Skip to main content


A Run of Contrary Results

From the outside, medical progress looks a lot easier than it really is. Well, I realize that’s true of a lot of things, but it’s especially true in progress against disease, and that’s especially especially true (as I’ve said here before) when you’re talking about dietary influences and what can be learned from them. Butter, whole milk, and other dairy products – bad for your cardiovascular health, right? Not according to this new observational study of 136,000 people across 21 countries. In fact, they found the opposite: higher dairy consumption was associated with a cardiovascular benefit. OK, salt consumption, at most any level, is associated with high blood pressure, right? Not according to this new study on 97,000 patients: the only correlation was seen with people who consumed over 5 grams per day of sodium (current guidelines sound the alarm at 2g and lower). Probiotics help to restore your gut flora after taking antibiotics, right? Not according to this new study on both mice and humans: treatment with a standard probiotic mix actually slowed down the reconstitution of the previous microbiome.

OK, dietary science is a mess – at least we can all agree on that. How about we shift over to a population of patients with a clearly defined disease? Omega-3 fatty acids are supposed to be good for cardiovascular health, right? And diabetic patients most definitely have a higher cardiovascular risk, so how about supplementing them? This new study looked at 15,480 diabetic patients without pre-existing cardiovascular trouble who got either omega-3 fatty acids or were given olive oil capsules as control. Over 7 years of observations – no significant difference between the two cohorts.

All right, enough dietary stuff in general. Let’s talk about giving specific compounds to patients and get some clarity, OK? Diclofenac is the most widely prescribed NSAID in the world, so it should be safe, right? Not according to this new study observing the entire population of Denmark over a twenty-year period. Unfortunately, the adverse event rate among those starting diclofenac went up by 50%, after controlling for everything else the authors could reasonably control for (take ibuprofen instead!) All right then, giving low-dose aspirin to older patients at risk for cardiovascular events is beneficial, right? Not according to this new study on 6200 moderate-risk patients (with 6200 in the control group), which found no meaningful differences in adverse events. And definitely not according to this new study with 9500 patients in each group – they also saw no benefit, but also report slightly higher mortality in the aspirin treatment group, apparently due to cancer.

Well now. I’m not sure if I’ve ever seen consensus medical advice take such a beating in such a short time (thanks to Eric Topol on Twitter for mentioning the whole range of these). I have no reason to celebrate any of these results, other than celebrating new and powerful data, which should always be welcome. But what all this does is illustrate just how tricky this whole business is. Aspirin! How long have we been recommending low-dose aspirin? And apparently not only recommending it to no avail, but quite possibly causing actual harm besides. Now, it may well be that in patients who have already experienced a cardiovascular event that low-dose aspirin may be worthwhile. But in the general population? Apparently not.

But I’m not just going to moan about how complicated everything is and how hard it is to get good answers. Both of those are true, big-time, but there’s actually something to celebrate here, odd as that may sound. What we’re seeing here is actual progress in human scientific knowledge, in all its messy infuriating glory. We know more than we did before. We’ve learned something. Even if you’re not sure about the situation in all of the cases mentioned above, what we’ve found is that you can (at the very least) run large well-executed trials and not get the results you’d expect, which is certainly something to think about. Science really does march on, and sometimes it steps on things along the way.

53 comments on “A Run of Contrary Results”

  1. Well, crep. So the extensive internal bleeding when I broke my ankle, while on a regiment of a baby aspirin a day, was for nothing?

  2. me says:

    The conclusion should not be “Science is so great, data is so great, let’s keep marching on! Wow isn’t this great? Oh, don’t take aspirin anymore, by the way haha.”

    It should be: “Something is fundamentally wrong with this area of science or its reporting. If this continues the general public will not trust experts’ opinions on anything.”

    In my opinion this all stems from the fundamental conflict of interest between good science and careerism. When only “exciting” results are publishable “publish or perish” will always lead to contradictory science and bad scientists.

    1. Derek Lowe says:

      I know what you’re saying. There’s a lot of stuff that gets over-reported, and there are a lot of temptations to overstate. But at the same time, some of these earlier recommendations were about as solid as they could be at the time. And note: we’re not only talking about stories in the press, but actual medical practice. The data were sufficient to convince physicians all over the world to recommend low-dose aspirin, with an expectation that the risk was low. Now the data are better. We have to be able to change our minds in these matters without being accused of bad faith:

    2. Hap says:

      I don’t think it’s so much the publication incentives (though they aren’t good either) but the tendency of people (doctors, dieticians, etc.) to focus on relatively simple biomarkers as panaceas for complex diseases or syndromes, and to hold on to them no matter what. If you act like your knowledge is absolute to get people to do what you think is best for them, and it turns out to be much more limited than you realize (or maybe care about), then people aren’t going to trust you, with reason.

