Social media debates about the harmfulness of certain foods like eggs, refined carbohydrates, unprocessed meat, dairy, seed oils and so on can seem intractable. But I’d like to think that the main thrust of this essay is something we can all agree on. It boils down to this:   

 
Whole plant foods are not harmful (outside of allergic reactions, sensitivities and excessive consumption); in fact, they are in all likelihood beneficial for human health.  
 
By ‘whole plant foods’ I am referring to fruits, vegetables, wholegrains, legumes, nuts and seeds.   

Whole Plants Aren’t to Blame for Chronic Diseases 

Why am I so confident that whole plant foods don’t lead to bad health outcomes? Because the available evidence is crystal clear: whole plant foods are (almost) never significantly associated with diseases of affluence in observational studies. 


Moreover, the absence of damning evidence isn’t due to a lack of research. Many large, long prospective cohort studies have observed the apparent effects of consuming fruits, vegetables legumes and nuts – often over decades. The quality of these studies certainly varies considerably; but the consistency of the results is striking.  When studies are combined to create more powerful evidence; whole plant foods appear to be neutral at worst. The fact that we don’t find statistically significant correlations between whole plant foods and diseases of affluence, gives us very good grounds to conclude that they cannot conceivably be responsible for them. Which foods contribute to the development of cardiometabolic diseases continues to be debated – but it would be irrational to blame unrefined plant foods.  
 
Even more reassuring: unrefined plant foods very seldom lead to detrimental changes in surrogate markers of health in clinical trials. By surrogate markers I am referring to intermediate outcome measures such as LDL “bad” cholesterol, blood pressure, BMI and so on (as opposed to so-called “hard outcomes”, which refer to changes in the rates of morbidity and mortality). Indeed, the totality of the relevant scientific evidence indicates that under usual circumstances unrefined plants are harmless. In fact, I would say that the idea that they are actively detrimental to human health is very implasuible.  

Whole plant foods are good for health  

Observational studies don’t merely find unrefined plants to be free from positive associations with diseases; they are correlated with good health outcomes with remarkable consistency. Over and over again large, long observational studies link unrefined plant foods of all stripes with a reduced risk of developing diseases of affluence. And randomized controlled trials more often than not corroborate these findings: usually showing that the consumption of whole plants leads to improvements in validated surrogate markers of health compared to refined carbohydrates and animal products.  
 

Claims about increases (or a decreases) in risk are always relative or comparative. Oftentimes, randomised controlled trials are comparing the effects a food has on markers of risk to the effects of another food. But in observational studies, the changes in risk usually refers to the differences in associated diseases rates between different levels of consumption of a food. For instance, researchers will often compare the health outcomes between those eating the highest amounts of a given food against those consuming the lowest levels. But scientists sometimes apply substitution models (a type of statistical analysis) to observational data to estimate what might happen to our health if we replaced one food (or nutrients) for another. For instance, scientists have looked into the potential effects of replacing 1% of calories from trans fats with the same amount of energy from polyunsaturated fats (usually found in vegetables and fish). Substitution analysis helps to elucidate the potential benefits of swapping one kind of food for another.  
 
Substitution models indicate that whole plant sources of fat, protein and carbohydrate may lead to better outcomes compared to alternative sources of these macronutrients and rarely (if ever) show the converse. To the best of my knowledge, no substitution analysis has ever predicted that replacing an animal-based food with a whole plant food would lead to a statistically significant increase in risk. That’s not to say that all animal-based foods increase risk relative to any given whole plant food (fish for example appears to be very healthy). But it does suggest that – as a rule of thumb – whole plant foods are better choices.  
 
Since I’m mostly relying on observational studies to support my conclusions, I can hear my critics chanting “CORRELATION IS NOT CAUSATION!”. And of course, I without hesitation accept this as an incontrovertible fact. We cannot be absolutely certain that whole plant foods have a positive influence on health (note: we can’t be certain of anything in my opinion): there is always some room for doubt. But taking into consideration the consistency of the observational and experimental scientific data, it is eminently reasonable to infer that whole plant foods actively reduce our risk of cardiometabolic diseases compared to other commonly eaten foods. 
 
If you accept the above (and you absolutely should!) it does not entail that you must eat a whole-food plant-based diet to be healthy. There are many ways of devising a healthy eating plan; but eating generous amounts of whole plant foods is common to them all.  

Research in support of whole plant foods 

Below I have gathered scientific research as it pertains to categories of whole plant foods. I will focus mostly on observational studies. In other words, we will be exploring the long-term associations between the consumption of these foods and one’s risk of disease and death. 

