Superb piece. The obesogenic study realy flips the usual framing. I've always thought higher heritability meant more genetic inevitability, but watching how the same twins show different heritability based on their home environment makes it obvious that the environment is the variable doing the work. The distinction betwene intrinsic and extrinsic deaths is such a clean example of reassignment logic.
Thank you Carole. I went back and forth about making this two essays, but it seemed more compelling to pair epigenetics with the obesity–environment study to show how much heritability depends on environment.
Lately I find myself writing 1/2 an oped about a problem and the other 1/2 about the solution. I used to simply mention the problem (10%) and go right into the solution (90%). But the details of problems are so outrageous, they seem to make the solution wanted even more, and right now....no waiting. I don't like dwelling on what's wrong, but how to do something about it. When doing my Phd. 20 years ago, I really disliked research funding only going to estimating the extent of problems, and never going to solutions. I hope things have changed.
Fantastic article, Bruce! As a pediatric obesity person attempting to explain the genetics of obesity, I frequent stumble all over myself. Certainly that is a sign of my own issues of conflating genes and heritability. I am going to shamelessly use your words moving forward. Great job!
People like dichotomy. Something is right and something is wrong. A person is either good or bad. Life is messier than that. But that white hat - black hat approach to life is perhaps one of the biggest challenges in everything: medicine, science, relationships.....
IMO: Putting financial incentives aside, I believe the main reason we have so much polarization in politics, family estrangement, and contested science is that people want the world to operate according to their simplistic view of right and wrong, with everything they believe being right and everything else being wrong.
They want "a pill for every ill" and they believe in 1-to-1 correlations because it's intellectually tidy. But life is messy, and a well-lived life requires humility and compassion.
Counter-Argument: Environment Doesn't Just Amplify Genetic Vulnerabilities—It Actively Shapes Population Genetics, Making Genes More Central to Disease
While Bruce Lanphear's article makes a compelling case that heritability estimates are often misinterpreted and that environmental changes drive the surge in chronic diseases like cancer, autism, obesity, and dementia—without requiring major genetic shifts—I disagree with the core implication that genes have "barely changed at all" over the past century. This overlooks how harmful environments and modern interventions dynamically alter population genetics through increased mutations and relaxed natural selection. Far from genes being static, the environment is actively reshaping our genome, accumulating deleterious variants and heightening genetic susceptibility to those same environmental insults. This creates a vicious cycle where genetics become increasingly responsible for disease prevalence, even as environments worsen.
1. Environmental Factors Directly Induce Mutations, Altering Population Genetics Over Generations
Lanphear argues that diseases have risen sharply without genetic changes, pointing to historical trends like the surge in heart disease or autism diagnoses. However, this underestimates how environmental mutagens—chemicals, pollution, radiation, and toxins—can cause heritable mutations in germline cells (sperm and eggs), leading to population-level genetic shifts. These mutagens don't just damage somatic cells (leading to individual cancers or other diseases); they can increase mutation rates in reproductive cells, passing on new variants to offspring. For example, exposure to air pollution, pesticides, or industrial chemicals has been linked to higher somatic and potentially germline mutation burdens, contributing to genetic diversity (or instability) in modern populations.
Over the 20th century, as industrialization and pollution intensified, these mutagens likely accelerated genetic changes far beyond baseline rates. While mutation accumulation is gradual, it's not negligible: studies estimate that environmental genotoxins shape somatic mutation landscapes in tissues, and if they affect germ cells, they could explain part of the rise in heritable disease risks. This isn't "sudden" as Lanphear dismisses, but cumulative—environment erodes genetic stability, making populations more prone to diseases that heritability estimates then capture as "genetic."