      1. Jim Hartley says:

        In my innocence I once posted the question on Quora, “Should climate scientists be less certain?” I was flamed as an ignorant boobaramus, but I was actually pointing out that ex cathedra pronouncements on climate change have not been notably successful at influencing public opinion. Is there a middle ground, where we can promulgate today’s best thinking on scientific/medical issues and also acknowledge a degree of uncertainty?

        1. Hap says:

          I don’t know. I think the level of certainty you make in your pronouncements probably needs to fit the level of certainty you have in your data. and the likelihood that other data could overturn it. It also depends on honest engagement by the people you’re talking to or being heard by – people who don’t want to listen or want to prevent you from being understood likely can’t be stopped by anything you would say anyway.

          If all you have is a hammer and you bang away insistently with it, chances are that I shouldn’t trust you with delicate or complicated equipment.

        2. Anonymous says:

          I just want to contrast, in a simplified way, some of the nutrition studies vs climate studies. In the nutrition studies, they often try to study the effect of a single variable (precision NaCl; generic “fat” intake; etc.) or analyze things down to ID the single most important dietary variable (in a given study). In the big climate studies (i.e., IPCC), the models are getting more and more complicated.

          The nutrition models are often very simple 2 variable models: eat less salt and obtain a better health outcome … or maybe not, so it seems.

          Some climate models started out simple (CO2 at Mauna Loa vs Temperature) and have become extremely complex with more variables and more data. (CO2 levels fluctuate all over the globe.) More complex models can improve precision but more variables reduce accuracy. IMO, the “improved” models increase confusion and decrease confidence of non-scientists and fuel the arguments of climate science deniers. (I remember a symposium where two big climate researchers were arguing about the effect of PM 2.5 in the upper atmosphere: Forcing! NOT forcing! Forcing! … Kind of like Brown vs Winstein on carbonium ions: Classical! NON-classical! Classical! …)

          Back to food … as long as the studies are well done and honestly and transparently reported, I am OK with the shifting recommendations … or even the admission that there is no sound basis to make a recommendation. The USDA used to maintain a database of Oxygen Radical Absorbance Capacity (ORAC) of foods but ceased to do so in 2012.

          “In 2012 USDA’s Nutrient Data Laboratory (NDL) removed the USDA ORAC Database for Selected Foods from the NDL website due to mounting evidence that the values indicating antioxidant capacity have no relevance to the effects of specific bioactive compounds, including polyphenols on human health. … There is no evidence that the beneficial effects of polyphenol-rich foods can be attributed to the antioxidant properties of these foods. The data for antioxidant capacity of foods generated by in vitro (test-tube) methods cannot be extrapolated to in vivo (human) effects and the clinical trials to test benefits of dietary antioxidants have produced mixed results. We know now that antioxidant molecules in food have a wide range of functions, many of which are unrelated to the ability to absorb free radicals.” (complete source: www ars usda gov / northeast-area/ beltsville-md-bhnrc/ beltsville-human-nutrition-research-center/ nutrient-data-laboratory/ docs/ oxygen-radical-absorbance-capacity-orac-of-selected-foods-release-2-2010/ )

          When I was an undergrad, I was part of WHO funded diet study to help to determine the minimum daily requirement of a particular essential nutrient. (For the non-scientists, “essential” means that humans cannot make that nutrient in their own bodies; it can only be obtained from an outside source, typically the foods they eat.) The entire diet and its output 🙂 was totally controlled and measured. Consider that the result of testing a bunch of healthy 18-22 year old US undergrads would affect the global WHO MDR for all of humanity. I followed all the rules of the study but also consider that some participants cheated on a regular basis.

          Who knows what will come next? By the time the year 2173 comes around, we might have yet another completely different take on what constitutes a healthy diet and life style. I have linked the Sleeper clip in my name.

          1. Harrison says:

            Nutrition science has been a mess for a very long time. The individual responses to various diets is a problem (unless they are all applied to 18-22 year old university students). The way nutrition science is translated into public policy is inconsistent at best. Layer on top of that the fact everybody eats and is in a sense conducting n=1 experiments everyday leads to a false sense being informed (I eat therefore I understand what is good for me).