Due to the relative scarcity of epidemiological research looking at the potential effects of specific fruits, vegetables, legumes, nuts and seeds; it would be inappropriate to try to estimate their effects on health. It’s safe to assume that some foods within these categories are healthier than others; but in the main it’s hard to draw any firm conclusions. Furthermore, there are good pragmatic reasons for eating a wide selection of whole plant foods – it helps to ensure that one gets a spectrum of vitamins and minerals.  

Wherever possible I have used systematic reviews and meta-analyses of prospective studies: these articles pool together many studies on a given research topic to create more powerful (and reliable) results. I want you to get a sense of just how consistently these broad categories of whole plant foods are associated with lower rates of diseases and premature death. It's possible to curate a bunch scientific papers to "support" any theory or idea, providing that one is suitable selective. But with whole plant foods one doesn't have to cherry-pick because the science on this topic unwaveringly favourable. 

Fruits and Vegetables

We have been told since childhood that fruit and veggies are good for health. The sweet taste of ripe fruit is appealing to most people, but vegetables are maligned by children (and some adults). Some parents find creative ways to conceal vegetables in order to bypass their children’s aversion to them. By the time we reach adulthood, most of us can eat vegetables without too much fuss. But is it really beneficial to acquire a taste for fruits and vegetables? 

The observational evidence suggests that the widely-held belief in the health-giving properties or fruits and (non-starchy) vegetables is justified. The relationships between these foods and disease have been extensively researched and statistically significant positive associations aren’t found. In fact, the research is especially favourable.  Meta-analyses of prospective studies link their consumption with a reduced risk of cardiovascular disease1-5, cancer1, type II diabetes6-8, and death from any cause1,9.  

What’s more, experimental trials seem to point in the same direction. A recent systematic review of RCTs concluded that consuming more than fruits and vegetables improves cardiovascular risk factors10. The benefits, in terms of favourable changes to markers of cardiovascular health, seem to peter out at about five servings10. However, a meta-analysis of 95 cohort studies found that ten servings (800g) was associated with the lowest risk of death1. In fact, not many individuals ate above this amount and thus we can hypothesize that eating more than ten servings may improve one’s risk even further!  
 
You may have noticed that the above comments were in reference to non-starchy vegetables. This is because much of the research on vegetables doesn’t include the starchy varieties. There is very little research on starchy vegetables aside from white potatoes. Unfortunately, white potatoes (the most commonly eaten and arguably most loved vegetable) is the only whole plant food which is fairly consistently associated with chronic disease – specifically diabetes type II. A 2018 meta-analysis of six large prospective cohort studies found that the highest total potato consumption (compared to the lowest) increased the risk of diabetes by 7.7%11. This result is seemingly mostly driven by French fries – the same study found that fried-potatoes increase one’s risk by a much larger 36.2% (again, comparing highest vs lowest).  
 
It stands to reason that French-fried potatoes may contribute to an increased risk of diabetes: they are more easily over-consumed; higher in calories; sometimes contain trans-fats. Importantly though, other forms of potato (such as baked, boiled, mashed) are independently associated with statistically significant increases in the risk of diabetes12-14. The consumption of one daily serving of boiled, mashed and baked potatoes may increase one’s risk of a diabetes by 9%. Furthermore, the Nurse’s Health Study cohort was used to look into the possible effects of potatoes on the risk of gestational diabetes (diabetes that occurs during pregnancy). The results were striking. After adjusting for many confounders (BMI, healthy eating, energy intake, exercise and more), they found that eating greater than one portion of (boiled/baked/mashed) potatoes a week was associated with an increased risk of developing diabetes during pregnancy. Consuming more than five portions a week may increase a woman’s risk of gestational diabetes by more than 50%13! Such large changes in risk are seldom seen in nutritional epidemiology, making this finding rather striking. 

However, I tend to think that these associations are spurious. The onset of diabetes is intimately related to weight gain and appears to be caused by the accumulation of fat in the liver and pancreas50. In my opinion, it is unlikely that potatoes – in and of themselves – promote weight gain: boiled, mashed or baked potatoes are very sating and difficult to overeat. I think that the link between potatoes and diabetes is probably due to the cooking method (when deep fried) and/or the types of food that they are routinely paired with (e.g., high fat sauces, fast food, butter other fried foods). 

Furthermore, the research on potatoes is not all bad. For instance, a meta-analysis of 95 cohort studies, found that they were linked to a lower risk of all-cause mortality1.  And another recent meta-analysis showed that potatoes aren’t associated with total- and cause-specific mortality15.  
 
To the best of my knowledge, no fruit is consistently associated with health problems. However, I would put whole coconuts (the unprocessed flesh) on hold for now since they are extraordinarily high in saturated fat (around 70% saturated fat by calorie). Certain types of saturated fat tend to increase the amount of “bad” cholesterol in our blood – a known risk factor for heart disease. According to my research, high-quality studies on whole coconuts do not exist – whether in terms of clinical trials or prospective cohort studies. Until we have good-quality research on this fruit, I think that caution dictates that it should be restricted. No other commonly eaten whole plant food, as far as I’m aware, contains such high amounts of saturated fat.  
 
