Meat Alternatives Associated with Lower Hip Fracture Risk in AHS-2

Today was a good day: As longtime readers of this blog will know, I love getting good news about my “bad” habits. 🙂

A study came out that fits in well with past findings, is well-written, and supports my proclivity to eat and recommend soy meats!

It was a report from Adventist Health Study-2 (AHS-2), conducted among caucasian Seventh-day Adventists living in the USA, comparing the intakes of many different foods and hip fracture risk after 5 years of follow-up (1).

First, there was a bit of bad news in that vegans had the highest rates of hip fracture at 3.0 per 1,000 person-years compared to 2.0 for non-vegan vegetarians (including semi-vegetarians) and 1.6 for non-vegetarians. The paper didn’t report whether this finding was statistically significant, but it was not a great trend, obviously. Read on for how to reduce your chances.

Here are findings from the fully adjusted model (2):

– Meat alternatives once a day or more (compared to less than once per week) were associated with a 66% reduced risk of hip fracture in the vegetarians (.34, .12-.95).

– Eating legumes once a day or more (compared to less than once per week) was associated with an 82% reduced risk in non-vegetarians (.18, .06-.54) and an 55% reduced risk in vegetarians (.45, .22-.94).

– Meat more than 3 times per week was associated with a 45% reduced risk in the non-vegetarians (.55, .36-.83), compared to less than once per week.

– Dairy, nuts, soy milk, and “tofu & soy cheese” were not associated with a lower risk. The average amount of tofu & soy cheese per day was only .1 among the vegetarians. They did drink close to a cup of soymilk per day (or “soya milk” as the authors, who apparently think they live in Europe, call it).

The authors emphasize the need for vegans to eat legumes in order to get enough of the essential amino acid lysine:

“[A]n individual who adheres to a vegan diet, which excludes meat and dairy products, will need at least two cups of cooked beans [per day] to meet the recommended lysine intake requirement…Lysine and hydroxylysine are the main amino acids in the cross-linking process of bone collagen…Lysine can also influence bone health through its end product carnitine. Carnitine supplements have been shown to improve bone density in some animal and human studies.”

Note that the requirement for 2 cups of cooked beans assumes no other lysine sources in the diet, which isn’t the case. You can can read more on lysine needs and how to meet them in Protein.

They go on to say:

“Among our participants, intake of meat analogues of at least one serving daily reduced the risk of hip fracture by up to 49%…The main protein ingredients in meat analogues are soya, wheat, gluten, eggs and milk. A typical serving of meat analogues (1 serving ~73 g in AHS-2) contains at least 10 g of protein, but can vary from 9 to 18 g.”

Or 30 g as Tofurky Italian sausage has!

They point out that their finding for meat being protective is backed up by other research but not all. After a brief analysis of the research they suggest that meat intake is associated with bone health when protein intakes are low. And, finally:

“Protein is recognized for its ability to improve [calcium] balance, suppress parathyroid hormone, increase lean body mass and increase production of the bone growth regulator insulin-like growth factor-1.”

This study had something for everyone – low-fat proponents can relish in the findings for legumes and meat alternative eaters can smugly continue in their bad habits!

I have not yet updated the article, Calcium and Vitamin D, with this new study yet, but hope to do so in the next few days.


Please share and/or like my posts! Thanks!

I greatly appreciate donations of any amount and it allows me to spend more time on nutrition (click here). Gift Cards – E-mail Delivery

Vegan for Life: Everything You Need to Know to Be Healthy and Fit on a Plant-Based Diet from


1. . Lousuebsakul-Matthews V, Thorpe DL, Knutsen R, Beeson WL, Fraser GE, Knutsen SF. Legumes and meat analogues consumption are associated with hip fracture risk independently of meat intake among Caucasian men and women: the Adventist Health Study-2. Public Health Nutr. 2013 Oct 8:1-11. [Epub ahead of print] | link

2. *Adjusted for fruits and vegetables intake, age, height, weight, gender, energy intake, physical activity, smoking, health status and total calcium intake.