            The obesity epidemic in the US today may have its roots in an effort to prevent cardiovascular tissue. The diet-heart hypothesis and subsequent studies led to the recommendation of lowering saturated fat intake. This led to a cottage industry of low-fat “health foods” where the fat was replaced with sugar. Did this cause the sky-rocketing Type II diabetes rate? Perhaps. The road to hell is paved with good intentions.

          2. Imaging guy says:

            Actually, nutrition and climates studies are quite similar. When you do a observational and randomized study, you not only collect the data for your main independent variable (e.g. salt intake), you collect other important independent variables (like age, sex, smoking status, history of heart attack, and etc.) and multiple dependent variables (all-cause mortality, cardiovascular mortality, cancer mortality, morbidities and etc.). It would be foolish not to collect all these variables given the cost of these studies. Then you do the modelling (also called regression/multivariable analysis/machine learning) to control these variables for causal analysis and prediction. In this day and age thousands of independent variables from proteomics, RNA and DNA sequencing can be used to do the modelling with machine learning method. The problem is that there are million ways to model these data and you can get any outcome you want. Given the inconsistencies we see in medical studies which are much more rigorous, I have great doubt about claims of man-made global warming obtained from climate models.

          3. MoMo says:

            Political correctness killed ORAC. You cant tell food manufacturers that their foods lack antioxidants, they start crying and need a safe space to melt down in.
            You can thank the lobbyists, Democrats and Obama for that stratagem.

          4. Anonymous says:

            MoMo: “Political correctness killed ORAC. You cant tell food manufacturers that their foods lack antioxidants.” What? The ORAC lists were removed because “ORAC values are routinely misused by food and dietary supplement manufacturing companies to promote their products and by consumers to guide their food and dietary supplement choices.” There is near universal (except for the food and supplement companies) agreement about that reason. No one is saying that the removal of the lists magically removed antioxidants from the foods themselves. They can still sell their acai berry extracts and broccoli smoothies.

  3. Bob Seevers says:

    “Science really does march on, and sometimes it steps on things along the way.” Sciences as Sideshow Bob.

    1. MoMo says:

      I call BS on that Anonymous- That’s what they told you. The Smart know better.

  4. MrXYZ says:

    This may be an odd question but can anyone think of a reason that these larger studies (any of them) might be less trustworthy (useful might be a better term here) than the earlier smaller studies? I don’t mean in a cynical “let’s keep the sample size small so we might get lucky with our p value” way but in terms of actual conclusions. For example, more individuals in a study will improve the statistics but does the fact that larger populations tend to be more diverse (have more variation) confound the results? Or to put it in a different way, is there any reason to think these studies are better other than just their size?

    I guess I know what I’m reading this weekend.

    1. Old Timer says:

      I don’t have a statistical response to your question, but I’ve always suspected that the main problem with these dietary studies or guidelines stem almost entirely from population heterogeneity. If you get a large enough sample size, all the effects will go away. Sodium is bad for you!? Of all chemicals on Earth that animals have to deal with regularly, sodium has to be at the top of the list. Do people really think millions of years of evolution couldn’t solve the sodium clearance problem?? Obviously, there is a subset of people with one of the many NaCl clearance systems out of whack. The general population doesn’t have to worry about it. Epidemiology doesn’t do enough to help with these strange diet discoveries and subsequent guidelines.

      1. Kismet says:

        It seems inappropriate to involve evolution here. First of all, why would evolution care about a modest increase in your blood pressure that leads to a modest long term increase in cardiovascular mortality? Second, a small increase in blood pressure could have small benefits for near term survival totally offsetting the risks.

        We have lots of indirect evidence that modest hypotension is beneficial but one can construct an evolutionary sensible argument why humans nevertheless have higher blood pressure, which is harmful to long term health. A decrease in blood pressure probably increases the risk of orthostatic hypotension and the risks associated with hypovolemia, all much more important in the wild than cardiovascular long term health. Again, you may think my speculation is ridiculous. But in that case your evolutionary speculation would be equally so.

        In most cases, the word evolution is too broad to explain anything.