Another concern: tinned fruit. For whatever reason, when fruits are canned, we do not see beneficial associations1,16. In fact, they may increase the risk of premature death1,17.  This may be at least partly related to the fact that these products often contain high amounts of added sugar. But for all we know, the canned fruit in juice (rather than syrup) may also be problematic since studies tend to lump them together. Thus, I’d avoid (or restrict) canned fruit for the time being.  
 
Some might worry about fruit juice since edible parts of fruit are removed in the juicing process creating a product which is very high in (natural) sugars. This position finds some support in pooled analysis of three large prospective cohort studies which found that fruit juice was associated with an increased risk of diabetes (whereas whole fruit was inversely associated with this disease)18. However, this study did not seem to distinguish between 100% fruit juice and sugar-sweetened juices or “juice drinks”. A 2020 meta-analysis of prospective observational studies and RCTs that expressly investigated the effects of 100% fruit juice, did not find that juice negatively impacted health markers (compared to placebo or a control drink) and statistically significant increases in disease risk were not found19. In fact, low levels of consumption associated with a lower risk of stroke and other cardiovascular events. All things considered; I don’t think 100% fruit juice should be renounced. However, I’d limit fruit juice to about one small glass per day or less and I’d give the same advice for vegetable juices, although they are very much understudied. 
 

Key points: 
 

  • Fruits and non-starchy vegetables are reliably associated with a reduced risk of chronic diseases.  
  • Higher amounts of fruits and vegetables may lower the risk of death (up to and possibly beyond 10 servings per day).   
  • 100% fruit and vegetable juices appear to be safe and are possibly healthful at low levels of consumption.  
  • Potatoes have been correlated with an increased risk of diabetes in some populations, but this association may be spurious. 
     

Exception: whole coconuts are very high in saturated fat and as such may pose a risk to cardiovascular health.  

Exception: canned fruit may increase one's risk of premature death.  
 

Legumes (including traditional soy products)

When I use the term ‘legume’, I am usually referring to pulses such as beans, peas, chickpeas, lentils and soy beans (as well as traditional products made from soy beans – tofu, soymilk, tempeh etc.). Peanuts technically fall into the botanical category of legume and are sometimes lumped together with pulses in scientific studies.  But I will not in the main address peanuts here since they are generally considered a culinary nut.   
 
Legumes aren’t usually eaten in high amounts in the West; but they are consumed as staples in Asian, Central American, South American, Middle Eastern and African countries. Legumes are an important part of plant-based diets because they are a good source of energy, nutrients and protein.  
 
The available observational data on legumes is very promising indeed: a 2019 comprehensive review (using 6 meta-analyses and 13 additional prospective cohort studies) found that legumes aren’t associated with an increases risk of cardiometabolic diseases. When comparing the highest versus lowest levels of consumption, legumes were associated with a roughly 10% decrease in the risk of cardiovascular disease, coronary heart disease, hypertension and obesity20. Furthermore, a 2017 meta-analysis found those eating the most legumes had a 7% reduced risk of death from all-causes21. There is also evidence indicating legumes may reduce one’s risk of prostate cancer22 and soy may reduce one’s risk of colorectal cancer23.  
 
When it comes to type II diabetes, total legume consumption appears to be in the main neutral (neither increasing nor decreasing risk)20,24. However, a recent meta-analysis found total legumes (beans, soy and peanuts) linked to a small increase in the risk of diabetes25. I think this is unlikely to reflect a genuine increase in risk, as the association was mostly driven by three European cohorts which consumed extremely small amounts of legumes (most individuals were eating a median intake of between 0-19g per day). It’s reassuring that an association was not found in those countries which enjoy legumes in much greater quantities (e.g., a Brazilian cohort consumed a median of 140g).  
 
What’s more, other meta-analyses have found sub-categories of legumes to be inversely associated with diabetes. For example, meta-analysis of four prospective cohort studies showed that non-soy legumes (beans, peas, lentils, chickpeas) may lower the risk of diabetes by 15%24. Further, a 2020 meta-analysis involving 15 cohorts with half a million people and more than 32k cases concluded that tofu (soybean curd) was linked with an 8% lower risk of diabetes26. Moreover, a large Singaporean prospective cohort study involving more than 43,000 people, found the consumption of unsweetened soy products inversely associated with diabetes: five servings per week was linked to a 28% lower risk27. A 2021 meta-analysis of RCTs indicated that soy milk (relative to dairy milk) consumption may help us to reduce blood pressure, total cholesterol, bad cholesterol and has beneficial effects on CRP (a marker of inflammation)28.
 
