Milk intake and risk of mortality and fractures in women and men: cohort studies
BMJ 2014; 349 doi: https://doi.org/10.1136/bmj.g6015 (Published 28 October 2014) Cite this as: BMJ 2014;349:g6015All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
There is no mention of whether the milk was processed, that is, homogenized. Raw milk is rarely used in studies anymore, even though our conceptions of milk as a healthy beverage come from times when the only milk was raw fresh milk, straight from the cow. The homogenization destroys the milk molecule, smashing it to smithereens. Those tiny fragments become toxic to the body. Hence the enormous increase in the numbers of those allergic to milk in just the past seventy years. For millenia humans drank milk with carefree abandon, and now we are allergic. I admire all your work in this study, but I think you missed the big idea. Dare you upset the Dairy Association by investigating the effects of homogenization?
Competing interests: No competing interests
In the comment of Karl Michaëlsson, Alicja Wolk and Liisa Byberg dated 4th of December 2014, they state the following.
“We note with interest that these authors also reference the incorrect calculations of crude mortality made by Staffan Hellstrand, who is consultant for the Federation of Swedish Farmers, an organization who is owner of milk industries (e.g., Arla Foods). In a previous response (http://www.bmj.com/content/349/bmj.g6015/rr/779932), we have corrected Staffan Hellstrand and guided him to the correct data for use. Despite erroneous figures, Astrup and Givens have chosen to repeat the presentation of Staffan Hellstrand’s incorrect calculations – an interesting action given the fact that Arne Astrup should be an objective key opinion leader in field of nutritional research by his role as editor of the American Journal of Clinical Nutrition. Hopefully, his action to criticize our study by distorted arguments is not related to his financial conflicts of interest, for example research support from Arla Foods (25).”
In a previous commentary I have shown some errors in this passage. In this commentary I will go deeper regarding some challenges in statistical analysis of systems where life is a key system characteristic.
Any one as read the original article; the first commentary of Stefan (not Staffan) Hellstrand; and the first reply of the authors of the original article on my first commentary in parallel; will easily see that the comment of the authors avoids the crucial point put forward in my first response; “that the paper itself with data supplement has not shown in a satisfactory way that the results obtained are results that are valid for the population of humans from which the samples were drawn.”
The figure I presented in the first comment and the calculations made are not incorrect. I clearly defined the measure “mortality” I introduced as the actual fraction that died per subsample. Noteworthy is that another article published in BMJ the 2nd December 2014, in a cohort study relating maternal overweight to risk of infant mortality, gives exactly the same definition of their measure “mortality rate” as the mortality measure I defined in my first comment, see Johansson et al. (2014).
Assume that this measure defined in this way gives incorrect calculations resulting in erroneous figures and conclusions, and that this is in conflict with criteria of good scientific quality regarding among other aspects objectiveness. Then there is a moral obligation for Michaëlsson et al. to guide the scientific community, including Johansson et al. (ibid.) and the editorial board and chief editor of the journal publishing their article, away from this incorrect measure generating erroneous results and conclusions to appropriate methods.
If it is so that the measure I used in my first response, which is the same as the mortality rate in Johansson et al. (ibid.), in some contexts are relevant in cohort studies of human health aspects, then Michaëlsson et al. have some more work to do to show why the measure they have chosen is better, and if the results obtained have such quality after considering strength and weaknesses in their material and the methodology they have chosen given the way they have applied it, that the conclusions they have drawn actually are supported.
The two figures below indicate the challenges they face. In the female cohort study women born in 1913 were included in the same data set as women born in 1951. The first figure shows clearly that women born in 1951 have another phenotype than women born in 1913. If there are differences in the age distribution between the groups with different milk-consumption in Michaëlsson et al. (2014), then this will complicate the statistical analysis.