      2. Ian Malone says:

        On sodium in particular, like our predilection for sugar, we may well have needed to get more of it. Paleolithic dietary information is hard to come by, both because there aren’t any paleolithic people to study and because there are lots of woo vendors telling you what your ideal paleo diet is (on the assumption that this is the only thing you need to be as healthy as ancient humans were, and live to your thirties), but suggests about 0.7g/day sodium was consumed, gorillas may resort to eating decaying wood to get enough (unfortunately both paywalled). As such there’s no reason to think we’re adapted to managing excess. (And as mentioned by Kismet, evolution doesn’t necessarily optimise for long and healthy life. Another example is we don’t really know why ApoE-e4 has the prevalence it does, aside from Alzheimer’s risk it has several other downsides in the long run, it may confer some kind of advantage earlier in life.)

    2. gwern says:

      At least with aspirin, the increase in all-cause mortality & cancer is puzzling because previous aspirin RCTs had consistently shown lower all-cause mortality and lower cancer risks in particular. The most obvious difference that pops out as a possible moderator is the age: this study had considerably older subjects. So you could tell a story that aspirin’s anti-inflammatory/cancer properties are useful in younger people and the bleeding risk too minor to offset it, and then decades later, the cost-benefit has flipped.

    3. Eric C. says:

      Epidemiologists like to call this internal validity versus generalizability. If I keep my sample small and heterogeneous — say, teetotalling middle-class white non-overweight American male undergraduates at a single university with the free time to enroll in a two-week residential study — then I can control or carefully measure lots of possible confounding variables. Genetic variation will be lower, age is controlled, gender is controlled, socioeconomic class is restricted, I can measure what they eat and how much they sleep, and so on. So results will be cleaner.

      The problem then is generalizability — how much can I extend this to non-whites, women, older or younger people, drinkers, let alone the rest of the world population. And, depending on the research questions and how well we understand the mechanism of what we’re interested in, the answer may be Not Much.

      Conversely, I can do large studies with a varied population. Results — if I can be confident of them — will be generalizable to any large population of which this is a representative sample. But getting clean results will be a muddy labyrinth. Even if I try to measure the variables that vary in this population, the world is still much too complex. Does race correlate with exposure to air pollution? Or maybe it interacts with the effects of pollution. Do socioeconomic factors — which we can only measure poorly, anyway — correlate with how inaccurate a food diary will be, and in what ways? Does eating tofu reduce risk of stroke, or does being the sort of person who eats tofu reduce risk of stroke? (Or does it increase risk of suicide, giving you fewer years in which to have a stroke?)

      So, less complexity versus less generalizability, pick your poison.

      (Statistical properties are a slightly different question. A larger sample will always give us a better measure of how each quantity varies in the population, and that can’t be bad.)

  5. Uncle Al says:

    Grapefruit is the Donald Trump of ingestions. “Fruit consumed three days before the medicine can still have an effect.”

  6. electrochemist says:

    I am not surprised that a large, observational study based on filling out questionnaires about dietary habits over 15 years produced yet another set of contradictory conclusions. There are so many factors that could influence results that no study could control for all of them (even if there were all even known). I skimmed the article that Derek linked involving 136,000 participants. I didn’t see, for example, anything about controlling for lactose intolerance in the populations. That is merely one example of genetic heterogeneity that would result in some participants deliberately consuming less dairy products than other participants. There are undoubtedly other factors like this to consider when looking at such a complex set of outcomes and trying to correlate them to simple (minded) inputs.

    My concern: the general public usually doesn’t understand the difference between these observational studies designed to probe dietary cause and effect relationships, vs. placebo-controlled, double-blind and more rigorously measured clinical trials conducted by the pharma industry. This naturally leads to tremendous public skepticism when decades of “conventional wisdom” is reversed.

    1. Or, to put it another way, we’ve got the scientific equivalent of Gresham’s Law:

      Bad science drives out good science.

  7. Ben Zene says:

    I always get hung up on validity of meta analyses (even if well-controlled) versus a double blind placebo-controlled study.

    Regarding controlling for variables, I am reminded of a compound we worked on that displayed non-hepatic metabolism in rats to a major metabolic (presumably gut microbiome-mediated). Using the same species rat at another site, we saw significant differences in metabolite formation. So even in the highly controlled environments and subjects, unforeseen variables will rear their heads.

  8. ScientistSailor says:

    Dietary science is not a mess. Eat whole foods, mostly plants, not too much. Done.

    1. Red Agent says:

      Where’s your data on that?

