Dietary effect strong when high genetic predisposition to weight gain
as we head
apparently unabatedly into genetic medicine, it was eye-opening to see a
study finding that a healthier diet was actually most effective in weight
management in people who are at the highest genetic risk for weight
gain (see doi:
10.1136/bmj.j5644).
Details:
--2 databases: 8828 women from the Nurses' Health Study and 5218 men form the Health Professionals Follow-Up Study
--genetic predisposition to weight gain determined on the basis of 77 variants associated with body mass index (BMI)
--dietary patterns were assessed by 3 diets: the Alternate Healthy Eating Index 2010 (AHEI-2010), Dietary Approach to Stop Hypertension (DASH), and the Alternate Mediterranean Diet (AMED)
--5 repeated measurements of 4-year changes in BMI and weight, from 1986-2006
Results:
--overall, the genetic risk score was associated with increases in BMI and weight every 4 years, with each additional 10 risk alleles (single nucleotide polymorphisms, SNPs) being associated with a 0.02 increase in BMI and 0.05 kg weight increase
--and, increases in dietary quality were associated with decreases in BMI and body weight
--over the 20 year follow-up:
--genetic association with change in BMI was significantly attenuated as adherence to AHEI-2000 increased (p=0.001 for the interaction for the Nurses' Health Study and p=0.005 for the interaction in the Health Professionals Follow-Up Study)
--changes in BMI per 10 genetic risk allele increment:
--was +0.07 with decreased AHEI-2010 score
--and -0.01 with increased AHEI-2010 score
--BMI and weight change among people, per 1 SD increment in AHEI-2010 score (healthier diet) by genetic risk:
--low genetic risk: -0.12 for BMI, -0.35 kg
--moderate genetic risk: -0.14 for BMI, -0.36 kg
--high genetic risk: -0.18 for BMI, -0.50 kg
--the above results were similar but weaker in those <65yo and never-smokers
--similar interactions for DASH diet but not for AMED
Commentary:
--one curious finding above was that AMED did not have this genetic association. the authors attribute that to the fact that, unlike AMED, AHEI-2010 had a continuous scale for evaluating dietary changes over time, so was better able to quantify these changes accurately
--there are studies (some by the same group as the above one) finding that certain foods themselves may further modify the relationship between diet and genetic risk in weight changes: coffee consumption can attenuate the genetic association, fried foods and sugar-sweetened beverages can strengthen the genetic association (ie, in the latter case, these foods enhance the genetic effect on weight)
--there are several caveats to the study and its conclusions: they assessed the known 77 genetic variants (SNPs) associated with BMI (are there other, undiscovered ones that may be more important? is there an unknown interplay between these SNPs themselves or with other SNPs which are not associated with BMI but are important?); is there reverse causation? (did people in these cohorts who gained weight then adopt a healthier diet to lose weight); and these types of longitudinal epidemiologic studies can only demonstrate association not causality (would need a clear RCT to see causation, but that would be pretty difficult: check genes in large numbers of people, prescribe randomized diets that people continue, and observe for 20 years, while at the same time controlling for smoking, exercise...). And can the results of these 2 large cohorts be generalized to those of non-European descent?
so, this study brings up a few important points:
--i am concerned that we clinicians are being wooed by genetic medicine, and that this onslaught has furthered ideologically the concept that genes are determinant
--and, they found above that the more adverse the genetic load to weight increase, the more powerful the association with a healthy diet (though the effect size was small)
--all of this reinforces the complexity of the interplay between genes and environment. However, this is even more complex, given their direct and short-term interaction as well: the gene/environment interaction is not necessarily fixed; there may well be epigenetic changes whereby the environment or medical comorbidities may turn on or off the function of genes, thereby altering the genetic effect.