The observational evidence on traditional soy products more generally would suggest that they are health promoting: these foods are associated with good (or neutral) health outcomes in prospective observational studies. It is widely believed that soy exerts feminising effects on men due to its relatively high phytoestrogen content – but this is untrue. A 2020 meta-analysis of clinical trials confirmed that even high levels of soy consumption doesn’t impact reproductive hormones in men29.  

However, there is evidence that soy may moderately increase of thyroid stimulating hormone (TSH) without affecting levels of thyroid hormones (fT3 and fT4)30. This may be important for those with pre-existing thyroid issues, such as subclinical hypothyroidism and it may wise for individuals with thyroid abnormalities to have their thyroid levels checked regularly, if consuming a lot of soy.  
 
I do think it’s wise to be wary of most modern, ultra-processed soy products: these foods typically contain refined ingredients that are probably harmful. Consider the fact that, as far as I know, there’s no evidence indicating that soy milk per se poses a health risk; but sweetened soy milk may increase one’s risk of diabetes27. Similarly, soy-based meat and dairy alternatives (faux chicken, burgers, cheeses, yoghurts etc.) typically contain refined starches, sugars and sources of saturated fat; and as such they can’t be vouched for to the same extent as traditional soy foods such as: (unsweetened) soy milk, tofu, tempeh, miso and natto.  

 

Key points: 
 

  • The consumption of legumes (including traditional soy products) is inversely associated with various health problems.  
  • Traditional soy-based foods are good choices; but some caution is called for in relation to some of the modern, processed soy products. 

Nuts (and Seeds)

There is a lot of recent quality observational data on nut and it’s all ridiculously positive! As mentioned earlier, peanuts aren’t ‘true nuts’; but they are often lumped with tree nuts and appear to have similarly beneficial effects31.  
 
Meta-analysis of numerous studies with hundreds of thousands of individuals show that one serving of nuts per day (about 28g; or roughly ? of a cup) is associated with lower incidence of: 

  • Death from all causes (between 12-27% reduction31,32)  
  • CVD mortality (21- 39%31,32) 
  • 29% CHD31
  • 15% total cancer31  
  • 52% respiratory disease deaths31  
  • 39% diabetes31.
     

Further, other meta-analyses have found that those eating the highest amounts of nuts (compared to those eating the lowest amounts) enjoyed: 

  • 19% lower risk of ACM33,34 
  • 44% lower risk of total CVD33
  • 25-27% lower risk of CVD death33,34
  • 34% of all CHD33 
  • 27-30% lower risk of CHD mortality33,34
  • 47% lower risk of sudden cardiac death33
  • Stroke mortality (mainly in relation to ischaemic stroke )34
  • Cancer mortality 13%-14%32,34.

 

Some promoters of a whole-food plant-based approach, warn against the consumption of nuts on the basis that they are higher in saturated fat. Quite frankly, this does not make sense. The main concern around saturated fat is its ability to increase one’s risk of cardiovascular disease by raising LDL-C (so-called “bad” cholesterol). However, as you can see, nuts are linked to a statistically significant reduced risk of total cardiovascular health problems! What’s more, A 2018 network meta-analysis of RCTs ranked nuts as the best food group for LDL-C reduction35. In fact, nuts were considered the best for overall improvement in markers of metabolic health35. Thus, the advice to avoid nuts based on their potential to cause or exacerbate heart disease is utterly unsupported by the best available evidence.  
 

There is very little observational data on culinary seeds (e.g., pumpkin, sesame, sunflower, flax etc.). In fact, I could not find a single prospective cohort study which looked into the possible health impact of seeds. There is also very little experimental data on culinary seeds as a food group. But flax and sesame have been studied with respect to their influence on health markers and the results are promising36,37,38. 

Despite the scarcity of scientific data, I tend to think that culinary seeds are unlikely to be problematic for health: 1) seeds are high in unsaturated fats (which are not linked to health problems); 2) the limited experimental evidence is favourable; 3) they are botanically similar to nuts (which are well-studied and associated with good health outcomes)  
 

Key points: 
 

  • There is a considerable amount scientific data on nut consumption and the results overwhelmingly favourable.  
  • Prospective observational studies consistently link nut consumption to significantly lower rates of several diseases.  
  • Randomised controlled trials show that nuts are particularly effective at improving surrogate markers of cardiometabolic health.  
  • The science on the potential health effects of seeds is rather limited, but it seems fair to assume that they are relatively safe.  