2. Medical Interventions Relax Natural Selection, Increasing the Prevalence of Deleterious Genetic Variants
A key flaw in the article is dismissing how human interventions—like vaccinations, antibiotics, and nutritional supplements—allow individuals with genetic vulnerabilities to survive and reproduce, thereby increasing those variants' frequency in the population. This is "relaxed selection," where modern medicine weakens the evolutionary pressures that once weeded out harmful alleles. In pre-modern eras, many genetic conditions (e.g., those causing early mortality or infertility) would limit reproduction; today, treatments enable survival, passing on those genes.
Your example of folate supplementation is spot on. Mandatory fortification of foods with folic acid (e.g., in grains since the 1990s in many countries) prevents neural tube defects, particularly in fetuses with MTHFR gene variants (like C677T), which impair folate metabolism. Homozygotes (TT) for MTHFR C677T have reduced enzyme activity, historically increasing risks of birth defects or miscarriage—but supplementation allows these pregnancies to succeed, potentially raising the allele's prevalence over generations. While direct longitudinal data on prevalence shifts is limited, the logic of relaxed selection applies: by saving lives, we're enriching the gene pool with variants that make future generations more reliant on interventions. This extends to other diseases; for instance, advanced care for dementia patients may allow related genetic risks to persist, contributing to rising incidence.
3. Gene-Environment Feedback Loops: Mutations Like MTHFR Heighten Sensitivity to Toxins, Amplifying Genetic Roles in Disease
Tying points together, these genetic changes create feedback where altered genomes make us more vulnerable to environments, flipping Lanphear's "heritability paradox" on its head. In harmful settings, high heritability signals not just amplified vulnerabilities but evolving genetics that exacerbate the issue.
Take MTHFR homozygotes and arsenic: Variants like C677T impair arsenic methylation, leading to higher retention of toxic inorganic arsenic or monomethylarsonic acid (MMA) in the body, increasing risks of skin lesions, cancer, or developmental issues. If folate supplementation increases MTHFR variant survival (as above), and arsenic exposure is rising environmentally (e.g., in water or food), then populations accumulate more sensitive individuals—making diseases seem more "genetic" because the gene pool has shifted. This isn't environment alone; it's environment sculpting genetics, which then heightens disease risks.
Similar loops apply to autism, cancer, or obesity: Mutagens like pollutants induce new variants, while medicine preserves others, leading to genomes less resilient to modern stressors.
Conclusion: Genes Aren't Inevitable, But Neither Are They Static—We Need a Nuanced View
Lanphear is right that prevention should focus on environments, not just genes, and that heritability doesn't mean destiny. But by downplaying genetic evolution, the article misses how environments and interventions are making genetics more relevant to disease. Over centuries, this could erode human fitness, as deleterious alleles accumulate. True optimism lies in addressing both: cleaning environments to reduce mutagenesis and considering how interventions might inadvertently shift genetics. Ignoring this interplay risks underestimating the long-term costs of our "progress."
Bogdan—thank you for this careful and generous critique. I’m grateful you took the time to engage so deeply with the essay and to expand the argument in ways that sharpen and extend it.
On your first point, I absolutely agree. In my own simplified way, this perspective is fully compatible with what I was trying to convey. The idea that environments don’t merely trigger disease but can actively shape population genetics over time is an important extension—and one that reinforces, rather than contradicts, the case for prevention.
On the second point, I also generally agree. Modern medicine does relax certain forms of natural selection, and that reality deserves honest discussion. The one nuance I’d add is that, in theory, some interventions—vaccines such as smallpox—become moot if we can truly eradicate the underlying threat. But your broader point about survival, reproduction, and the persistence of vulnerability is well taken.
On the third point, I agree again, and I see this as entirely consistent with how I describe heritability: genes and environments locked in feedback loops, with heritability reflecting that interaction rather than genetic inevitability.
Thank you again for pushing the essay further and for articulating these ideas so clearly. I appreciate the dialogue and the opportunity to learn from it.
Genes matter....and they can be directly influenced by the environment especially with regards to developing embryos/fetuses. Occupational exposures of either parent can create de novo mutations that can change the life course of the developing infant. We must never forget the environment and its interplay with our genomes.
