). This ~100 IQ points per ~$1 spent. Each point increase in IQ test scores raises income by ~$100-$1000 per year after holding a variety of factors constant and so benefit-cost ratios of select interventions might be very high. Increasing IQ from current mean (100) to the top 2 percent of society (130) increases wages by ~$10k,[4] and so raising the IQ of 100M people would to such an extent would raise world GDP by at least $1T.
Intelligence and healthy cognitive neurodevelopment is an important prerequisite for a flourishing life and suboptimal neurodevelopment might also have high social costs.
For instance, Caplan argues that education raises support for capitalism, free markets, globalization, civil liberties and tolerance, and reduces racism and sexism. Correcting for intelligence cuts education's impact by ~1/3.[5]
As a result of suboptimal child neurodevelopment we likely lessen the effectiveness of education for hundreds of millions or even billions—I review the evidence below. In contrast, early childhood education has mixed evidence.[6] However, we spend a lot of money on early childhood education, but very little on cognitive development. This might be because education is much less controversial and that the idea that intelligence is fixed to a substantial extent in early childhood or in utero is uncomfortable and controversial (as it might be conflated with genetic differences in intelligence between cultures).
Suboptimal child neurodevelopment might have drastic general equilibrium effects as well. People with low IQs might not be able to understand and answer conditional hypothetical questions [7] (e.g. ''How would you have felt yesterday evening if you hadn't eaten breakfast or lunch?' 'What do you mean? I did eat breakfast and lunch.' 'Yes, but if you had not, how would you have felt?' 'Why are you saying that I didn't eat breakfast? I just told you that did.' 'Imagine that you hadn't eaten it, though. How would you have felt?' 'I don't understand the question.').
Consider the literature on the 'fetal origins hypothesis' and child development, where early environment affects cognitive development and later life outcomes.[8] Some of the effect sizes are downright incredible (not using this term lightly) and its implications might be 'big, if true' and so deserve further study.
For example, salt iodization is cheap and might improve (population-level) cognitive development and IQ. [9],[10] In India, children in iodine-poor districts who were exposed to a decrease in iodine-fortified salt in early life that was almost random, are '1-8 percentage points less likely to be able to do any math or reading'.[11]
In China, females in the rural areas with a 1SD reduction in goiter rates (a proxy for iodine deficiency) have a ~15% increase in cognitive ability. Females who benefit from the new salt also obtain 0.5 additional years of schooling and have a higher educational attainment.[12]
There are other natural experiments that show an even more dramatic effect of salt iodization in the past,[13] up to an increase of 15 IQ points. In contrast, while there are some RCTs that suggest that improving iodine deficiency in children improves cognitive functions (summarized in a meta-analysis[14]), the evidence is not as clear-cut as sample sizes in RCTs are much lower, participants were not followed over their whole development and it is deemed unethical to deprive people of micronutrients they need.
Even in 2019, ~5m newborns will be affected by full-blown iodine deficiency disorders.[15] However this underestimates the economic deficit resulting from suboptimal iodine intake.
Iodine deficiency affects 780M people globally. Even in rich countries there might be substantial deficiency.[16]
Figure: Iodine nutrition based on the median urinary iodine concentration, by country. Figure adapted from Zimmermann et al. [17]
Other micronutrient deficiencies are also still widespread:
Global Distribution and Disease Burden Related to Micronutrient Deficiencies in 2011[18]
Iron deficiency might affect school enrollment.[19]
IQ might also be affected by acute and transient micronutrient deficiencies, and so multivitamin & iodine supplementation might raise IQ.[20]
There are many other problems that can improve neurodevelopment.