The second figure illustrates the statistical challenge due to the wide span in the age of women entering the study. The blue line shows the share of all women that lives born in 1915 at different ages, and the red line the same for women born in 1955 in Sweden. They are based on data from Statistics Sweden that are closest to the birth-year for the oldest and youngest women in the study, respectively. The X-marks are for women with the age of 74 when entering the study, and the age of 39, respectively. For women born in 1915, 61.2% are still alive at the age of 74. The study endured for around 20 years. At the age of 94, about 10% were still alive: i.e. 16% of the ones alive at the age of 74. Here the fraction that died during the study was 1 – 0.16 = 0.84.
At the age of 39, 96.2% of all women born in 1955 are still alive, and 20 years later, at the age of 59, around 93% are alive: i.e. 97% of the ones alive at the age of 39. Here the fraction that died during the study was 1 – 0.97= 0.03. If there are differences in the age distribution between the groups with different milk-consumption in Michaëlsson et al. (2014), then this will further complicate the statistical analysis.
From my experiences from analysis of biological systems in agricultural sciences I conclude that some more efforts are needed before we know how Michaëlsson et al. (2014) should be interpreted.
And with the comments of the authors as I have discussed above my final conclusion is that there is a need to further examine to what degree the material and methods of the original article informs of impacts to expect on mortality and fractures of different levels of milk consumption. With that I hold the question open that the authors may have made a valuable contribution. Maybe they have, maybe not. So I end with a question to the authors:
If I specify underlying data that I would like to get access from you in order to further probe the relevance of the statistical processing of data, generating the results presented in the original article, will you support me with that?
References:
Johansson, S., Villamor, E., Altman, M., Edstedt Bonamy, A-K; Granath, F., Cnattingius, S. 2014. Maternal overweight and obesity in early pregnancy and risk of infant mortality: a population based cohort study in Sweden. BMJ 2014;349:g6572.
Competing interests: No competing interests
In the comment of Karl Michaëlsson, Alicja Wolk and Liisa Byberg dated 4th of December 2014, they state the following.
“We note with interest that these authors also reference the incorrect calculations of crude mortality made by Staffan Hellstrand, who is consultant for the Federation of Swedish Farmers, an organization who is owner of milk industries (e.g., Arla Foods). In a previous response (http://www.bmj.com/content/349/bmj.g6015/rr/779932), we have corrected Staffan Hellstrand and guided him to the correct data for use. Despite erroneous figures, Astrup and Givens have chosen to repeat the presentation of Staffan Hellstrand’s incorrect calculations – an interesting action given the fact that Arne Astrup should be an objective key opinion leader in field of nutritional research by his role as editor of the American Journal of Clinical Nutrition. Hopefully, his action to criticize our study by distorted arguments is not related to his financial conflicts of interest, for example research support from Arla Foods (25).”
First, the most well-known Swede with the name Staffan Hellstrand is a song-writer. My friends know that neither singing nor lyrics is my fields of competence.
My name is Stefan Hellstrand, nothing else.
Second, I am not a consultant for the Federation of Swedish Farmers. It is correct that I have made some jobs for them, but I have made jobs for numerous of organisations, at the homepage of my enterprise www.ekostrateg.se (in Swedish only) information about my field of expertise, organisations I have worked for, networks are obtained.
Third, The Federation of Swedish Farmers – LRF is to my knowledge not an owner of milk industries including Arla. The organisation of the farmers cooperatives have for many decades, close to one century, been divided in one branch dealing with agricultural politics which is LRF and before LRF RLF. The other branch with sub-branches had and has response for agricultural industries. My father, uncle and grandfather contributed in the creation and development of these organisations, which together with consumer cooperatives, labour unions, and other organisations empowering the people since the 1900th century was a substantial part of the explanation why Sweden together with Japan had the fastest growing economy in the world 1870 to 1970, (see SOU 1991:82 for detailed analysis).
Fourth, the authors have not in a previous response “corrected Staffan Hellstrand and guided him to the correct data for use.” This i will elaborate on further in a following commentary.
Fifth, the figure I presented in the first comment and the calculations made are not incorrect. I will show this in a following commentary.