I agree directionally with the article. It's vexing that environment and history get forgotten so quickly.
That said, though, I think that this article would be improved by acknowledging that there are numerous diseases and conditions that are almost perfectly heritable and are therefore much closer to the layman understanding of heritability.
I think it's important to lay the distinction. Genetic diseases, or traits, are almost uncountable and they do vary along a spectrum in terms of the genetic contribution.
Luca: A fair point. I try to keep my essays below 1250 words and this one was already more than 1500 words. When evidence shows that environmental influences are driving increases in chronic disease—which is true for most chronic diseases—than prevention should be our primary response. When people suffer from conditions with no identifiable environmental cause—or from diseases with a clearly defined genetic origin, such as Huntington’s disease—our focus appropriately shifts to treatment and long-term management. Too often, we neglect upstream environmental risk factors. When heritability findings are interpreted to mean that, say, 50% of life expectancy is “genetic” and therefore unavoidable—rather than a statistical estimate of variation within a given environment—we misread the science and risk missing opportunities for prevention.
Excellent post! I saw the provocative Science headline but haven’t had a chance to dive in. But my gut feeling was exactly what you describe. We know “heritability” is context specific and not a biological constant as people commonly assume…. I like the example of in a population where everyone smokes, lung cancer is 100% genetic.
This is such an important corrective, and you make it with the right mix of rigor and moral clarity: heritability is not destiny, and it’s certainly not causation. 
As a physician-scientist, the line that landed for me is your population-level logic: genes don’t change fast enough to explain rapid shifts in disease incidence, but environments do; sometimes within months. When we reduce lead exposures, cardiovascular outcomes improve; when air pollution drops, deaths fall; when smoking declines, lung cancer follows. None of those wins required “better genes.” 
I also appreciate you naming why the heredity narrative persists: it’s tidy, it fits biomedical training, and—crucially—it’s politically and commercially convenient because it shifts responsibility away from regulation and prevention. That’s not an academic quibble; it shapes what gets funded, what gets measured, and what patients are told is “just your family history”.
Your closing questions are exactly the ones we should be teaching trainees to ask: what changed before the disease rose, do rates fall when exposures fall, are there critical windows, and do safer alternatives exist? That’s the playbook that built modern public health and it’s the playbook we need for today’s chronic disease wave.
I suspect many diseases are attributed to "genes" because we don't count slow poisons, which have multiplied exponentially - and I don't use that word lightly - over the past century.
Tracy: I suspect you are correct. As Gundle’s research shows, the tobacco industry deliberately funded studies to explain why some people developed lung cancer while others did not. By emphasizing genetic susceptibility, the industry could recast lung cancer as a problem of individual biology rather than a consequence of smoking itself. That shift—away from cigarettes and toward genes—helped distract from the causal role of smoking and allowed companies to continue selling a product they knew was carcinogenic. https://surl.li/lghxtz
Very nice article! I particularly like the heritability paradox with "poor" environment enhancing genetic effects in twins. It's interesting that Science published a Twin study which is primitive (the design was invented before Watson and Crick) and requires no genomic measures. In GWAS studies where the genome is measured, the heritability estimates always fall below the Twins study. What genetics calls "missing heritability" is probably the heritability paradox you describe. i.e. shared "poor" environment.
A great article. I must confess I did not come away with the distinction between genetics and heritability from my classes with the esteemed Dr. Suzuki in the early '80s...my comprehension lacking...not his teaching. It has been articles like this and a brilliant article https://substack.com/home/post/p-176634675 by Dr. Robert Wright regarding chickens, white corn and yellow corn that have nailed the understanding for me. Many thanks.