    2. AR says:

      [Citation Needed]

  9. cynical1 says:

    Does this mean I can start smoking again?

    1. Druid says:

      see? it is really complicated. I am always amazed that some smokers are not harmed, or at least they live into their 90’s. My parents smoked until I left home (?) and lived into 90’s, but I sometimes need an inhaler ‘cos of 21 years of passive smoking. So, do you live alone? Do you have lucky genes? (eg CYP2D6 effective metabolizer) red hair? Did you gain weight when you stopped? How long do you want to live? Has stopping made you miserable? You won’t live for ever, whatever you give up, so your choice (your insurers might take another view).
      Perhaps the very successful use of drugs to normalize blood pressure in hypertensives has negated any benefits from reduced dairy or salt or prophylactic aspirin. Perhaps the early studies included many people at risk from childhood diseases which are less common. You can’t repeat experiments in biology because people change, so maybe it is best to heed only the recent studies if they are competent.

    2. myma says:

      No. You should squander you $5/day habit on starbucks instead.

  10. Jeff says:

    Dietary science is a mess — but the Food Babe is going to make it all better.

    (Ducking and running for cover)

  11. anon the II says:

    I remember when I was in high school and got nominated, along with a bunch of other people, for a small college scholarship. We were sitting there answering questions from a panel of stern old men trying to ascertain just how much booze and drugs we’d experimented with. When pressed, the guy next to me says “Well, I always say, ‘If it’s good to ya, it’s good for ya!'” Neither of us got the scholarship but for me, that little bit of wisdom has been as good a guide to what I imbibe as any stinkin’ study.

  12. Yvar says:

    Speaking specifically to one of these studies, my daughter has a medical condition that includes low blood pressure. She takes salt to raise it, and it is well known in this community of patients and doctors that grams of salt are needed to have any noticeable impact (she has a bottle of solid PBS capsules by her bed, since that amount of unbuffered salt can wreck your stomach!). It is therefore not surprising to me that other people’s blood pressure isn’t impacted by normal amounts of salt, but it is sad and bothersome that the general medical community doesn’t seem to pick up on the lessons learned by these smaller patient populations. Or at least not at the rate I’d like to see!

    1. AQR says:

      Is the disorder POTS (postural orthostatic tachycardia syndrome)? A member of my family has it and, in addition to an adrenergic agonist, consumption of excessive amounts (for the normal individual) of high salt foods was one of the necessary steps to control the symptoms.

      1. Yvar says:

        It is POTS. Salt definitely helps my daughter’s symptoms, but it has to be a lot (she hasn’t been able to get enough just with diet). Exercise is the other prescribed treatment, and it also seems to be helping when symptoms allow.

  13. Moi says:

    I’ve always believed that my Grandma nailed pretty much all outcomes medical studies on the benefits of otherwise of foods with the addage “A little bit of what you like does you good

    1. Peter S. Shenkin says:

      Well, she probably thought that a lot of what you like is bad for you.

      But at least we can say that grandma was half right.

  14. n says:

    FWIW: I have been fortunate to sit through the PURE presentations at both last years Euro Soc Card mtng, and the same only last month. It was pretty clear from some of the polls that many of the attendees don’t agree with the premise that saturated fats have a limited (or nil) impact on cv disease. Common agreement seems to be simple carbs aren’t that great for you (esp sugars), but the jury is still out re:sat fats with the camps being polarised. I suspect there is clear confirmation bias in both groups when viewing data. ps=Derek, your summary is spot-on!

    1. HFM says:

      The worst part about that field is that all the variables are confounded. There’s an obvious lifestyle cluster that leads to bad outcomes – the IDGAF diet (high in saturated fat, sugar, salt, protein, fried foods, etc, and low in vegetables, fiber, and various nutrients), plus substance use, little exercise, and poor self-care. You can pick out any one of those things, and even if you attempt to control for the others, you’ll get a “signal” that your favorite culprit is independently increasing risk.

      The other worst part is that humans are adaptable. You can literally eat nothing but ribeyes, or nothing but bananas, and your biochemistry will just roll its metaphorical eyes at you and then adapt. It can take decades for any adverse effects to show up. And no matter what you tell people when you’re attempting to do a diet intervention study, they’ll be back to their starting diet in six months, maximum. It’s a mess.