The last two sentences in the quotation above from Michaëlsson et al. are the most disturbing ones. I read it as an effort to weaken the strength of the argumentation of Astrup and Givens in that part where they have made use of my previous comment by implying that I repeat what the dairy industry wants me to say, and that I do that in a rather amateurish way. If I have misunderstood Michaëlsson et al. in this part I apologise for that, and ask for a clarification about what they mean.
I was in 2008 contacted by a group of around 40 scientists internationally that planned a qualified scientific contribution regarding what a more sustainable agriculture globally can be. They asked me to be the man responsible for that part that treated animal production systems in a more sustainable agriculture globally, and co-responsible for the issue of how to measure sustainability performance in such complex systems that agricultural systems are given the combined biophysical and socioeconomic contexts and their variation in time and space. Both issues are challenging not only for me but for the scientific community. How comes that that proposal came to me and my little enterprise? Does it reflect that qualified scientists internationally appreciate my overall competence regarding agriculture, food production, milk production and human health in a sustainability context?
Prof David Pimentel Cornell University, on the behalf of the editorial committee, revived my proposal and summarised it with the word “excellent”. David is author of around 700 scientific papers and 20 books. This gave Hellstrand (2013).
From the content of Hellstrand (2006; 2013) it is obvious that I work independent of the Swedish Farmers Federation and the dairy industry. At the same time I acknowledge that they have an important task to secure the economic viability of Swedish farmers including milk producers in their delivery of one of the most important ecosystem services needed to secure human wealth, namely food, in accordance with principles for sustainability put forward by Odum (1989), OECD (2001), Millennium Ecosystem Assessment (MEA 2005), FAO (2006), UN Millennium Development Goals (UN 2008), TEEB (2010). It is a reason that food in Swedish is called “livs-medel”, that is means for life.
I regard myself as quite well informed about the systems and issues treated in the quotation in the beginning of this comment. Given the substantial discrepancies between the content of the quotation and my own understanding of these systems, my last question is to what degree a similar discrepancy is at hand in the article itself in the way it describes the investigated system and the investigated system itself, i.e. reality.
The article of Michaëlsson et al. exists in a context. Hellstrand (2013) shows that the way ruminant production systems, where milk production is the most important one, are treated on an operative level in scientific articles and policy-areas in Sweden and internationally is in conflict with known properties of concerned systems in the disciplines that have the competence of excellence. That creates a situation where these “contributions” are in conflict with Odum (1989), OECD (2001), Millennium Ecosystem Assessment (MEA 2005), FAO (2006), UN Millennium Development Goals (UN 2008), TEEB (2010). This has resulted in a situation where the strategy for bioenergy from IPCC (2012) may threaten global food security in 20 years (Hellstrand 2013).
Hellstrand (ibid.) shows that one part of this problem is official proposals for cost-recommendations for good health of people and the environment in Sweden. If there is substance in Michaëlsson et al., this shall be acknowledged. If their results mainly reflects weaknesses, and possible flaws it is equally important that this is shown, so that they do not in the prolongation contribute to cost recommendations that threaten global food security and sustainability.
References
FAO. 2006. Livestock’s long shadow, Rome.
Hellstrand, S. 2006. A Multi-Criteria Analysis of Sustainability Effects of Increasing Concentrate Intensity in Swedish Milk Production 1989-1999. Environment, Development and Sustainability, Kluwer Academic Publisher, Volume 8, Number 3, Pages: 351 – 373.
Hellstrand, S. 2013. Animal production in a sustainable agriculture. Environment, Development & Sustainability 15:4 999-1036.
IPCC. 2012. Renewable Energy Sources and Climate Change Mitigation. Special Report of the Intergovernmental Panel on Climate Change. Eds: Edenhofer, O., Pichs-Madruga, R: & Y. Sokona.
MEA. 2005. http://www.millenniumassessment.org/en/About.aspx. Accessed 6th of February 2009.
Odum, E.P. 1989. Ecology and Our Endangered Life-Support Systems. Massachusetts: Sinauer Associates, Inc., Publishers.
OECD. 2001. Policies to Enhance Sustainable Development. Meeting of the OECD Council at Ministerial level, 2001.
SOU. 1991. Drivkrafter för produktivitet och välstånd, SOU 1991:82.