Thank you Doreen: I’ve also learned a great deal about heritability from Bob Wright. What’s striking is not how complex heritability is, but how deceptively simple it is—and how few people truly understand what it does and does not mean. Despite this, heritability estimates are routinely taken at face value, leading to confident claims that “80% of this” or “60% of that” is due to genes. Have you read Besse’s essay on twin studies? It was an eye-opener for me. You can find it here: https://substack.com/home/post/p-156985605
In reading your work, I am reminded that progress or lack thereof, is a measure of how well we craft inquiries and degree of scrutiny we give to responses. I am also reminded that the challenge always is to ask the right question. Thank you Dr. Lanphear.
Absolutely! "Heritability is an oddly comforting statistic. It has a crisp number attached to it. It sounds precise. Scientific. Authoritative. But heritability does not mean what most people—including many journalists, physicians, and scientists—think it means."
Your points are all great. Yet, it's even worse. Heritability estimates rely on the assumption that there is no interaction between genes and environment (GxE). In the presence of GxE, heritability estimates are inherently inflated, sometimes dramatically. Virtually no heritability studies validate that assumption. And since it's an input assumption, such studies cannot possibly show that assumption is true.
Heritability estimates show *associations* between genetics and outcomes, not *causality*. As every scientist knows, association does not mean causation. But many conveniently forget this when it comes to genetics and the environment. Conclusion: genetics can confer increased *susceptibility* but are generally not sufficient to cause the outcome. That's consistent with that your post says.
And, heritability estimates do not work as expected when the outcome is binary. For example, autistic or not. These estimates work on *variance*, not absolute values. There is solid analytical published evidence about the dramatic overestimates that occur with binary outcomes.
Superb piece. The obesogenic study realy flips the usual framing. I've always thought higher heritability meant more genetic inevitability, but watching how the same twins show different heritability based on their home environment makes it obvious that the environment is the variable doing the work. The distinction betwene intrinsic and extrinsic deaths is such a clean example of reassignment logic.
Thanks Bruce, for opening my eyes once again, to be able to see hidden connections I had no idea were there.
My two favorite lines
"Genetics explain who’s at risk, but the environment determines who gets sick."
"Many of today’s chronic diseases are not inevitable. They are preventable." (this line I will quote in an oped, thanks)
Thank you Carole. I went back and forth about making this two essays, but it seemed more compelling to pair epigenetics with the obesity–environment study to show how much heritability depends on environment.
Lately I find myself writing 1/2 an oped about a problem and the other 1/2 about the solution. I used to simply mention the problem (10%) and go right into the solution (90%). But the details of problems are so outrageous, they seem to make the solution wanted even more, and right now....no waiting. I don't like dwelling on what's wrong, but how to do something about it. When doing my Phd. 20 years ago, I really disliked research funding only going to estimating the extent of problems, and never going to solutions. I hope things have changed.
Fantastic article, Bruce! As a pediatric obesity person attempting to explain the genetics of obesity, I frequent stumble all over myself. Certainly that is a sign of my own issues of conflating genes and heritability. I am going to shamelessly use your words moving forward. Great job!
Thank you, Chris! I can picture you stumbling for a moment—and then righting yourself, that beautiful smile already back on your face.
"Intellectually tidy" - neat phrase.
People like dichotomy. Something is right and something is wrong. A person is either good or bad. Life is messier than that. But that white hat - black hat approach to life is perhaps one of the biggest challenges in everything: medicine, science, relationships.....
IMO: Putting financial incentives aside, I believe the main reason we have so much polarization in politics, family estrangement, and contested science is that people want the world to operate according to their simplistic view of right and wrong, with everything they believe being right and everything else being wrong.
They want "a pill for every ill" and they believe in 1-to-1 correlations because it's intellectually tidy. But life is messy, and a well-lived life requires humility and compassion.