Wholegrains

By definition unrefined grains (otherwise known as wholegrains) contain all of their intrinsic parts. That’s to say, unlike refined grains, wholegrains have not had the bran and germ mechanically removed. Refined grains are left with the most energy dense and nutritionally impoverished part of the plant (this is called the endosperm). Both cereal grains (like oats, rice, and wheat) and pseudo-cereals (such as quinoa and buckwheat) are classified as wholegrains as long as the bran and the germ haven’t been removed. Processed foods like bread and pasta which have been made from milled unrefined grains also fall into this category.   
 
While refined grains are often associated with poor health outcomes, wholegrains are emphatically not. In fact, I am not aware of a single prospective cohort study that links wholegrain consumption with a statistically significant increased risk of any disease! By that very fact, I think it is incredibly unlikely that wholegrains are responsible for chronic diseases.  
 
On the contrary, meta-analyses of prospective cohort studies show over and over again that the consumption of wholegrains is inversely associated with many diseases of affluence, including: cardiovascular disease5,39, including ischemic stroke40,41, obesity42, diabetes41,43,44, colorectal cancer41,45.  
 
Moreover, meta-analyses have found that wholegrains are associated with a lower risk of dying from:  

  • Any cause39,42,46,47,48  
  • CVD46,47,48  
  • Cancer39,47,48 
  • Respiratory disease39
  • Infectious diseases39.
     

Grains would appear to be a boon for health, as long as their more fibrous aspects are eaten along with the endosperm. The process of refining grains seems to radically transform a perfectly healthy ingredient into a potentially hazardous junk food. Epidemiological studies don’t in the main find total grains and refined grains associated with a reduced risk of disease and death. In fact, refined grains are at times associated with an increase in the risk of several diseases.  
 
Thus wholegrains – whether eaten intact or milled – appear to be safe and healthful. But while I unreservedly endorse wholegrains, I cannot – to the same extent – vouch for commercially-available products which contain them. Many industrialised ‘wholegrain’ products such as breakfast cereals, breads, snack bars, crackers and so on contain a mix of refined and unrefined flours. It seems highly plausible – given the dubious status of refined grains – to think that these ‘hybrid’ foods are less than ideal. What’s more, grain-based products often contain other undesirable ingredients such as free sugars and preservatives. Food manufacturers are incentivized to add refined flours and factory-made ingredients because they improve palatability and extend shelf life. Admittedly, some of the commercially available grain products are neither ultra-processed nor include refined flour; but sadly, these are exceptional. The best way to ensure that grain-based foods are made from 100% wholegrains and are free from suspect ingredients is to make them yourself. 
 
But most of the research on wholegrains include hybrid products and despite this they are unvaryingly associated with neutral or better health outcomes. Products which contain as little as 25% wholegrains may be classified as wholegrains in epidemiological studies49. This is speculation on my part, but it seems likely that the research may underestimate the benefits of wholegrains and 100% wholegrain products.  
 

Key points: 
 

  • There is a considerable amount epidemiological research on wholegrains and the results overwhelmingly suggests that they are good for us. 
  • Where possible choose 100% wholegrain products.  

 If you have any comments or criticisms let me know on Facebook or Twitter. 

[Updated: 10/05/2022]

                                                                                Back to Articles


References: 

 1. Aune, D., Giovannucci, E., Boffetta, P., Fadnes, L.T., Keum, N., Norat, T., Greenwood, D.C., Riboli, E., Vatten, L.J. and Tonstad, S., 2017. Fruit and vegetable intake and the risk of cardiovascular disease, total cancer and all-cause mortality—a systematic review and dose-response meta-analysis of prospective studies. International journal of epidemiology, 46(3), pp.1029-1056.  

2. Zhan, J., Liu, Y.J., Cai, L.B., Xu, F.R., Xie, T. and He, Q.Q., 2017. Fruit and vegetable consumption and risk of cardiovascular disease: A meta-analysis of prospective cohort studies. Critical reviews in food science and nutrition, 57(8), pp.1650-1663. 

3. Gan, Y., Tong, X., Li, L., Cao, S., Yin, X., Gao, C., Herath, C., Li, W., Jin, Z., Chen, Y. and Lu, Z., 2015. Consumption of fruit and vegetable and risk of coronary heart disease: a meta-analysis of prospective cohort studies. International journal of cardiology, 183, pp.129-137. 

4. Dauchet, L., Amouyel, P. and Dallongeville, J., 2005. Fruit and vegetable consumption and risk of stroke: a meta-analysis of cohort studies. Neurology, 65(8), pp.1193-1197. 

5.  Bechthold, A., Boeing, H., Schwedhelm, C., Hoffmann, G., Knüppel, S., Iqbal, K., De Henauw, S., Michels, N., Devleesschauwer, B., Schlesinger, S. and Schwingshackl, L., 2019. Food groups and risk of coronary heart disease, stroke and heart failure: a systematic review and dose-response meta-analysis of prospective studies. Critical reviews in food science and nutrition, 59(7), pp.1071-1090. 