TEEB. 2010. Mainstreaming the Economics of Nature: A Synthesis of the Approach, Conclusions and Recommendations of TEEB.
UN. 2008. UN Millennium Goals, http://www.un.org/millenniumgoals/poverty.shtml, accessed 2008-08-25.
Competing interests: No competing interests
May I thank the authors for serial responses to criticisms? An example worthy of emulation.
Second suggestion.
Would the authors AND their critics please produce a brief STATEMENT of AGREED RECOMMENDATIONS that might help the consumers to decide their own course of action in, Central Sweden and if possible, elsewhere?
Thank you
Competing interests: No competing interests
In her commentary, Danielle Lemay stresses the conceivable confounding effect of genetic backgrounds of those able to digest lactose compared with those who cannot. Unlike the situation in United States, we would like to emphasize that the lactase persistence mutation is particularly common in the populations of the Scandinavian countries. The proportion in Sweden at time of cohort entries is estimated to have been approximately 95% (1, 2). Given this high proportion and therefore a comparably homogenous population in regards to the lactase persistence mutation in our cohorts, we found it unlikely that the genetic mutation is an underlying explanation for our findings. In addition, we found a clear dose response relation between milk consumption and the outcomes. Thus, women who reported consumption of 3 glasses of milk or more per day had higher rate of death compared with women who had a consumption of 2 up to 3 glasses of milk per day. Moreover, we already in our Discussion (3) explained the conceivable strengths but also weaknesses with a Mendelian randomization study design based on genetic variation in lactase persistence. It is probably weak as an instrumental variable, it has possible pleiotropic effects, the dose-response relation with milk is not readily observed with such a design and neither is the consideration of type of dairy product consumed.
During the study period, there was a change in milk consumption in the society at large and this trend with decreasing milk consumption was also captured in our cohort, as Gert Doekes and Lutzen Portengen remark. Concurrently, there has been a shift towards higher consumption of fermented milk products and cheese both in the society and in our cohort. If these changes in dietary intake were related with new fractures or other morbidities during follow-up, we could have introduced bias by confounding or reverse causation. However, there was no such link and we believe that our data reflect changes observed in the society. By not taking these changes in to account in time-updated exposure analyses we would have more exposure misclassification which, in general, introduces a conservative bias. The differences in person-years and cases when comparing the time-updated analysis in Table 2 and the baseline analysis (from 1987) in supplementary Table H are due to censoring of participants with implausible energy intake at time of the second questionnaire. If, however, these were kept in the analyses, all estimates remain essentially identical (see attached Table).
In a closed cohort, the total mortality might be relatively higher after the baseline examination compared with that after a follow-up examination. This is because those with high risk of mortality, including the oldest, have died before the follow-up examination and because those participating in the follow-up examination are on average healthier than those who do not participate. We would like to thank Gert Doekes and Lutzen Portengen for drawing our attention to supplementary Table I where we mistakenly had included a shorter follow-up time. However, the crude mortality rate is not very much higher by use of the baseline in 1997 (13.5/1000 person-years at risk) compared with that after 1987 (12.7/1000 person-years at risk), likely due to reasons discussed above. We provide an updated appendix file where supplementary Table I and additionally the number of cases and person-years for some outcomes in supplementary Table D have been corrected.
We thank Ulf Emanuelson for his comment. In short, as already mentioned in our manuscript and in previous commentaries, a non-differential misclassification of the exposure – existent to some degree in almost any type of prospective cohort study – will normally render conservatively biased estimates. Therefore, our higher rates of death and fracture with high milk consumption are not likely explained by measurement error of the exposure.
Liisa Byberg
Karl Michaëlsson
References
1. Torniainen S, Hedelin M, Autio V, Rasinpera H, Balter KA, Klint A, et al. Lactase persistence, dietary intake of milk, and the risk for prostate cancer in Sweden and Finland. Cancer Epidemiol Biomarkers Prev. 2007;16(5):956-61.