Lead poisoning is harming the cognitive ability and educational opportunities of hundreds of millions of children: Lead poisoning is affecting children 800M children globally—that's around 1 in 3—have levels of lead in their blood at or above the level that requires intervention. Most of these children live in Africa and Asia.[21]
Leaded fuel reduced the IQ of everyone born before 1990 by ~4.25%. Millennials are the first to be born with unleaded gas.[22]
Environmental lead levels from leaded gasoline are still around in cities today, and cause continued neurotoxicity.[23]
Figure. Children's average blood lead levels by country (μg/dL) source
The Toxic Truth, UNICEF: 'Lead is more dangerous for children than it is for adults. Children absorb it more readily and they drink more water relative to their body weight than adults do. It particularly affects children's developing brains and has long-term effects, including deficits in cognitive ability, lower educational attainment, behavioral disorders, increased tendencies for violent crime, and reduced lifetime earnings. A pooled analysis of lead exposure and intellectual function found that children with higher (30 μg/dL) blood lead levels had IQ levels of almost seven points less than children with lower (2.4 μg/dL) blood lead levels. And evidence from the US found that high blood lead levels before the age of six are strongly associated with poor academic achievement in primary school grades.'
Since lead poisoning affects children in poor countries most, and since lead poisoning is detrimental to children's educational achievement, perhaps the education sector should invest more in research to understand the long-term educational effects on exposure to lead and in cross-sector efforts to prevent and control exposure to lead.
National average IQ and parasite stress correlates strongly r ≈-0.8 (p<0.0001).[24]
Infectious disease exposure is associated with neurocognitive function in humans and one study finds that an index aggregating exposure to several infectious diseases like toxoplasmosis was associated with neurocognitive function.[25]
For instance, there's some disagreement around how much (long) Covid affects cognition, but generally, the cognitive decline following acute viral infections and strong evidence for smell distortions post Covid, suggest that Covid affects the brain and thus might plausibly affect cognition. One study finds 0.04 SD IQ difference between people without respiratory symptoms (on the order of ~1 IQ point). Given that there are many millions of infections the IQ loss might be substantial.
Other reviews on this topic find the following interventions:
Predictors of IQ
Predictors of IQ | Effect Size |
Genetic factors | |
Genes from twin studies | R2 = 50% |
SNP heritability (from GCTA studies) | R2 = 20% |
Polygenic Risk Score | R2 = 10% |
Genetic syndrome (here: Fragile X) | β* ~ −30 pts |
Prenatal exposures | |
Maternal tobacco consumption during pregnancy (>1 pack/d) | β* = −2 pts. in univariate analysis |
but β = 0 pts. in multivariate analysis | |
Maternal alcohol consumption during pregnancy (heavy drinking) | β = −8 pts |
Drug exposure during pregnancy (e.g. high-dose valproate (>800 mg /d) | β = −10 pts |
Birth factors | |
Preterm birth (25-37 GA) | β = −10 pts |
Small for gestational age | β = −4 pts |
Apgar scores <7 at 1 and 5 min | β = −1.2 pts |
Parental and social factors | |
Parental education | β = +0.7 pts./year of parental education |
Birth rank (first vs. second born) | β* = +3 pts |
Breastfeeding (yes vs. no) | β* = +3.4 pts. in univariate analysis |
and β = +2.62 when controlling for maternal IQ | |
Parent-child interaction | β = +0.8 pts. / HOME score points |
Malnutrition | β = −3.53 pts. for early-onset persistent stunting in univariate analyses |
and β = −2.10 pts. when controlling for other risks factors | |
Screen exposure | β = −0.5/−0.7 pts. / daily hour of exposure for the within-subject association |
E.g. 'Study after study shows higher intelligence (2-4 points), more favorable personality traits (for example, 33% less extreme shyness) and greater height (0.6 cm) in winter/spring babies (1, 2, 3, 4).'
'Below are all the interventions considered, my probability that the effect is real, and the number of IQ points gained. 'Probability it works' is my totally made up wild guess as to how likely it is to be a real causal effect. 'Probability it applies' is my totally made up wild guess as to how likely it is an average reader can benefit from it. For example, licorice has a low probability of application, because you probably weren't planning to eat a ton of licorice anyway. Some things have partial probabilities of application; you may get a little benefit from changing your location if you had been planning on living in California, but not the full benefit you would get if you had been planning on living in Mississippi.