45 Responses to “Meat Alternatives Associated with Lower Hip Fracture Risk in AHS-2”

  1. Eric Brooks Says:

    One passage of this is confusing – two cups of cooked legumes, how often?:

    “An individual who adheres to a vegan diet, which excludes meat and dairy products, will need at least two cups of cooked beans to meet the recommended lysine intake requirement…Lysine and hydroxylysine are the main amino acids in the cross-linking process of bone collagen…Lysine can also influence bone health through its end product carnitine. Carnitine supplements have been shown to improve bone density in some animal and human studies.”

  2. Jack Norris RD Says:

    Eric, A day.

  3. ghul Says:

    hi jack interesting study, one more question do you think wheat products could reduce the risk of hip fractures as well?

  4. Jack Norris RD Says:


    They included wheat products in the meat alternatives, but I don’t know what percentage came from wheat. Seitan (and other wheat gluten meats) is high in lysine, but it is very low in tryptophan. My sense would be that vegans get enough tryptophan from other foods that it wouldn’t matter much.

  5. Eric Brooks Says:

    Ok. One more question. Do legumes that have had their hulls removed, like red lentils, have less lysine or are they ok to fill the requirement?

  6. Jack Norris RD Says:


    No idea. I use the amounts given in the USDA database and they don’t mention red lentils or hulled lentils.

  7. Richard Says:

    Hi Jack,

    First of all, thanks to you and Ginny for Vegan For Life—because of the great information in that book I’ve been a vegan for nearly a year!

    My mom is huge into animal rights and is considering going vegan herself, but I have to admit that this study scared her (and me) a little. I tried to explain to her that, as best I understand the data, the risk of hip fracture is pretty low for everybody, vegan or not, and that we can make that risk even lower by eating 2 cups of beans a day. However, since I have practically no science background, I’m not sure I’m summarizing this study accurately.

    Would it be better, for example, to eat 2 cups of beans today *and* meat imitations as well? My mom’s a fairly light eater, and the thought of 2 cups of beans a day is a little intimidating to her, to say nothing of adding meat imitations. (Even though she’s omni, she eats meat only rarely, and always has, so going from that to daily consumption of soy meats would be a big change for her, to say nothing of having to eat more beans.)

    She and I also had the thought that the 2 cups/day recommendation is a baseline, that depending on age and gender one might do completely fine to eat less than 2 cups. How does this general recommendation work when applied to various age and gender population subsets?

    Thank you so much—this information will be a great relief to her and to me as well.

  8. Jack Norris RD Says:


    I should have been more careful to point out that the authors assumed no other source of lysine in the diet but the 2 cups of cooked beans. In fact, I have adjusted the article above. A serving of cooked beans is usually considered to be 1/2 cup (not 1 full cup) and I recommend 3-4 servings of high lysine foods a day (which includes legumes but also 1 cup of quinoa). See my recommendations here:

    If your mom is an especially light eater, then if she likes Tofurky sausages, that is about the best bang for the buck for vegan protein that I’ve seen. It might even rival straight protein powder in terms of volume, especially since the powder would need to be mixed with a liquid.

  9. Jason Harrison Says:

    I thought that Insulin-like growth factor-1 was associated with cancer. Not as a carcinogen, but as a “might help cancer grow”. My understanding is based on my readings at YMMV.

  10. Jack Norris RD Says:


    IGF-1 has been associated with cancer. I have not done a full literature review on it, but from what I can tell, the evidence is inconclusive. It would be interesting to see if, in AHS-2, there is any association between meat alternatives and cancer. We might have to wait awhile for such results.

  11. Health Says:

    Was BMI adjusted for in this study? A higher BMI may protect against hip fracture, and as we know in the AHS-2 lacto-ovo vegeatrains had a higher BMI than vegans, and meat eaters had a higher BMI than lacto-ovo vegetarians.

  12. Jack Norris RD Says:


    The results were adjusted for weight.