2. Ji J, Sundquist J, Sundquist K. Lactose intolerance and risk of lung, breast and ovarian cancers: aetiological clues from a population-based study in Sweden. Br J Cancer. 2014.
3. Michaelsson K, Wolk A, Langenskiold S, Basu S, Warensjo Lemming E, Melhus H, et al. Milk intake and risk of mortality and fractures in women and men: cohort studies. BMJ. 2014;349:g6015.
Competing interests: No competing interests
It was with great interest I read the paper by Michaëlsson et al., but I cannot understand why one single paper creates such fuzz when meta analyses of multiple studies hasn’t found the same association?! I have browsed through the various comments and replies, but it doesn’t seem as if anyone has picked up what I consider one of the most problematic aspect with the study. The authors refer to a correlation of 0.7 between their food frequency questionnaire (FFQ) and the gold standard reference and that this is considered good. However, this means that less than 50% of the variation in actual consumption is explained by the FFQ. To me it seems highly unlikely that such a poor exposure classification, and especially one that was done some 10-20 years before the failure, is likely to yield any true associations at all and thus I’d treat the association that was found with a great deal of caution.
Competing interests: No competing interests
Dear Editor,
The most remarkable findings in the study of Michaelsson et al. (BMJ 2014; 349 doi: 10.1136) are the strong associations between milk consumption and mortality in the mammography cohort. High HR’s up to 1.9 were found in analyses using time-updated milk consumption data, combining consumption reported at baseline in 1987-1990 with later food frequency questionnaire data from 1997. To “investigate possible bias…. introduced by using time updated information” (Mat &Meth), the authors also calculated HRs based on baseline consumption data only (for the whole cohort and the complete study period; Supplement Table H), and in the subcohort with FFQ 1997 data, as baseline for the period 1997-2010 (Table I). The HRs for all outcomes in these analyses were much lower than those obtained using the time-updated exposure variables, and similar to or lower than HRs in the men’s cohort. For CV- and Cancer-specific mortality and fractures many of the HRs were in fact close to 1.0 and, with few exceptions, no longer significant.
The authors interpret the results of this sensitivity analyses as evidence that using data from the 1997 FFQ has resulted in lower measurement error and de-attenuation of risk estimates. However, using time-updated exposure variables is not without pitfalls, and we feel that there are strong indications that these sensitivity analyses actually confirm that using the updated information may have introduced (confounding) bias. By combining of data in Tables 1 and 2 and Supplementary tables H and I it is easy to show that:
1. There was a strong reduction in milk consumption over the study period
2. Mortality rates were likely lower in the second part of the study (from 1997 onwards) than in the first part.
Ad 1) The distribution of subjects or p-yrs over the four milk consumption levels in the whole cohort at baseline (N in Table 1; p-yrs in Table H), in the subcohort with FFQ data in 1997 (p-yrs, Table I) and in the final analyses (p-yrs, Table 2) reveals a substantial change in reported milk consumption during the study period (see attached table).
As expected, the relative size of the four baseline subpopulations corresponds quite well with the distribution of person-years in Table H, which confirms that there were no huge differences in crude mortality or other loss-to-follow-up between the groups. In contrast, Table I shows a strong shift to less milk consumption, with the original 25% and 9% in the highest consumption categories reduced to less than 8% and 2% respectively. Thus among the ~65% women with follow-up data, many had their p-yrs from the period before 1997 assigned to a higher consumption category than their p-yrs after 1997. The authors have noticed this strong shift in milk consumption during the study, but only mentioned it in a discussion of possible comorbidity-associated changes in milk consumption (Table J)
Ad 2) The raw mortality rate calculated over the full population and entire follow-up (12.6/1,000 p-yrs; see table 2) is higher than the mortality rate for the population that filled out the 1997 FFQ and was followed up from 1997-2010 (11.1/1,000 p-yrs, table I in the supplement). That is counter-intuitive because the average age over which these cohorts were followed up is expected to be approximately 7-10 years higher in the latter (sub-)cohort. Because the paper does not provide information separately for women that did not respond to the 1997 FFQ it is difficult to exclude the possibility that the difference is due to differential mortality among those that filled out or did not fill out the FFQ, but that seems somewhat unlikely because the difference would have to be quite large (and it would signal problems with using the time-updated exposure variable as well).