The best case scenario — in which every single study I cited was the gospel truth, someone reading this had been planning to do everything wrong, and now they are planning to do everything right — nets us an extra 90 IQ points, meaning that it would be enough to bring an average IQ 100 child (which of course the child who did everything wrong would not have been) to an IQ 190 child, which is around the level historians speculate Isaac Netwon could've been at.'
From: The Health-Education Link | Hauke Hillebrandt and Hayden Wilkinson
Intervention type | Testing location | Effect on schooling | Source |
Preventing Iodine deficiency (supplementation or fortification of staple foods) | Tanzania | 62.50-109.80 additional years per US$100 spent on supplements (does not include cost of delivery), plus an increase in IQ by 6.9-10.2 IQ points for approximately 179-196 children | Field, Robles, & Torero, 2009.[27] Bougma, Aboud,, Harding, Marquis. 2013.[28] |
Information on returns of education (to parents) | Madagascar | 43.48 additional years per US$100 spent | Nguyen, 2008.[29] |
Measles vaccination | Ethiopia | Approximately 14 additional years per US$100 spent | Anekwe, et al. 2015.[30] Driessen, et al. 2015.[31] |
Deworming through primary schools | Kenya | 12.50-34.25 additional years per US$100 spent | Miguel & Kremer, 2004.[32] Hicks, Kremer, Miguel, 2014.[33] |
National-level malaria control | Sub-Saharan Africa | 7.58 additional years per US$100 spent | Kuecken, Thuilliez, & Valfort, 2015.[34] |
Combination iron fortification and deworming in preschools | India | 2.73 additional years per US$100 spent | Bobonis, Miguel, & Puri-Sharma, 2006.[35] |
Building village-based schools | Afghanistan | 1.51 additional years per US$100 spent | Burde & Linden, 2013.[36] |
Free primary school uniforms | Kenya | 0.71 additional years per US$100 spent | Evans, Kremer, & Ngatia, 2009.[37] |
Information on returns of education (to male students) | Dominican Republic | 0.24 additional years per US$100 spent | Jensen, 2010.[38] |
Merit scholarships for girls | Kenya | 0.16 additional years per US$100 spent | Kremer, Miguel, & Thornton, 2009.[39] Friedman, et al., 2011.[40] |
Girls' conditional cash transfers (CCTs) (minimum amount) | Malawi | 0.09 additional years per US$100 spent | Baird, McIntosh, & Özler, 2011.[41] |
Girls' CCTs (average amount) | Malawi | 0.07 additional years per US$100 spent | Baird, McIntosh, & Özler, 2011.[42] |
Girls' unconditional cash transfers (UCTs) (average amount) | Malawi | 0.02 additional years per US$100 spent | Baird, McIntosh, & Özler, 2011.[43] |
Progresa CCT for primary school attendance | Mexico | 0.01 additional years per US$100 spent | Schultz, 2000.[44] |
Camera monitoring of teachers' attendance | India | No significant change in years of schooling, but 21 percentage point decrease in teacher absenteeism (unknown cost per school) | Duflo, Hanna, & Ryan, 2012.[45] |
Computer assisted learning curriculum | India | No significant change in years of schooling, but would increase mathematics scores of 6.6 students by 0.35 standard deviations | Banerjee, et al,. 2007.[46] |
Remedial tutoring by community volunteers | India | No significant change in years of schooling | Banerjee, et al., 2007.[47] |
Menstrual cups for teenage girls | Nepal | No significant change | Oster & Thornton, 2011.[48] Oster & Thornton, 2012.[49] |
Child sponsorship | Uganda, Guatemala, Philippines, India, Kenya, Bolivia | 0.04 additional years per US$100 spent | Wydick, Glewwe, & Rutledge, 2010.[50] |
Treating iron deficiency (fortification)[51] | China | Raised the test scores of anemic students by about 0.2 standard deviations (exact cost per student unknown but likely very low) | Luo, et al., 2012.[52] |
Figure: Impact on years of schooling per US$100 spent for various interventions, based on available evidence. Partly adapted from J-PAL.[53] Note that some of these estimates are based on only one paper and should be taken as indications rather than exact measures. However, though inexact, there are orders of magnitude differences between the different interventions