  13. Richard Says:


    Thank you so much for the information. It’s going to be a big help to my mom and me.

    I’m still a little unclear on the big picture message to take away from this study: Should vegans be especially worried about hip fracture? And if so, is eating the proper protein/lysine you talk about sufficient to remove most of the hip fracture risk? 3.0 fractures per 1000 person-years sounds to me like 3 out of 10 people living 100 years as vegan getting a hip fracture, which sounds like a lot. Of course, despite looking it up on Google I can’t claim to know exactly what a person-year is, so I’m sure that’s not at all an accurate picture.

    Anyway, I don’t mean to belabor the point but since my mom has a legitimate concern about doing anything that would raise her risk of hip fracture, including going vegan, I’d love to be able to tell her that she could undo most of the risk (89% of it?) by eating plenty of lysine-rich protein. Is that the ultimate take-away here?

  14. Jack Norris RD Says:

    Let’s see, the study was for people 30 and above, and they normally remove people after age 90, though among SDA’s, the lifespan is closer to 80. So, I’d say that’s 50 years per person, not 100. At that rate being vegan would roughly raise your chance of risk from 1 in 13 to 1 in 7. I would say that if your mom is getting enough calcium and enough protein, her risk from being vegan would be ameliorated.

    Note that the comparison wasn’t between 2 cups of legumes per day, but rather it was between eating 1 serving per day with eating less than 1 per week. That said, I’d still recommend 3-4 servings of high lysine foods per day.

  15. Richard Says:

    Thank you, Jack, that has really helped me understand the study a lot better. I’m sure my mom will really appreciate it as well.

  16. Tyler Says:

    The title of this post and using this study to motivate the consumption of mock meats seems a bit strange to me. The study shows that the consumption of legumes reduces increases risk much more, shouldn’t that be the focus?

    Also, with mock meats, there is a pretty big range in healthfulness. What sort of mock meats were being consumed in this study?

  17. Jack Norris RD Says:


    This conversation has barely begun and I’ve already had enough of it. 🙂

  18. Tyler Says:

    “This conversation has barely begun and I’ve already had enough of it.”

    Alright…..I won’t disrupt your shilling any further.

  19. Dave Says:

    Few corrections and comments:
    1) The fracture rate for for non-vegan vegetarians was 2.0 (not 2.2).
    2) The fracture rate were only adjusted for age.
    3) By my calculation, the difference between vegan and non-vegetarians is highly statistically significant.
    4) Daily consumption of of legumes in vegetarian was associated with 55% less fractures (not 89%)
    5) The highest category of meat consumption was more than 3 times per week or more (not 3 per day).

  20. Jack Norris RD Says:


    Thanks, I fixed the mistakes.

  21. Donna Barski Says:

    Dear Jack,

    Just wondering…why didn’t you answer Tyler’s well thought out and reasonable question?

  22. Jack Norris RD Says:


    Because I see that you are a real person and someone I have emailed previously, rather than a troll, I will answer you.

    Tyler has commented a lot on my blog posts and I have generally not found arguing with him to be worth my time.

    > Also, with mock meats, there is a pretty big range in healthfulness. What sort of mock meats were being consumed in this study?

    In my post above, I gave the information on mock meats provided by the authors of the study. They said, “The main protein ingredients in meat analogues are soya, wheat, gluten, eggs and milk. A typical serving of meat analogues (1 serving ~73 g in AHS-2) contains at least 10 g of protein, but can vary from 9 to 18 g.”

    Tyler’s other question was merely a rhetorical criticism and people can agree or disagree with him as they choose.

  23. Donna Barski Says:

    Dear Jack,

    Thank you for answering my question in regard to the post from Tyler. I do unerstand that there are people who participate in blogs for the sole purpose of stirring up controversy.

    I have always appreciated your efforts to present accurate health information, and to not sugar coat the results of nutrition reserch. When I read factual updates from you, I know that I am then able to adjust my family’s meals accordingly, so we can maintain our already good health on a plant based (vegan) diet.