Obviously we cannot exclude the possibility that the reduced mortality over the second part of the follow-up is due to the lowered milk consumption, but the fact that the same reduction in mortality is also apparent within the lowest (reference) category of exposure makes that somewhat unlikely. In short, by using a time-varying exposure estimate that shows a strong decreasing trend over time the authors may have introduced potential confounding by parallel trends in mortality. Without proper adjustment for this potential confounding effect, the claim that the difference in effect estimates for analyses either using or discarding the 1997 exposure information is due to measurement error attenuation seems to be insufficiently founded.
Competing interests: No competing interests
Michaëlsson and colleagues [1] have reported a correlation between high milk intake and higher mortality. The higher mortality of fluid milk consumers may be unrelated to milk consumption, but may instead reflect the different genetic backgrounds of those able to digest lactose compared with those who cannot.
DNA mutations that led to the ability to drink milk into adulthood had such a profound affect on the survival of our ancestors that this allele rapidly became dominant. This happened repeatedly in independent populations [2-4]. Lactase persistence has exerted stronger selective pressure than any other known human gene [5].
It is well known that positive selection of a highly favorable mutation enables nearby deleterious mutations also to become more frequent [6]. This “hitchhiking” of deleterious mutations has even been observed near the lactase (LCT) gene [7]. As genome wide association studies point to thousands of DNA variants that contribute to cardiovascular disease [8] and osteoporosis [9], there is reason to believe that those who can consume lactose will have a different disease risk from those who cannot, independent of whether they actually drink milk. This may be especially true among Europeans, given the strong sweep of their 13,910*T allele [10].
In the study by Michaëlsson and colleagues [1], it is likely that a higher proportion of the study subjects drinking 0 or 1 glass of milk per day were lactose intolerant, compared with study subjects drinking more glasses of milk per day. It is also likely that the lactose intolerant subset of the study population consumed more cheese and fermented milk products. The presence of one or two lactase persistence alleles has been shown to have a dosage effect on lactase activity [11]. Therefore, it could even be argued that the presence of one or two lactase persistence alleles provides a dosage effect on both the ability to consume more glasses of milk per day and the exposure to more unfavorable genetic mutations. Thus, the genetic background of study subjects is a confounding variable that may explain the association between high milk intake and higher mortality.
Further study of the genetic backgrounds of those who can and cannot digest fluid milk is needed to assess disease risk independent of modern-day milk consumption. When those who can consume milk do not, the disease risk associated with their genetic heritage is unchanged, and it is possible they may make things worse by not drinking milk. Long-term prospective randomized trials are sorely needed. In the meantime, observational studies involving milk consumption need to consider the genetic backgrounds of their study subjects.
Danielle G. Lemay, PhD, Genome Center, University of California-Davis. Email: dglemay@ucdavis.edu
References
1. Michaelsson K, Wolk A, Langenskiold S, Basu S, Warensjo Lemming E, Melhus H, Byberg L: Milk intake and risk of mortality and fractures in women and men: cohort studies. BMJ 2014, 349:g6015.
2. Tishkoff SA, Reed FA, Ranciaro A, Voight BF, Babbitt CC, Silverman JS, Powell K, Mortensen HM, Hirbo JB, Osman M et al: Convergent adaptation of human lactase persistence in Africa and Europe. Nature genetics 2007, 39(1):31-40.
3. Enattah NS, Jensen TG, Nielsen M, Lewinski R, Kuokkanen M, Rasinpera H, El-Shanti H, Seo JK, Alifrangis M, Khalil IF et al: Independent introduction of two lactase-persistence alleles into human populations reflects different history of adaptation to milk culture. American journal of human genetics 2008, 82(1):57-72.
4. Jones BL, Raga TO, Liebert A, Zmarz P, Bekele E, Danielsen ET, Olsen AK, Bradman N, Troelsen JT, Swallow DM: Diversity of lactase persistence alleles in Ethiopia: signature of a soft selective sweep. American journal of human genetics 2013, 93(3):538-544.