Outcome | Strength of human evidence | Probability of causation, % | |
Prenatal PBDEs | IQ loss and intellectual disability | Moderate to high | 70–100% |
Prenatal organophosphate pesticides | IQ loss and intellectual disability | Moderate to high | 70–100% |
Multiple prenatal exposures | Attention-deficit disorder | Low to moderate | 20–69% |
Multiple prenatal exposures | Autism spectrum disorder | Low | 20–39% |
Table. Exposure–outcome associations with a probability of evidence for causation identified up to 2015
There might be stark cross-cultural differences in IQ due to suboptimal neurodevelopment caused by poverty that could be too politically touchy to discuss, despite it being intuitive that, say, early childhood nutrition might affect intelligence on a population level. This makes the cause quite neglected. However, crucially, DCN does not look at genetic differences in cognition—rather it only examines environmental and thus malleable causes.
On the flip side, childhood development seems to be a pretty popular area of philanthropy and so there is already a lot of money going into it.
Given that this literature around 'fetal origins hypothesis' has been around for so long and people still don't talk about it much, maybe the best thing a philanthropist could do might be to go a bit 'for the throat of the problem' and go more meta by:
Shovel ready grantees might be:
The Gates foundation has funded two interesting projects on epidemiology- that could inspire analogous projects in DCN:
Global Burden of Disease study visualization tool
However, while the Global Burden of Disease does include neurological diseases that are developmental, like Autism and tries to quantify the disease burden with methodologies like the disability-adjusted life year (DALY), this might not capture many cases of suboptimal mental development and many relevant social costs. And so while we have 'global burden of disease study', we do not yet have a 'Global Burden of suboptimal intelligence study'. This study measures how many IQ points are lost due to various causes and risk factors, just like the Global Burden of Disease study at the Institute for Health Metrics (See visualization here).
For instance, a child who through poor nutrition in childhood has their IQ reduced from 140 to 125, and thus still being normal functioning, perhaps even have a higher self-reported well-being, but this is not captured by the GBD.
A philanthropist could fund scientists to try to comprehensively map the Global Burden of suboptimal neurodevelopment from the biological to the social (more below).
In a second step, the Gates Foundation funded the Disease Control Priorities Project (DCP) (which inspired Toby Ord's seminal EA paper The Moral Imperative toward Cost-Effectiveness in Global Health).[58] The DCP tried to find estimate the cost-effectiveness of different global health interventions with measures like $ per DALY averted and then rank them by their cost-effectiveness.
Analogously, one fund a Neurodevelopmental Priorities Project that aims to find what interventions are most cost-effective to improve intelligence/ cognition i.e. how many IQ points per $. It might look at different global mental interventions to prevent say IQ loss (e.g. reducing lead and pesticide exposure, improving infant nutrition, micronutrient fortification, etc.), and then see how much it would cost to implement the interventions to arrive at a cost-effectiveness metric such as $ per IQ point loss averted. In addition to finding high benefit-cost ratios, one might also look use the GBD-equivalent suggested earlier to where the highest benefit minus cost can be found and the efficient of the two are (e.g. relatively cost-effective interventions i.e. high benefit-cost ratios in terms of say $ per IQ point loss averted that can scale to avert billions of lost IQ points). Crucially, DCN must not necessarily be about IQ, but could be about other cognitive dysfunction that is on the normal spectrum and has social costs or creates harm. One might imagine suboptimal neural development that increases depression or anxiety.