    For the animals,


  24. Eric Brooks Says:

    One more question, on combining proteins.

    Many years ago I read (as I’m sure most have read) from myriad sources, that complimentary proteins do not need to be combined in the same meal to deliver complete protein in the diet.

    Taking this to heart I experimented with not combining complimentary proteins in each meal, and I immediately, and also consistently over many meals, noticed that every time I ate grain carbs without legumes, my hunger was simply not satiated. No matter how much grain based food I ate, I stayed hungry.

    The difference was so dramatic that I very quickly went back to combining complimentary proteins in every meal and the endless hunger problem dutifully went away as soon as I did this.

    My theory is that when we sit down for a meal our body registers what we eat, and continues to desire more food if the proteins are not balanced so that it can compensate for the protein lacking in a given meal. For all the body knows, there is now a shortage of a given protein and it is maintaining hunger to compensate for that assumption.

    So the question is, is this theory correct?

    If so we need to keep advising people to combine proteins in each meal.

    I personally now swear by doing so, and I am pretty certain I am better for it.

  25. Jack Norris RD Says:


    Seems reasonable.

  26. Jason Harrison Says:

    Whoa, “seems reasonable”? On what basis, of known and verified biology do we havr evidence for Eric’s theory of “protein consumption summation”?

    I’d be willing to entertain a model of gut microflora signalling amino acid consumption and/or amino acid needs for the microflora, and then their outputs signaling food drives or satiety. But I have no idea how quickly these communication channels work. So, those with a microflora population signalling a need for a broad spectrum of amino acids might experience a hunger drive to continue to consume until all microflora signal satiety.

    But that’s still leaving a lot of science to be nailed down.

    I would also note that hunger, satiety, and fullness are very poorly defined terms in this, and most discussion. My guess is that you will feel full on a diet of steamed potatoes (eat with salsa, kimchi, or ketchup if you want more flavour). Potatoes contain only the necessary amino acids (the ones your body can’t produce), water, fiber, and starch. No protein, no fat.

  27. Dave Says:

    Just a reminder folks, combining proteins is absolutely not necessary. Eat sensibly, meet the RDAs, but please to potential vegans: don’t worry, ever, about the given amino acid composition of a given meal. It’s just not necessary. Taking note of the total protein intake at a meal is a better number to start examining, if you want to talk protein. That said, I’m glad the above writer feels better with his new dietary adjustment, and yes, if you are having some issue, why not try it? The science is clear, however: your body pools its amino acids and stores its protein. As long as the protein is coming in, the specific timing does not matter.

    What has always made the vegan diet “work” for us has been a heavy emphasis on legumes. Legumes at every meal, basically. When you remove flesh from your diet and with it the protein, fat, and iron those foods contain, you need to replace it with something similar. Legumes and mock meats fit the bill perfectly. There are so many ways and forms that you can add them into your diet.

    I second the Tofurkey sausage recommendation as a good item to add in once in awhile. They are great. You can’t even start a soy isolate argument there!

  28. Dan Says:


    I’m curious – why do you recommend such a heavy reliance on legumes at each and every meal?

    I think just 1 or 2 meals per day should contain legumes, but not necessarily every meal. And one could have 3 servings of legumes in a single meal!

    It may be a contradiction to state that you don’t need to eat a complete protein source (i.e. all essential amino acids) at every meal and then turn around and say but yes, eat a complete protein source (called legumes) at every meal.

    I know I am just an n=1 anecdote, but in the morning, for breakfast, I consume nuts, seeds, grains in many different forms – but no legumes – and this keeps me full for a good 6 hours. Specifically a smoothie made with flaxseed, hemp protein powder, diverse nuts, almond milk, other seeds, cocoa nibs, berries, wheat germ, wheat bran, etc — this would be significantly lysine-restricted, and maybe (according to Jack’s comment above), tryptophan-restricted too, but it really gets me through without hunger, etc, until 2pm or even later, despite a midday workout as well!