5. Bersaglieri T, Sabeti PC, Patterson N, Vanderploeg T, Schaffner SF, Drake JA, Rhodes M, Reich DE, Hirschhorn JN: Genetic signatures of strong recent positive selection at the lactase gene. American journal of human genetics 2004, 74(6):1111-1120.
6. Hartfield M, Otto SP: Recombination and hitchhiking of deleterious alleles. Evolution; international journal of organic evolution 2011, 65(9):2421-2434.
7. Chun S, Fay JC: Evidence for hitchhiking of deleterious mutations within the human genome. PLoS genetics 2011, 7(8):e1002240.
8. Ndiaye NC, Azimi Nehzad M, El Shamieh S, Stathopoulou MG, Visvikis-Siest S: Cardiovascular diseases and genome-wide association studies. Clinica chimica acta; international journal of clinical chemistry 2011, 412(19-20):1697-1701.
9. Urano T, Inoue S: Genetics of osteoporosis. Biochemical and biophysical research communications 2014, 452(2):287-293.
10. Itan Y, Powell A, Beaumont MA, Burger J, Thomas MG: The origins of lactase persistence in Europe. PLoS computational biology 2009, 5(8):e1000491.
11. Flatz G: Gene-dosage effect on intestinal lactase activity demonstrated in vivo. American journal of human genetics 1984, 36(2):306-310.
Competing interests: Dr. Lemay receives funding from the California Dairy Research Foundation. She is also the Executive Editor of “SPLASH! milk science update”, the monthly e-newletter of the International Milk Genomics Consortium (http://milkgenomics.org/splash-newsletter).
Early in my pediatric career I became aware of the clinical harm from cows' milk and its broad and potent immunologic effects. The death from Hodgkin's disease of a dairy-farming friend prompted a somewhat whimsical study of geographic correlations between dietary factors and lymphoma mortality (Lancet 2: 1184, 1976). From published figures of per caput protein consumption in 15 OECD countries, 1955-56, and age -adjusted death rates published by WHO I was surprised to find a strong correlation between milk/beef consumption and aggregate mortality rate for lymphosarcoma, reticulosarcoma and Hodgkin's disease, 1963-65: r=0.79 (p<0.0005). I theorized that the association was biologically plausible based on immunologic stress and/or exposure to oncogenic viruses in cows. This little paper has had a respectable citation history and subsequent studies have indicated that the association could well be causal....Now we find from a cohort study that milk consumption increases the risk of all-cause mortality, particularly cancer mortality, in a dose-response manner....Interesting!
Competing interests: No competing interests
Re: Milk intake and risk of mortality and fractures in women and men: cohort studies
I read with great interest this puzzling milk study. The results contradict previous prospective cohort studies. This begs the question, what is truly behind the observed association in this alarming study?
Many inherent weakness of the study have been addressed by several commentators. However, none seem to have addressed milk's effect on inflammation and oxidative stress.
I ran a literature search at PubMed and extracted the results for randomized controlled trials. Eleven RCTs reporting outcomes on inflammation and oxidative stress were found (1-11). Picture emerging from these randomized studies is in contrast to what Michaelsson et al. reported. Milk seems to have neither adverse nor beneficial effect on the parameters of inflammation or oxidative stress, when compared to active control, ie. soy milk, soy protein or red meat (table enclosed). Indeed, in one RCT milk improved markers of oxidative stress against soy protein (7).
It can be concluded that milk seems to be neutral in terms of inflammation and oxidative stress in isocaloric subsitution trials lasting 4-16 weeks. In other words, randomized trials have not confirmed observed pro-inflammatory and pro-oxidative effect of milk in the reported Swedish cohorts.