6. Wang, P.Y., Fang, J.C., Gao, Z.H., Zhang, C. and Xie, S.Y., 2016. Higher intake of fruits, vegetables or their fiber reduces the risk of type 2 diabetes: A meta?analysis. Journal of diabetes investigation, 7(1), pp.56-69. 

7. Li, M., Fan, Y., Zhang, X., Hou, W. and Tang, Z., 2014. Fruit and vegetable intake and risk of type 2 diabetes mellitus: meta-analysis of prospective cohort studies. BMJ open, 4(11), p.e005497.

8.Halvorsen, R.E., Elvestad, M., Molin, M. and Aune, D., 2021. Fruit and vegetable consumption and the risk of type 2 diabetes: a systematic review and dose–response meta-analysis of prospective studies. BMJ Nutrition, Prevention & Health, 4(2).  

9. Wang, X., Ouyang, Y., Liu, J., Zhu, M., Zhao, G., Bao, W. and Hu, F.B., 2014. Fruit and vegetable consumption and mortality from all causes, cardiovascular disease, and cancer: systematic review and dose-response meta-analysis of prospective cohort studies. Bmj, 349.   

10. Toh, D.W.K., Koh, E.S. and Kim, J.E., 2020. Incorporating healthy dietary changes in addition to an increase in fruit and vegetable intake further improves the status of cardiovascular disease risk factors: A systematic review, meta-regression, and meta-analysis of randomized controlled trials. Nutrition reviews, 78(7), pp.532-545.  

11. Zhang, Y., Dingyun, Y.O.U., Nanjia, L.U., Donghui, D.U.A.N., Xiaoqi, F.E.N.G., Astell-Burt, T., Pan, Z.H.U., Liyuan, H.A.N., Shiwei, D.U.A.N. and Zuquan, Z.O.U., 2018. Potatoes consumption and risk of type 2 diabetes: A meta-analysis. Iranian journal of public health, 47(11), p.1627. 

12. Muraki, I., Rimm, E.B., Willett, W.C., Manson, J.E., Hu, F.B. and Sun, Q., 2016. Potato consumption and risk of type 2 diabetes: results from three prospective cohort studies. Diabetes care, 39(3), pp.376-384.

13. Bao, W., Tobias, D.K., Hu, F.B., Chavarro, J.E. and Zhang, C., 2016. Pre-pregnancy potato consumption and risk of gestational diabetes mellitus: prospective cohort study. bmj, 352.

14. Schwingshackl, L., Schwedhelm, C., Hoffmann, G. and Boeing, H., 2019. Potatoes and risk of chronic disease: A systematic review and dose–response meta-analysis. European journal of nutrition, 58(6), pp.2243-2251.. 

15. Wang, D.D., Li, Y., Bhupathiraju, S.N., Rosner, B.A., Sun, Q., Giovannucci, E.L., Rimm, E.B., Manson, J.E., Willett, W.C., Stampfer, M.J. and Hu, F.B., 2021. Fruit and vegetable intake and mortality: results from 2 prospective cohort studies of US men and women and a meta-analysis of 26 cohort studies. Circulation, 143(17), pp.1642-1654. 

16. Aasheim, E.T., Sharp, S.J., Appleby, P.N., Shipley, M.J., Lentjes, M.A., Khaw, K.T., Brunner, E., Key, T.J. and Wareham, N.J., 2015. Tinned fruit consumption and mortality in three prospective cohorts. PloS one, 10(2), p.e0117796. 

17. Fardet, A., Richonnet, C. and Mazur, A., 2019. Association between consumption of fruit or processed fruit and chronic diseases and their risk factors: a systematic review of meta-analyses. Nutrition reviews, 77(6), pp.376-387.    

18. Muraki, I., Imamura, F., Manson, J.E., Hu, F.B., Willett, W.C., van Dam, R.M. and Sun, Q., 2013. Fruit consumption and risk of type 2 diabetes: results from three prospective longitudinal cohort studies. Bmj, 347. 

19. D’Elia, L., Dinu, M., Sofi, F., Volpe, M. and Strazzullo, P., 2021. 100% Fruit juice intake and cardiovascular risk: a systematic review and meta-analysis of prospective and randomised controlled studies. European Journal of Nutrition, 60(5), pp.2449-2467. 

20. Viguiliouk, E., Glenn, A.J., Nishi, S.K., Chiavaroli, L., Seider, M., Khan, T., Bonaccio, M., Iacoviello, L., Mejia, S.B., Jenkins, D.J. and Kendall, C.W., 2019. Associations between dietary pulses alone or with other legumes and cardiometabolic disease outcomes: an umbrella review and updated systematic review and meta-analysis of prospective cohort studies. Advances in Nutrition, 10(Supplement_4), pp.S308-S319.  