    I am just curious, not criticizing you or others. I was under the impression that there were other sources of protein in the vegan diet than just legumes – not that I am knocking legumes (I do think they are great, and in addition to what you mentioned, also contain copious amounts of fiber and calcium). I just don’t see a need to eat legumes at every single meal. Are we now saying that legumes are the main (“majority”) source of protein in a well-done vegan diet? Grains, nuts, seeds, corn, peas, nut and seed butters, rice, specific non-starchy vegetables ….

  29. Jack Norris RD Says:


    Seitan seems to be unusual in that it has very little tryptophan. I spoke to a wheat gluten researcher who told me it gets destroyed in the process of making it. Wheat has more. I haven’t checked wheat bran or wheat germ.

    > Grains, nuts, seeds, corn, peas,

    Just so you know, peas are legumes and fairly high in protein.

    > Are we now saying that legumes are the main (“majority”) source of protein in a well-done vegan diet?

    Generally, I would say that, though it’s possible to plan a vegan diet without them.

  30. Dan Says:

    > Are we now saying that legumes are the main (“majority”) source of protein in a well-done vegan diet?

    >> Generally, I would say that, though it’s possible to plan a vegan diet without them.

    I don’t think there’s any reason to plan a vegan diet without legumes, as there are ways to get around the gas/bloating issue (as you point out in your book); I was more making the point that they need not be consumed at every meal (I think).

    Elsewhere you’ve mentioned that a lot of low-lysine foods can add up to a goodly amount of lysine in people who are consuming higher-calorie diets who exercise.

    The sheer diversity of legume products does make it easy for them to become the centerpiece protein source of a vegan diet though.

    >I spoke to a wheat gluten researcher

    Wow, talk about your microspecialization in nutrition research! 🙂

  31. Dave Says:

    Thank you, Dan.

    You’re right that it’s not necessary to go legume crazy to be healthy on a vegan diet. You could plan a healthy vegan diet with zero legumes, I believe.

    That said, we find that we eat legumes all the time. It seems that one way or another, a legume typically “appears” in our meals. Whether it’s hummus,or some soy product or peanut butter, green beans or peas, they’re popping up all the time. We love legumes because they are so satisfying; they really fill you up in a good way.

    I’m a large physically active person, and when I made the switch from many years of lacto-ovo to veganism about 5 years ago, I was concerned that the vegan diet would not work for a serious athlete, that I couldn’t build muscle, etc. Of course, we all know [now] how laughable that is, but it took Jack’s writing to convince me that the vegan diet was a sound idea. Athletes like Jurek and Danzig and so many others further dispel any illusions about the adequacy of the vegan diet for athletes. Anyway, at the time of transition, I saw legumes as an easy way to replace dairy and egg protein without going soy crazy or using powders, so that probably started our legume emphasis. That and our love of Indian food!

    My wife has had low iron as a vegan, but that is the only issue we’ve had. That’s another reason we like legumes, and we’ve also adopted Jack’s tips on not drinking tea/coffee with meals and to include a source of vitamin C. I might also add that calcium blocks iron absorption, so if one supplements calcium, don’t do it at the same time as an iron-heavy meal or an iron supplement.

    Conclusion: I believe that it’s easier to thrive on a vegan diet if one incorporates many varied sources of legumes into the diet. Not necessary, but has worked well for us.

  32. Dave Says:

    Oops, my above “soy crazy” comment makes no sense the way it’s written.

    Of course, soybeans are legumes; I was thinking about non-soy legumes as I was writing that sentence. Sorry!

  33. Dan Says:

    Legumes also lower HbA1c in type 2 diabetics – randomized trial by Jenkins et al in Arch Intern Med this past year.

    Given that there aren’t that many vegan food choices that are rich in protein (contradicting what I said before), versus say meat/eggs/dairy, it does make sense to eat lots of legumes.