It is of interest that authors do not address effects of milk on cancer in the meta-analyses of cohort studies. World Cancer Research Fund (WCRF) constantly updates the meta-analyses on diet and cancer. WCRF has concluded on basis of its recent meta-analysis that milk and calcium are likely protective against colorectal cancer (strength of evidence: probable) (12)
According meta-analyses of WCRF milk is not associated with breast cancer (13), albeit inverse association of dairy has been reported in an other meta-analysis (14). Furthermore, WCRF concludes that association of milk and prostate cancer is minor or non-existing (15). If anything, previous meta-analyses suggest that milk should protective against cancer in women. Once again a finding contradicting the results of Swedish researchers.
In addition, a recent systematic review on milk and osteoporotic fractures failed to find one single study associating milk and fractures. Actually two studies showed protective effect of milk on fracture incidence (16).
Randomized studies in humans regarding inflammation and oxidative stress and meta-analyses of prospective cohorts contradict the observed effects in the Swedish milk study. Further studies are urgently warranted and caution should be exercised when interpreting results of this cohort study.
1) Drouin-Chartier J-P, et al. Impact of milk consumption on cardiometabolic risk in postmenopausal women with abdominal obesity. Nutrition Journal. 2015;14:12.
2) Serra M et al. . Effects of 28 days of dairy or soy ingestion on skeletal markers of inflammation and proteolysis in post-menopausal women. Nutr Health. 2012 Apr;21(2):117-30. doi: 10.1177/0260106012467243.
3) Rebholz CM et al. . Effect of soybean protein on novel cardiovascular disease risk factors: a randomized controlled trial. Eur J Clin Nutr. 2013 Jan;67(1):58-63.
4) Miraghajani M et al. Soy Milk Consumption, Inflammation, Coagulation, and Oxidative Stress Among Type 2 Diabetic Patients With Nephropathy. Diabetes Care. 2012;35(10):1981-1985
5) Napora JK, et al. Dose Isoflavones do not improve Metabolic and Inflammatory Parameters in Androgen Deprived Men with Prostate Cancer. Journal of andrology. 2011;32(1):40-48.
6) Beavers KM et al. Soy and the exercise-induced inflammatory response in postmenopausal women. Appl Physiol Nutr Metab. 2010 Jun;35(3):261-9.
7) Zemel MB, et al. Effects of dairy compared with soy on oxidative and inflammatory stress in overweight and obese subjects. Am J Clin Nutr. 2010 Jan;91(1):16-22
Beavers KM et al. Soymilk supplementation does not alter plasma markers of inflammation and oxidative stress in postmenopausal women. Nutr Res. 2009 Sep;29(9):616-22.
8) Charles C, et al. Effects of High-Dose Isoflavones on Metabolic and Inflammatory Markers in Healthy Postmenopausal Women. Menopause (New York, NY). 2009;16(2):395-400
9) Greany KA, e tal. Consumption of isoflavone-rich soy protein does not alter homocysteine or markers of inflammation in postmenopausal women. Eur J Clin Nutr. 2008 Dec;62(12):1419-25.
10 ) Fanti P, et al. Positive effect of dietary soy in ESRD patients with systemic inflammation--correlation between blood levels of the soy isoflavones and the acute-phase reactants. Nephrol Dial Transplant. 2006 Aug;21(8):2239-46.
11) Bryan-Borchers T et al. Beerman KA. Soy isoflavones modulate immune function in healthy postmenopausal women. Am J Clin Nutr. 2006 May;83(5):1118-25
12) http://www.wcrf.org/int/research-we-fund/continuous-update-project-findi...
13) http://www.wcrf.org/int/research-we-fund/continuous-update-project-findi...
14) Dong J et al. Dairy consumption and risk of breast cancer: a meta-analysis of prospective cohort studies. Breast Cancer Res Treat. 2011;127(1):23-31
15) http://www.wcrf.org/int/research-we-fund/continuous-update-project-findi...
16) Bolland Mark J, Leung William, Tai Vicky, Bastin Sonja, Gamble Greg D, Grey Andrew et al. Calcium intake and risk of fracture: systematic review BMJ 2015; 351 :h4580. (Appendix 3, table D)
Competing interests: Lectures and expert articles for companies /organisations which have commericial interest in dairy products.