21. Li, H., Li, J., Shen, Y., Wang, J. and Zhou, D., 2017. Legume consumption and all-cause and cardiovascular disease mortality. BioMed research international, 2017.  

22. Li, J. and Mao, Q.Q., 2017. Legume intake and risk of prostate cancer: a meta-analysis of prospective cohort studies. Oncotarget, 8(27), p.44776.( https://www.oncotarget.com/article/16794/text/ ) 

23. Zhu, B., Sun, Y., Qi, L., Zhong, R. and Miao, X., 2015. Dietary legume consumption reduces risk of colorectal cancer: evidence from a meta-analysis of cohort studies. Scientific reports, 5(1), pp.1-7. 

24. Becerra-Tomás, N., Papandreou, C. and Salas-Salvadó, J., 2019. Legume consumption and cardiometabolic health. Advances in Nutrition, 10(Supplement_4), pp.S437-S450.  

25. Pearce, M., Fanidi, A., Bishop, T.R., Sharp, S.J., Imamura, F., Dietrich, S., Akbaraly, T., Bes-Rastrollo, M., Beulens, J.W., Byberg, L. and Canhada, S., 2021. Associations of total legume, pulse, and soy consumption with incident type 2 diabetes: Federated meta-analysis of 27 studies from diverse world regions. The Journal of nutrition, 151(5), pp.1231-1240.  
 
26. Tang, J., Wan, Y., Zhao, M., Zhong, H., Zheng, J.S. and Feng, F., 2020. Legume and soy intake and risk of type 2 diabetes: A systematic review and meta-analysis of prospective cohort studies. The American journal of clinical nutrition, 111(3), pp.677-688.  

27. Mueller, N.T., Odegaard, A.O., Gross, M.D., Koh, W.P., Mimi, C.Y., Yuan, J.M. and Pereira, M.A., 2012. Soy intake and risk of type 2 diabetes mellitus in Chinese Singaporeans. European journal of nutrition, 51(8), pp.1033-1040. 

28. Sohouli, M.H., Lari, A., Fatahi, S., Shidfar, F., G?man, M.A., Guimarães, N.S., Sindi, G.A., Mandili, R.A., Alzahrani, G.R., Abdulwahab, R.A. and Almuflihi, A.M., 2021. Impact of soy milk consumption on cardiometabolic risk factors: A systematic review and meta-analysis of randomized controlled trials. Journal of Functional Foods, 83, p.104499. 

29. Reed, K.E., Camargo, J., Hamilton-Reeves, J., Kurzer, M. and Messina, M., 2021. Neither soy nor isoflavone intake affects male reproductive hormones: An expanded and updated meta-analysis of clinical studies. Reproductive Toxicology, 100, pp.60-67. 

30. Otun, J., Sahebkar, A., Östlundh, L., Atkin, S.L. and Sathyapalan, T., 2019. Systematic review and meta-analysis on the effect of soy on thyroid function. Scientific reports, 9(1), pp.1-9.  

31. Aune, D., Keum, N., Giovannucci, E., Fadnes, L.T., Boffetta, P., Greenwood, D.C., Tonstad, S., Vatten, L.J., Riboli, E. and Norat, T., 2016. Nut consumption and risk of cardiovascular disease, total cancer, all-cause and cause-specific mortality: a systematic review and dose-response meta-analysis of prospective studies. BMC medicine, 14(1), pp.1-14.   

32. Grosso, G., Yang, J., Marventano, S., Micek, A., Galvano, F. and Kales, S.N., 2015. Nut consumption on all-cause, cardiovascular, and cancer mortality risk: a systematic review and meta-analysis of epidemiologic studies. The American journal of clinical nutrition, 101(4), pp.783-793.    

33. Mayhew, A.J., de Souza, R.J., Meyre, D., Anand, S.S. and Mente, A., 2016. A systematic review and meta-analysis of nut consumption and incident risk of CVD and all-cause mortality. British Journal of Nutrition, 115(2), pp.212-225. 

34. Chen, G.C., Zhang, R., Martínez-González, M.A., Zhang, Z.L., Bonaccio, M., Van Dam, R.M. and Qin, L.Q., 2017. Nut consumption in relation to all-cause and cause-specific mortality: a meta-analysis 18 prospective studies. Food & function, 8(11), pp.3893-3905. 

35. Schwingshackl, L., Hoffmann, G., Iqbal, K., Schwedhelm, C. and Boeing, H., 2018. Food groups and intermediate disease markers: a systematic review and network meta-analysis of randomized trials. The American journal of clinical nutrition, 108(3), pp.576-586., 

36. Khalesi, S., Paukste, E., Nikbakht, E. and Khosravi-Boroujeni, H., 2016. Sesame fractions and lipid profiles: a systematic review and meta-analysis of controlled trials. British Journal of Nutrition, 115(5), pp.764-773. 
 