  34. Dan Says:

    Speaking of “the endless hunger problem”, as Eric put it so eloquently above, I have found that adding alot of raw vegetables to each meal really helps me keep fuller, longer. It has also stopped me from snacking on nuts completely. The ones I have experimented with are green beans, spinach, broccoli, carrots, red cabbage – though not all of these at every meal, of course. It is undoubtedly the fiber load despite the very low caloric content that is keeping me full, and I’m sure I’m getting the benefit of all those lovely phytonutrients and vitamins as well. Foolish of me not to think of this earlier! (of course this doesn’t replace protein, but I do think it helps to keep me satiated longer between meals).

  35. Susie Says:

    Jack, love your blog and have your book. Have been a vegan for 2 years and appreciate any info to maintain good nutrition. Question—I take 1500 mg of lysine per day to make sure I get this. Good or bad idea? Thanks, Susie

  36. Jack Norris RD Says:

    Hi Susie,

    > I take 1500 mg of lysine per day to make sure I get this. Good or bad idea?

    My sense is that 1,500 mg of lysine is safe. But if you eat a typical vegan diet with plenty of legume products, then I might be inclined to cut the dosage down to more like 500 mg which would give you a buffer of protection but not be so much that it could cause any sort of amino acid imbalance. I don’t actually know if 1,500 could cause an imbalance but when I was in RD school they taught that supplementing with single amino acids could cause an imbalance and so I always try to err on the side of not taking any more than necessary. I have never looked into the research on this matter to see what the theory was based on – my memory is that it was based on studies in people getting isolated amino acids in their tube feedings which might not be applicable to people eating whole foods.

    Sorry – I hope I didn’t do more harm than good with this answer. 🙂

  37. Susie Says:

    Thanks for the quick response, Jack! Susie

  38. Susie Says:

    one more question…I know that soy is a legume. I eat a fair amount of Gardein and similar “chicken” products. So do these count as legumes? And does any nut count, or only peanuts? Thanks!!

    I should probably get out your book and look this up!

  39. Jack Norris RD Says:


    Soy products count as legumes. Soy meats are probably the highest plant foods in lysine that there are. Only peanuts count in the nut category.

  40. Dan Says:

    The only harmful amino acid supplement that I am aware of is L-arginine. Two studies — one in heart attack patients and the other in patients with peripheral vascular disease — both showed very poor outcomes in patients randomized to L-arginine. I’d have to dig these out but they were published in JAMA and Circulation. This is why I dislike taking supplements that don’t have long-term safety outcome data – actually the above trials were 2 years or less.

  41. Dan Says:

    Hi Jack,
    I thought peanuts were a legume, not a nut.

  42. Jack Norris RD Says:


    That’s what I meant — peanuts are the only nuts that are legumes. I should probably cut and paste what I’m commenting on.

  43. Dan Says:


    >>That’s what I meant — peanuts are the only nuts that are legumes.

    I realized after that that’s what you meant, but didn’t want anyone to take what you said out of context (“Only peanuts count in the nut category.”), thereby depriving themselves of this valuable legume.

  44. Susie Says:

    Dan, interesting comment about arginine. I am in wound healing, and my lit review on arginine is that it is a precursor to nitric oxide, which as you probably know, is a potent vasodilator. There are multiple studies showing supplementation of L-arginine (up to 3 grams per day) improves both coronary and peripheral blood flow, decreases oxidative stress in the vessels, and improves overall circulation in patients with cardiomyopathy. We supplement arginine in patients with peripheral arterial disease and non healing wounds in conjunction with peripheral pumps which stimulate the release of NO from vessels. So I am interested in any data that says that arginine is damaging.

  45. Dan Says:

    Hi Susie,

    Here are the two I am aware of. There may be others that I don’t know about, as I don’t follow this field closely.

    JAMA. 2006 Jan 4;295(1):58-64. L-arginine therapy in acute myocardial infarction: the Vascular Interaction With Age in Myocardial Infarction (VINTAGE MI) randomized clinical trial. url:

    Circulation. 2007 Jul 10;116(2):188-95. Epub 2007 Jun 25.
    L-arginine supplementation in peripheral arterial disease: no benefit and possible harm. url:

Leave a Reply