Counter-Argument: Environment Doesn't Just Amplify Genetic Vulnerabilities—It Actively Shapes Population Genetics, Making Genes More Central to Disease
While Bruce Lanphear's article makes a compelling case that heritability estimates are often misinterpreted and that environmental changes drive the surge in chronic diseases like cancer, autism, obesity, and dementia—without requiring major genetic shifts—I disagree with the core implication that genes have "barely changed at all" over the past century. This overlooks how harmful environments and modern interventions dynamically alter population genetics through increased mutations and relaxed natural selection. Far from genes being static, the environment is actively reshaping our genome, accumulating deleterious variants and heightening genetic susceptibility to those same environmental insults. This creates a vicious cycle where genetics become increasingly responsible for disease prevalence, even as environments worsen.
1. Environmental Factors Directly Induce Mutations, Altering Population Genetics Over Generations
Lanphear argues that diseases have risen sharply without genetic changes, pointing to historical trends like the surge in heart disease or autism diagnoses. However, this underestimates how environmental mutagens—chemicals, pollution, radiation, and toxins—can cause heritable mutations in germline cells (sperm and eggs), leading to population-level genetic shifts. These mutagens don't just damage somatic cells (leading to individual cancers or other diseases); they can increase mutation rates in reproductive cells, passing on new variants to offspring. For example, exposure to air pollution, pesticides, or industrial chemicals has been linked to higher somatic and potentially germline mutation burdens, contributing to genetic diversity (or instability) in modern populations.
Over the 20th century, as industrialization and pollution intensified, these mutagens likely accelerated genetic changes far beyond baseline rates. While mutation accumulation is gradual, it's not negligible: studies estimate that environmental genotoxins shape somatic mutation landscapes in tissues, and if they affect germ cells, they could explain part of the rise in heritable disease risks. This isn't "sudden" as Lanphear dismisses, but cumulative—environment erodes genetic stability, making populations more prone to diseases that heritability estimates then capture as "genetic."
2. Medical Interventions Relax Natural Selection, Increasing the Prevalence of Deleterious Genetic Variants
A key flaw in the article is dismissing how human interventions—like vaccinations, antibiotics, and nutritional supplements—allow individuals with genetic vulnerabilities to survive and reproduce, thereby increasing those variants' frequency in the population. This is "relaxed selection," where modern medicine weakens the evolutionary pressures that once weeded out harmful alleles. In pre-modern eras, many genetic conditions (e.g., those causing early mortality or infertility) would limit reproduction; today, treatments enable survival, passing on those genes.
Your example of folate supplementation is spot on. Mandatory fortification of foods with folic acid (e.g., in grains since the 1990s in many countries) prevents neural tube defects, particularly in fetuses with MTHFR gene variants (like C677T), which impair folate metabolism. Homozygotes (TT) for MTHFR C677T have reduced enzyme activity, historically increasing risks of birth defects or miscarriage—but supplementation allows these pregnancies to succeed, potentially raising the allele's prevalence over generations. While direct longitudinal data on prevalence shifts is limited, the logic of relaxed selection applies: by saving lives, we're enriching the gene pool with variants that make future generations more reliant on interventions. This extends to other diseases; for instance, advanced care for dementia patients may allow related genetic risks to persist, contributing to rising incidence.
3. Gene-Environment Feedback Loops: Mutations Like MTHFR Heighten Sensitivity to Toxins, Amplifying Genetic Roles in Disease
Tying points together, these genetic changes create feedback where altered genomes make us more vulnerable to environments, flipping Lanphear's "heritability paradox" on its head. In harmful settings, high heritability signals not just amplified vulnerabilities but evolving genetics that exacerbate the issue.
Take MTHFR homozygotes and arsenic: Variants like C677T impair arsenic methylation, leading to higher retention of toxic inorganic arsenic or monomethylarsonic acid (MMA) in the body, increasing risks of skin lesions, cancer, or developmental issues. If folate supplementation increases MTHFR variant survival (as above), and arsenic exposure is rising environmentally (e.g., in water or food), then populations accumulate more sensitive individuals—making diseases seem more "genetic" because the gene pool has shifted. This isn't environment alone; it's environment sculpting genetics, which then heightens disease risks.