37. Pan, A., Yu, D., Demark-Wahnefried, W., Franco, O.H. and Lin, X., 2009. Meta-analysis of the effects of flaxseed interventions on blood lipids. The American journal of clinical nutrition, 90(2), pp.288-297. 

38. Khosravi?Boroujeni, H., Nikbakht, E., Natanelov, E. and Khalesi, S., 2017. Can sesame consumption improve blood pressure? A systematic review and meta?analysis of controlled trials. Journal of the Science of Food and Agriculture, 97(10), pp.3087-3094. 

39. Aune, D., Keum, N., Giovannucci, E., Fadnes, L.T., Boffetta, P., Greenwood, D.C., Tonstad, S., Vatten, L.J., Riboli, E. and Norat, T., 2016. Whole grain consumption and risk of cardiovascular disease, cancer, and all cause and cause specific mortality: systematic review and dose-response meta-analysis of prospective studies. bmj, 353. 

40. Chen, J., Huang, Q., Shi, W., Yang, L., Chen, J. and Lan, Q., 2016. Meta-analysis of the association between whole and refined grain consumption and stroke risk based on prospective cohort studies. Asia Pacific Journal of Public Health, 28(7), pp.563-575.

41. Reynolds, A., Mann, J., Cummings, J., Winter, N., Mete, E. and Te Morenga, L., 2019. Carbohydrate quality and human health: a series of systematic reviews and meta-analyses. The Lancet, 393(10170), pp.434-445. 

42. Schlesinger, S., Neuenschwander, M., Schwedhelm, C., Hoffmann, G., Bechthold, A., Boeing, H. and Schwingshackl, L., 2019. Food groups and risk of overweight, obesity, and weight gain: a systematic review and dose-response meta-analysis of prospective studies. Advances in Nutrition, 10(2), pp.205-218.

43. Aune, D., Norat, T., Romundstad, P. and Vatten, L.J., 2013. Whole grain and refined grain consumption and the risk of type 2 diabetes: a systematic review and dose–response meta-analysis of cohort studies. European journal of epidemiology, 28(11), pp.845-858.  

44. Schwingshackl, L., Hoffmann, G., Lampousi, A.M., Knüppel, S., Iqbal, K., Schwedhelm, C., Bechthold, A., Schlesinger, S. and Boeing, H., 2017. Food groups and risk of type 2 diabetes mellitus: a systematic review and meta-analysis of prospective studies. European journal of epidemiology, 32(5), pp.363-375. 

45. Vieira, A.R., Abar, L., Chan, D.S.M., Vingeliene, S., Polemiti, E., Stevens, C., Greenwood, D. and Norat, T., 2017. Foods and beverages and colorectal cancer risk: a systematic review and meta-analysis of cohort studies, an update of the evidence of the WCRF-AICR Continuous Update Project. Annals of Oncology, 28(8), pp.1788-1802.   

46. Zhang, B., Zhao, Q., Guo, W., Bao, W. and Wang, X., 2018. Association of whole grain intake with all-cause, cardiovascular, and cancer mortality: a systematic review and dose–response meta-analysis from prospective cohort studies. European journal of clinical nutrition, 72(1), pp.57-65. 

47. Wei, H., Gao, Z., Liang, R., Li, Z., Hao, H. and Liu, X., 2016. Whole-grain consumption and the risk of all-cause, CVD and cancer mortality: a meta-analysis of prospective cohort studies. British Journal of Nutrition, 116(3), pp.514-525.  

48. Benisi-Kohansal, S., Saneei, P., Salehi-Marzijarani, M., Larijani, B. and Esmaillzadeh, A., 2016. Whole-grain intake and mortality from all causes, cardiovascular disease, and cancer: a systematic review and dose-response meta-analysis of prospective cohort studies. Advances in Nutrition, 7(6), pp.1052-1065. 

49. Healthgrain Forum http://orcid. org/0000-0002-9585-0141 Ross Alastair B Alastair. Ross@ chalmers. se van der Kamp Jan-Willem King Roberto Lê Kim-Anne Mejborn Heddie http://orcid. org/0000-0001-6313-6691 Seal Chris J Thielecke Frank, 2017. Perspective: A definition for whole-grain food products—Recommendations from the Healthgrain Forum. Advances in Nutrition, 8(4), pp.525-531.  

50. Taylor, R., Ramachandran, A., Yancy, W.S. and Forouhi, N.G., 2021. Nutritional basis of type 2 diabetes remission. bmj, 374.