--one common clinical scenario that we see a lot is the overweight kids of overweight parents, where the family (and oftentimes us, as well) dismiss the possibility of weight change given that it seems to run in the family. This study, i think, helps right this by showing that these patients may well be even more positively affected by a healthier diet. based on the other studies noted above, it might make sense to focus more heavily on trying to eliminate sugar-sweetened beverages and fried foods, since these food items seem to elevate the genetic effect on weight gain
--and, as a side-line, in my mind this adds further fodder to the advice not to get genetic testing done (eg 23-and-me), since we do not know how to deal with much of the information generated, and this testing really does reinforce the conception that genes are determinant. I am also concerned that the genetic structure of a person does not necessarily determine even the actual function of the genes, given the potential for epigenetic changes during one’s life. see http://gmodestmedblogs.blogspot.com/2017/04/23andme-genetic-analysis-approved-for.html
Details:
--2 databases: 8828 women from the Nurses' Health Study and 5218 men form the Health Professionals Follow-Up Study
--genetic predisposition to weight gain determined on the basis of 77 variants associated with body mass index (BMI)
--dietary patterns were assessed by 3 diets: the Alternate Healthy Eating Index 2010 (AHEI-2010), Dietary Approach to Stop Hypertension (DASH), and the Alternate Mediterranean Diet (AMED)
--5 repeated measurements of 4-year changes in BMI and weight, from 1986-2006
Results:
--overall, the genetic risk score was associated with increases in BMI and weight every 4 years, with each additional 10 risk alleles (single nucleotide polymorphisms, SNPs) being associated with a 0.02 increase in BMI and 0.05 kg weight increase
--and, increases in dietary quality were associated with decreases in BMI and body weight
--over the 20 year follow-up:
--genetic association with change in BMI was significantly attenuated as adherence to AHEI-2000 increased (p=0.001 for the interaction for the Nurses' Health Study and p=0.005 for the interaction in the Health Professionals Follow-Up Study)
--changes in BMI per 10 genetic risk allele increment:
--was +0.07 with decreased AHEI-2010 score
--and -0.01 with increased AHEI-2010 score
--BMI and weight change among people, per 1 SD increment in AHEI-2010 score (healthier diet) by genetic risk:
--low genetic risk: -0.12 for BMI, -0.35 kg
--moderate genetic risk: -0.14 for BMI, -0.36 kg
--high genetic risk: -0.18 for BMI, -0.50 kg
--the above results were similar but weaker in those <65yo and never-smokers
--similar interactions for DASH diet but not for AMED
Commentary:
--one curious finding above was that AMED did not have this genetic association. the authors attribute that to the fact that, unlike AMED, AHEI-2010 had a continuous scale for evaluating dietary changes over time, so was better able to quantify these changes accurately
--there are studies (some by the same group as the above one) finding that certain foods themselves may further modify the relationship between diet and genetic risk in weight changes: coffee consumption can attenuate the genetic association, fried foods and sugar-sweetened beverages can strengthen the genetic association (ie, in the latter case, these foods enhance the genetic effect on weight)
--there are several caveats to the study and its conclusions: they assessed the known 77 genetic variants (SNPs) associated with BMI (are there other, undiscovered ones that may be more important? is there an unknown interplay between these SNPs themselves or with other SNPs which are not associated with BMI but are important?); is there reverse causation? (did people in these cohorts who gained weight then adopt a healthier diet to lose weight); and these types of longitudinal epidemiologic studies can only demonstrate association not causality (would need a clear RCT to see causation, but that would be pretty difficult: check genes in large numbers of people, prescribe randomized diets that people continue, and observe for 20 years, while at the same time controlling for smoking, exercise...). And can the results of these 2 large cohorts be generalized to those of non-European descent?
so, this study brings up a few important points:
--i am concerned that we clinicians are being wooed by genetic medicine, and that this onslaught has furthered ideologically the concept that genes are determinant
--and, they found above that the more adverse the genetic load to weight increase, the more powerful the association with a healthy diet (though the effect size was small)
--all of this reinforces the complexity of the interplay between genes and environment. However, this is even more complex, given their direct and short-term interaction as well: the gene/environment interaction is not necessarily fixed; there may well be epigenetic changes whereby the environment or medical comorbidities may turn on or off the function of genes, thereby altering the genetic effect.
--one common clinical scenario that we see a lot is the overweight kids of overweight parents, where the family (and oftentimes us, as well) dismiss the possibility of weight change given that it seems to run in the family. This study, i think, helps right this by showing that these patients may well be even more positively affected by a healthier diet. based on the other studies noted above, it might make sense to focus more heavily on trying to eliminate sugar-sweetened beverages and fried foods, since these food items seem to elevate the genetic effect on weight gain
--and, as a side-line, in my mind this adds further fodder to the advice not to get genetic testing done (eg 23-and-me), since we do not know how to deal with much of the information generated, and this testing really does reinforce the conception that genes are determinant. I am also concerned that the genetic structure of a person does not necessarily determine even the actual function of the genes, given the potential for epigenetic changes during one’s life. see http://gmodestmedblogs.blogspot.com/2017/04/23andme-genetic-analysis-approved-for.html
to be added to the mailing list, please let me know at gmodest@uphams.org
Comments
Post a Comment
if you would like to receive the near-daily emails regularly, please email me at gmodest@uphams.org