Observational Studies Can Point the Way, but Only Clinical Trials Can Prove Causality

An image depicting healthful foods in a clinical setting

New observational data by Wang et al.1 highlight the health benefits of fruits and vegetables. These foods are known for being sources of vitamins and minerals and for their phytochemical content.2 Soybeans are also known for their high phytochemical content, especially that of isoflavones.3 But the question of whether fruits and vegetables are protective in and of themselves, as opposed to simply being markers of a healthy lifestyle, remains as pertinent today as it was when it was first raised decades ago.

In the above referenced research, intake of ≈5 servings per day of fruit and vegetables, or 2 servings of fruit and 3 servings of vegetables, was associated with the lowest mortality.1 In comparison with the reference intake level of 2 servings/day, daily intake of 5 servings of fruit and vegetables was associated with a 13% reduction in total mortality, a 12% reduction in CVD mortality, and a 10% reduction in cancer mortality. These results by Wang et al.,1 come from a combined analysis of two cohort studies, one involving 66,719 women from the Nurses’ Health Study (NHS) and the other, 42,016 men from the Health Professionals Follow-up Study (HPFS).  During the 30 years of follow-up in the NHS there were 18,793 deaths and during the 28 years of follow-up in the HPFS there were 15,105 deaths.  Wang et al. also meta-analyzed data from 26 prospective cohort studies including the NHS and HPFS.  Results indicated that the consumption of 5 servings of fruits and vegetables per day was associated with a 13% reduction in mortality.

Interestingly, consuming more than 5 servings/day was not associated with greater benefit.  Also, higher intakes of most subgroups of fruits and vegetables were associated with lower mortality, but this was not the case for starchy vegetables such as peas and corn. In addition, intakes of fruit juices and potatoes were not associated with total and cause-specific mortality.

Epidemiologic studies are not designed to establish cause and effect relationships, but rather, to determine whether two observations are related to one another.  With that recognition in mind, one can certainly speculate about the nature of the relationship between fruits and vegetables and mortality based on the above-described results.  Why is it that peas, corn, potatoes, and fruit juice, were not associated with mortality? Wang et al.1 suggest several possible biological mechanisms as to why these foods were not associated with mortality, but other fruits and vegetables were.  While these mechanisms may account for the difference, it is also possible that the consumption of peas, corn, potatoes, and fruit juice is not reflective of a healthy lifestyle in the same way the consumption of other fruits and vegetables associated with mortality is. According to 2014 data, potatoes and corn were consumed 5 times and ~2 times more often than broccoli, respectively.4 Consumption of kale was even lower than broccoli. The consumption of orange juice was almost 5 times greater than the consumption of oranges. Aside from their choice of fruits and vegetables, people who regularly consume orange juice but not oranges, or who eat corn but not broccoli, might differ in many ways that affect mortality from people who eat oranges and broccoli. Researchers go to great lengths to control for these differences, that is, to control for potentially confounding variables, but it is impossible to control for everything that might affect an outcome.

Furthermore, it is puzzling that consuming daily >5 servings of fruits and vegetables was not associated with greater reductions in mortality than 5 servings.  Perhaps, it is because the mechanism responsible for their proposed protective effects is fully operational at 5 servings, so that no additional benefit occurs beyond that amount.  But it is certainly coincidental that 5 servings/day is the number that has been promoted by health promotion programs as being beneficial.

The above discussion is not to suggest that the consumption of fruits and vegetables does not in and of itself reduce mortality.  It very well might do just that.  But it is to suggest that definitive data about the protective effects of these foods can come only from clinical trials. This is also true for understanding about soyfoods and health.

Soyfoods have been rigorously investigated for the past 30 years. This research has produced a wide range of hypotheses about the health benefits of these foods. These hypotheses can only be definitively evaluated by conducting clinical trials. There may be exceptions, but for the most part, additional epidemiologic observations are not going to further advance our understanding of the role of soy in disease prevention.

Finally, if in fact, fruits and vegetables are simply markers of a healthy lifestyle and not directly protective, the data still show that a healthy lifestyle is important for longevity. Identifying precisely what that lifestyle is will continue to be the focus of researchers around the world.


  1. Wang DD, Li Y, Bhupathiraju SN, et al. 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. 2021.
  2. Van Duyn MA, Pivonka E. Overview of the health benefits of fruit and vegetable consumption for the dietetics professional: selected literature. J Am Diet Assoc. 2000;100:1511-21.
  3. Franke AA, Custer LJ, Wang W, et al. HPLC analysis of isoflavonoids and other phenolic agents from foods and from human fluids. Proc Soc Exp Biol Med. 1998;217:263-73.
  4. Produce for Better Health Foundation. State of the Plate, 2015 Study on America’s Consumption of Fruit and Vegetables, Produce for Better Health Foundation, 2015.
Dr. Mark Messina

Author Dr. Mark Messina

PhD in Nutrition, Director of Nutrition Science and Research, Soy Nutrition Institute Global. Expert in soyfoods and isoflavones.

More posts by Dr. Mark Messina

Comments will be reviewed by the website moderator, but are not guaranteed to be posted.