Similar loops apply to autism, cancer, or obesity: Mutagens like pollutants induce new variants, while medicine preserves others, leading to genomes less resilient to modern stressors.
Conclusion: Genes Aren't Inevitable, But Neither Are They Static—We Need a Nuanced View
Lanphear is right that prevention should focus on environments, not just genes, and that heritability doesn't mean destiny. But by downplaying genetic evolution, the article misses how environments and interventions are making genetics more relevant to disease. Over centuries, this could erode human fitness, as deleterious alleles accumulate. True optimism lies in addressing both: cleaning environments to reduce mutagenesis and considering how interventions might inadvertently shift genetics. Ignoring this interplay risks underestimating the long-term costs of our "progress."
Bogdan—thank you for this careful and generous critique. I’m grateful you took the time to engage so deeply with the essay and to expand the argument in ways that sharpen and extend it.
On your first point, I absolutely agree. In my own simplified way, this perspective is fully compatible with what I was trying to convey. The idea that environments don’t merely trigger disease but can actively shape population genetics over time is an important extension—and one that reinforces, rather than contradicts, the case for prevention.
On the second point, I also generally agree. Modern medicine does relax certain forms of natural selection, and that reality deserves honest discussion. The one nuance I’d add is that, in theory, some interventions—vaccines such as smallpox—become moot if we can truly eradicate the underlying threat. But your broader point about survival, reproduction, and the persistence of vulnerability is well taken.
On the third point, I agree again, and I see this as entirely consistent with how I describe heritability: genes and environments locked in feedback loops, with heritability reflecting that interaction rather than genetic inevitability.
Thank you again for pushing the essay further and for articulating these ideas so clearly. I appreciate the dialogue and the opportunity to learn from it.
Genes matter....and they can be directly influenced by the environment especially with regards to developing embryos/fetuses. Occupational exposures of either parent can create de novo mutations that can change the life course of the developing infant. We must never forget the environment and its interplay with our genomes.
Thank you Rick. I agree. We need more GxE studies to appreciate how genes and environment interact. That will be challenging, but long overdue.
I have suspected this for a while but hasn't read it explained so well before, thank you.
I agree directionally with the article. It's vexing that environment and history get forgotten so quickly.
That said, though, I think that this article would be improved by acknowledging that there are numerous diseases and conditions that are almost perfectly heritable and are therefore much closer to the layman understanding of heritability.
I think it's important to lay the distinction. Genetic diseases, or traits, are almost uncountable and they do vary along a spectrum in terms of the genetic contribution.
Luca: A fair point. I try to keep my essays below 1250 words and this one was already more than 1500 words. When evidence shows that environmental influences are driving increases in chronic disease—which is true for most chronic diseases—than prevention should be our primary response. When people suffer from conditions with no identifiable environmental cause—or from diseases with a clearly defined genetic origin, such as Huntington’s disease—our focus appropriately shifts to treatment and long-term management. Too often, we neglect upstream environmental risk factors. When heritability findings are interpreted to mean that, say, 50% of life expectancy is “genetic” and therefore unavoidable—rather than a statistical estimate of variation within a given environment—we misread the science and risk missing opportunities for prevention.
Excellent post! I saw the provocative Science headline but haven’t had a chance to dive in. But my gut feeling was exactly what you describe. We know “heritability” is context specific and not a biological constant as people commonly assume…. I like the example of in a population where everyone smokes, lung cancer is 100% genetic.
This is such an important corrective, and you make it with the right mix of rigor and moral clarity: heritability is not destiny, and it’s certainly not causation. 
As a physician-scientist, the line that landed for me is your population-level logic: genes don’t change fast enough to explain rapid shifts in disease incidence, but environments do; sometimes within months. When we reduce lead exposures, cardiovascular outcomes improve; when air pollution drops, deaths fall; when smoking declines, lung cancer follows. None of those wins required “better genes.” 