      My personal suspicion is that you can eat all the well-salted ribeyes you want, or all the sugar-coated bananas you want, but combining the two makes you fat and sick. I’ve been running that experiment for years (strict keto for neurological reasons). I’m watching with interest, but…yeah, the field is a mess.

  15. hn says:

    As both a scientist (chemist) and patient, I find these studies very frustrating. So am I supposed to keep taking my aspirin and fish oil? Are we wasting money on this kind of research that would be better invested in my lab?

    1. bks says:

      And what are conscientious primary care physicians to tell their patients?

  16. PI, i have it going on says:

    You are overthinking this buddy….just do your work….stop trying to be a “smart” pundit’s idiotic and you are setting a real bad example for the young….that somehow thinking ( not doing ) gets you there.

  17. pi, i prohibit freedom of speech in my lab says:

    Der boy needs to shove his own analyses up his a hole.

  18. Jay Jay says:

    If taking aspirin leads to more GI bleeding episodes, it would increase detection of cancers since these individuals are more likely to be examined w gastro-duodeno/colonoscopy. Also via screening for fecal occult blood as is widely done.
    Might not affect survival tough.

  19. TinMan says:

    “If one tortures a dataset long enough, it will confess to anything!” – Andrew Lo

  20. Watson says:

    Were the salt investigations conducted with Iodized or Kosher salt? Did they even control for that? Formulated as fine, or perhaps more like Baleine sea salt?

    Oh the many questions that could be asked for even the simplest study.

  21. MattF says:

    …and ibuprofen is bad for your kidneys.

  22. In Vivo Veritas says:

    And then there’s crap like this, which just exacerbates the existing credibility gap created when scientists oversell and pop media overimterprits scientific results…..

  23. dearieme says:

    I tend to assign many medical pronouncements into two classes. (i) Claims that invite the response “oh don’t be so silly” from the layman given to critical thought. I’d put the dietary fats and salt stuffs there. (ii) Claims that as a layman I am likely to have little ability to form an independent opinion of, for example the use of Diclofenac.

    The news that for the asymptomatic, aspirin might be even worse than statins makes me wonder if I should believe anything that cardiologists say about drugs. I do hope that, say, beta-blockers do more harm than good but should I be confident? (I do know that circumstances have recently been identified where they are a Bad Thing,)

    Should I pay any attention to the shifting definitions of Diabetes Type II, hypertension, or “normal” BMI? What should I do when I learn that the “overweight” and the “mildly obese” outlive the “normal”?

    1. Chris Phoenix says:

      If you think about the math, the BMI formula does not pass the “don’t be so silly” test.

      Your body mass is not proportional to the square of your height.

      Yes, it’s really that stupid.

      IIRC the entire Yale women’s crew team had a BMI of “obese” and this made the news soon after BMI started being promoted, and they kept using BMI.

  24. exGlaxoid says:

    Whatever the science says, people are living longer than ever before, so perhaps the invention of penicillin and other antibiotics, farming efficiency tools, generally better living conditions, and other items that science and engineering have provided was actually worth the side-effects of climate change, pollution, trace pesticides/chemicals in food, and other issues that I hear constantly claimed will kill us all. Not to say that we cannot improve on the past mistakes made, but so far, humans seem to be doing pretty well the last 100 years.

    1. anon says:

      Yeah, it might look like humans did well in the last century. It is probably true.
      But it is also completely unsustainable. The planet simply does not have enough resources to sustain the current global economy, let alone the further exponential (!!) growth mandated by the loan/interest driven monetary/economic system. Something’s gotta give sooner or later. Something’s GONNA give sooner or later.

    2. Ian Malone says:

      Definitely most people are healthier, there’s no reason to stop tweaking though.

      However you’ve lumped in two items that are much more of concern to our species than the others. Antibiotics were probably the biggest win, and the agricultural revolution. Sadly antibiotics seem to be on the way out, and climate change is a slow process that hasn’t really shown its teeth yet. Only the most hysterical would claim it will kill us all, but mass population shifts and/or reversing some of those agricultural gains will not be pleasant.

  25. JG4 says:

    “Omega-3 fatty acids are supposed to be good for cardiovascular health, right?”

    It may matter which ones and how much.

    Amarin fish oil capsule shows dramatic benefit for cardiovascular patients, potentially upending market

    see also:

    Bone-derived hormone reverses age-related memory loss in mice

    Human bones produces an important hormone as a result of load-bearing exercise. The work at Bone Health Technologies may produce a similar result.

Comments are closed.