Eventually, the field might only look at 'biological causes' but also examine how intelligence and executive functioning is reduced by:
Eventually it could also construe 'development' more widely and look at IQ loss across the lifespan (according to famous neuroscientist Eric Kandel something a double digit percentage of all adults over 65 experience 'mild cognitive decline' something that is not dementia but might not be captured by the GBD, but could potentially even helped by therapeutics like nootropics, but is currently not deemed treatable.
Consider for inspiration this figure of a model of from 'The Mental Wealth of Nations' model:[62]
Establishing the field of Developmental Cognitive Neuroepidemiology might help with human flourishing across the lifespan.
[1] “Uauy and Dangour (2006); Georgieff (2007); Stewart et al. (2008); Eppig et al. (2010); Cotman and Berchtold (2002).”
[2] Field, Erica, Omar Robles, and Maximo Torero. 2009. "Iodine Deficiency and Schooling Attainment in Tanzania." American Economic Journal: Applied Economics, 1(4): 140-69.
[3] Bougma, Karim et al. "Iodine and mental development of children 5 years old and under: a systematic review and meta-analysis." Nutrients 5.4 (2013): 1384-1416.
[4] https://psycnet.apa.org/record/2007-12577-007
[6] (Barnett et al., 2018; McCoy et al., 2017; Morris et al., 2018; van Huizen & Plantenga, 2018) as reviewed in Use of Quasi-Experimental Research Designs in Education Research: Growth, Promise, and Challenges - Maithreyi Gopalan, Kelly Rosinger, Jee Bin Ahn, 2020
[15] Estimating the Health and Economic Benefits of Universal Salt Iodization Programs to Correct Iodine Deficiency Disorders | Thyroid
[16] The new emergence of iodine deficiency in the UK: consequences for child neurodevelopment - Margaret P Rayman, Sarah C Bath, 2015
[17] Zimmermann, Michael B, Pieter L Jooste, and Chandrakant S Pandav. 'Iodine-deficiency disorders.' The Lancet 372.9645 (2008): 1251-1262.
[20] Raising IQ among school-aged children: Five meta-analyses and a review of randomized controlled trials - ScienceDirect
[21] The Toxic Truth: Children’s Exposure to Lead Pollution Undermines a Generation of Future Potential
[22] Ethan Mollick on Twitter: "Bad news: Leaded fuel reduced the IQ of everyone born before 1990 by ~4.25%. Millennials are the first to be born with unleaded gas. Worse news: a new paper shows environmental lead levels from leaded gasoline are still around in cities today, and cause continued neurotoxicity. https://t.co/TAWAIU2nTH" / Twitter
[23] Environmental lead levels from leaded gasoline are still around in cities today, and cause continued neurotoxicity.
[25] Infectious disease burden and cognitive function in young to middle-aged adults - ScienceDirect
[27] Field, Erica, Omar Robles, and Maximo Torero. 2009. "Iodine Deficiency and Schooling Attainment in Tanzania." American Economic Journal: Applied Economics, 1(4): 140-69.
[28] Bougma, Karim et al. "Iodine and mental development of children 5 years old and under: a systematic review and meta-analysis." Nutrients 5.4 (2013): 1384-1416.
[29] Nguyen, Trang. "Information, role models and perceived returns to education: Experimental evidence from Madagascar." Unpublished manuscript 6 (2008).
[30] Anekwe, Tobenna D et al. "The causal effect of childhood measles vaccination on educational attainment: A mother fixed-effects study in rural South Africa." Vaccine 33.38 (2015): 5020-5026.
[31] Driessen, Julia et al. "Comparing the health and social protection effects of measles vaccination strategies in Ethiopia: An extended cost-effectiveness analysis." Social Science & Medicine 139 (2015): 115-122.
[32] Miguel, Edward, and Michael Kremer. "Worms: identifying impacts on education and health in the presence of treatment externalities." Econometrica 72.1 (2004): 159-217.