I also appreciate you naming why the heredity narrative persists: it’s tidy, it fits biomedical training, and—crucially—it’s politically and commercially convenient because it shifts responsibility away from regulation and prevention. That’s not an academic quibble; it shapes what gets funded, what gets measured, and what patients are told is “just your family history”.
Your closing questions are exactly the ones we should be teaching trainees to ask: what changed before the disease rose, do rates fall when exposures fall, are there critical windows, and do safer alternatives exist? That’s the playbook that built modern public health and it’s the playbook we need for today’s chronic disease wave.
I suspect many diseases are attributed to "genes" because we don't count slow poisons, which have multiplied exponentially - and I don't use that word lightly - over the past century.
Tracy: I suspect you are correct. As Gundle’s research shows, the tobacco industry deliberately funded studies to explain why some people developed lung cancer while others did not. By emphasizing genetic susceptibility, the industry could recast lung cancer as a problem of individual biology rather than a consequence of smoking itself. That shift—away from cigarettes and toward genes—helped distract from the causal role of smoking and allowed companies to continue selling a product they knew was carcinogenic. https://surl.li/lghxtz
Very nice article! I particularly like the heritability paradox with "poor" environment enhancing genetic effects in twins. It's interesting that Science published a Twin study which is primitive (the design was invented before Watson and Crick) and requires no genomic measures. In GWAS studies where the genome is measured, the heritability estimates always fall below the Twins study. What genetics calls "missing heritability" is probably the heritability paradox you describe. i.e. shared "poor" environment.
Thank you Bob. I've learned from the expert (you).
A great article. I must confess I did not come away with the distinction between genetics and heritability from my classes with the esteemed Dr. Suzuki in the early '80s...my comprehension lacking...not his teaching. It has been articles like this and a brilliant article https://substack.com/home/post/p-176634675 by Dr. Robert Wright regarding chickens, white corn and yellow corn that have nailed the understanding for me. Many thanks.
Thank you Doreen: I’ve also learned a great deal about heritability from Bob Wright. What’s striking is not how complex heritability is, but how deceptively simple it is—and how few people truly understand what it does and does not mean. Despite this, heritability estimates are routinely taken at face value, leading to confident claims that “80% of this” or “60% of that” is due to genes. Have you read Besse’s essay on twin studies? It was an eye-opener for me. You can find it here: https://substack.com/home/post/p-156985605
Thanks Bruce…I have not read Bessis’ post. Will tuck into it tonight. The topic is fascinating and helpful when reading articles on autism.
In reading your work, I am reminded that progress or lack thereof, is a measure of how well we craft inquiries and degree of scrutiny we give to responses. I am also reminded that the challenge always is to ask the right question. Thank you Dr. Lanphear.
Absolutely! "Heritability is an oddly comforting statistic. It has a crisp number attached to it. It sounds precise. Scientific. Authoritative. But heritability does not mean what most people—including many journalists, physicians, and scientists—think it means."
Your points are all great. Yet, it's even worse. Heritability estimates rely on the assumption that there is no interaction between genes and environment (GxE). In the presence of GxE, heritability estimates are inherently inflated, sometimes dramatically. Virtually no heritability studies validate that assumption. And since it's an input assumption, such studies cannot possibly show that assumption is true.
Heritability estimates show *associations* between genetics and outcomes, not *causality*. As every scientist knows, association does not mean causation. But many conveniently forget this when it comes to genetics and the environment. Conclusion: genetics can confer increased *susceptibility* but are generally not sufficient to cause the outcome. That's consistent with that your post says.
And, heritability estimates do not work as expected when the outcome is binary. For example, autistic or not. These estimates work on *variance*, not absolute values. There is solid analytical published evidence about the dramatic overestimates that occur with binary outcomes.