[33] Hicks, Joan Hamory, Michael Kremer, and Edward Miguel. "Commentary: Deworming externalities and schooling impacts in Kenya: a comment on Aiken et al.(2015) and Davey et al.(2015)." International journal of epidemiology 44.5 (2015): 1593-1596.
[34] Kuecken, Maria, Josselin Thuilliez, and Marie-Anne Valfort. "Does malaria control impact education? Evidence from Roll Back Malaria in Africa." (2015).
[35] Bobonis, Gustavo J, Edward Miguel, and Charu Puri-Sharma. "Anemia and school participation." Journal of Human resources 41.4 (2006): 692-721.
[36] Burde, Dana, and Leigh L Linden. "Bringing education to Afghan girls: A randomized controlled trial of village-based schools." American Economic Journal: Applied Economics 5.3 (2013): 27-40.
[37] Evans, David, Michael Kremer, and Muthoni Ngatia. "The impact of distributing school uniforms on children’s education in Kenya." World Bank, mimeo (2008).
[38] Jensen, Robert. "The (perceived) returns to education and the demand for schooling." Quarterly Journal of Economics 125.2 (2010).
[39] Kremer, Michael, Edward Miguel, and Rebecca Thornton. "Incentives to learn." The Review of Economics and Statistics 91.3 (2009): 437-456.
[40] Friedman, Willa et al. "Education as Liberation?." 7 Apr. 2011.
[41] Baird, Sarah, Craig McIntosh, and Berk Özler. "Cash or condition? Evidence from a cash transfer experiment." The Quarterly Journal of Economics (2011): qjr032.
[42] ibid.
[43] ibid.
[44] Schultz, T Paul. "Impact of PROGRESA on school attendance rates in the sampled population." February. Report submitted to PROGRESA. International Food Policy Research Institute, Washington, DC (2000).
[45] Duflo, Esther, Rema Hanna, and Stephen P Ryan. "Incentives work: Getting teachers to come to school." The American Economic Review (2012): 1241-1278.
[46] Banerjee, Abhijit et al. "Remedying education: Evidence from two randomized experiments in India." 26 Dec. 2005.
[47] ibid.
[48] Oster, Emily, and Rebecca Thornton. "Menstruation, sanitary products, and school attendance: Evidence from a randomized evaluation." American Economic Journal: Applied Economics (2011): 91-100.
[49] Oster, Emily, and Rebecca Thornton. "Determinants Of Technology Adoption: Peer Effects In Menstrual Cup Take‐Up." Journal of the European Economic Association 10.6 (2012): 1263-1293.
[50] Wydick, Bruce, Paul Glewwe, and Laine Rutledge. "Do International Child Sponsorship Programs Work? Evidence from Six Countries Using a Regression Discontinuity Design." (2010).
[51] "FSI | REAP - Nutrition and Educational Performance in Rural ..." 16 Mar. 2016 <Nutrition and Educational Performance in Rural China's Elementary Schools: Results of a Randomized Control Trial in Shaanxi Province | FSI>
[52] Luo, Renfu et al. "Nutrition and educational performance in rural China’s elementary schools: Results of a randomized control trial in Shaanxi Province." Economic development and cultural change 60.4 (2012): 735-772.
[53] Dhaliwal, Iqbal, Esther Duflo, Rachel Glennerster, and Caitlin Tulloch. " Comparative Cost-Effectiveness Analysis to Inform Policy in Developing Countries" In Education Policy in Developing Countries, edited by Paul Glewwe, 285-388. Chicago: University of Chicago Press, 2014.
[54] Youssef Oulhote - Assistant Professor of Epidemiology - University of Massachusetts Amherst | LinkedIn
[55] Neurological and Developmental Outcome in Extremely Preterm Children Born in England in 1995 and 2006: The EPICure Studies | Request PDF
[59] Brain Drain: The Mere Presence of One’s Own Smartphone Reduces Available Cognitive Capacity | Journal of the Association for Consumer Research