Plant Polyphenols Lignans and Cardiovascular Disease

David J.A. Jenkins, MD, PHD; Cyril W.C. Kendall, PHD; John L. Sievenpiper, MD, PHD


J Am Coll Cardiol. 2021;78(7):679-682. 

Broadly, there are 4 classes of phenolic compounds[1] that are found in plants that total >8,000 polyphenolic and phenolic compounds. These compounds include phenolic acids flavonoids (including isoflavones), stilbenes (eg, resveratrol), and lignans (that are converted by gut bacteria into enterodiol and enterolactone). In this issue of the Journal, dietary lignans are the focus of a study by Hu et al.,[2] linking their intake to coronary heart disease (CHD). These phenolic compounds have functions in the plant, such as prevention from ultraviolet light damage and antiseptic and antifungal activity.[3] Much has been written on potential mechanisms by which phenolics may prevent or reduce chronic disease, but data indicating direct links or associations with these diseases are lacking. The paper by Hu et al.[2] is therefore of particular interest in demonstrating a strong association with dietary total lignan intake and CHD events (hazard ratio: 0.85; 95% confidence interval [CI]: 0.79–0.92) in the 214,108 men and women of the 3 cohorts of the Health Professionals Follow-Up Study, Nurses' Health Study, and Nurses' Health Study II (6,283 nonfatal and 3,961 fatal myocardial infarctions). Study strengths include the large numbers of participants involved, the frequency of follow-up, the assessment, and use of multiple covariates, including ethnicity and family history of myocardial infarctions and also inclusion of the healthy plant-based diet index, as many of the foods that are good sources of lignans also part of a healthy plant-based diet (eg, whole grains [matairesinol], fruit and vegetables [pinoresinol, lariciresinol], grapes and wine [secoisolariciresinol]).

Other smaller studies referenced by Hu et al.[2] have been undertaken to assess the association of lignan intake with cardiovascular disease (CVD), but these cohorts have involved small numbers of participants that have likely limited the establishment of significant associations. A Dutch cohort of 16,165 women aged 49–70 years from the EPIC (European Prospective Investigation into Cancer and Nutrition) study, who were followed over 6 years, had 372 CHD and 149 stroke events, but the study failed to find an overall association between total lignan intake and CVD (CHD or stroke); however, an association was observed for smokers as having a higher CVD risk in which the lignan intake was negatively associated with CVD.

Further, a similarly small study of 570 Dutch men of the Zutphen cohort were followed for 15 years, during which time 392 men died (from CHD, CVD, and cancer). Although there was no association between lignan intake and CVD, there was a significant negative association between the grain lignan, matairesinol, and CHD.

These 2 studies perhaps indicate a signal for a lignan affect, but the participant events were too few to come to definite conclusions.

There are, however, meta-analyses on total flavonoids and also isoflavones that are notable in having the large number of participants and events required to see the associations of plant phenolics with disease outcomes. These larger analyses have shown negative associations among flavonoid intake, CHD, and stroke.

The CHD flavonoid meta-analysis from 2015 included 14 cohorts and 452,564 men and women who developed 7,233 CHD events, in which the relative risk (RR) for the association of total flavonoid with CHD was 0.85 (95% CI: 0.79–0.910) and there was no significant heterogeneity (I2 = 26%).[4]

A further CHD study in 2016 included 11 cohorts and 356,627 participants who developed 5,126 strokes, and the RR for stroke, comparing extreme quantiles for flavonoid intake, was 0.89 (95% CI: 0.82–0.93; P = 0.006; with no heterogeneity, I2 = 0%), and there was a significant dose-response relationship.[5]

These meta-analyses demonstrate that with adequate numbers of participants in the analysis, phenolic intakes, as flavonoids in the present discussion, appeared to show significant associations with key aspects of CVD.

Hence, we can therefore have more confidence in studies in which there are significant numbers of participants, 3 as in the present study, in which participant numbers were increased by combining 3 cohorts.

Obviously, smaller studies may also be informative, especially when linked to randomized controlled trials with defined outcomes and defined diets. Most notable in this respect is the PREDIMED (Prevención con Dieta Mediterránea) trial with >7,000 participants in which a Mediterranean diet, naturally high in flavonoids, was supplemented over a 4-year period with nuts and extra virgin olive oil that are also good sources of flavonoids. This diet was compared with a low-fat American Heart Association–type diet with much lower levels of flavonoids and demonstrated a highly significant CVD reduction.[6] Importantly, in an observational study on this trial, there was a strong association between polyphenol intake and the risk of CVD in the comparison of quintile 5 versus quintile 1. It demonstrated a 46% reduction in CVD associated with total polyphenol intake, 60% reduction with flavonoids, 49% for lignans, and 51% for hydroxybenzoic acid, with extra virgin olive oil contributing significantly to the latter 2 phenolics.[7]

Much work has been undertaken on the soy isoflavones, as a further example of plant phenolics with antioxidant and selective estrogen receptor modulating activity. As with the present study, the Nurses' Health Study and Nurses' Health Study II were combined with the Health Professional Study cohort, resulting in a total of 210,700 participants and 8,359 CHD cases. Analysis of these data for the association of isoflavone intake with CHD gave a hazard ratio in the comparison of the extreme quintiles of 0.87 (95% CI: 0.81–0.94; P = 0.008).[8] These data provide additional evidence for the association of certain plant phenolics with CVD.

There are no randomized controlled trials related to CVD outcomes and plant phenolics. What trials there are relate to the proposed mechanisms by which plant phenolics and foods high in plant phenolics reduce CVD risk. Studies have assessed the effect of test meals and diets high in plant phenolics, including nuts, on increasing flow-mediated vasodilatation.[9] Plant phenolics may have marked antioxidant activity and are selective estrogen receptor modulators, sometimes referred to as phytoestrogens.[10] The antioxidant activity of these compounds has attracted attention for their possible roles in both CVD and cancer prevention, possibly through preventing oxidative damage to low-density lipoprotein cholesterol, and so reducing low-density lipoprotein cholesterol uptake by the scavenger system.[11,12] They may also reduce oxidative damage to DNA with the potential for reducing initiation and transformation to malignant cells. The anti-inflammatory effects have been assessed by C-reactive protein and interleukin-6 levels. For soy, significant reductions in C-reactive protein have been shown when soy foods were taken as part of a cholesterol-lowering diet.[13] Increased polyphenol intake has been shown to reduce platelet activation and aggregation, inhibit pro-matrix metalloproteinases involved in plaque growth and instability, and improve endothelial function.[9,14] Soy polyphenols (isoflavones) have been suggested to be part of the mechanism by which soy foods lower serum cholesterol. However, diets comparing soy foods with high or low isoflavone levels reduced serum cholesterol equally,[12] and the soy affect in lowering serum cholesterol is now attributed to the protein, rather than to the isoflavone content of soy.

The evidence is building that there is an association between polyphenol intake and chronic disease, especially for CVD. Plant polyphenols may be important components of healthy plant-based diets that contribute to freedom from chronic diseases such as CVD, diabetes, and possibly cancer and so are associated with a reduction in all-cause mortality. However, although the associations are promising, it would be helpful to have trial data, both for total and individual polyphenols. At the same time, such trials would be difficult to undertake, and sustained adherence may not be possible to achieve even over short periods of time, as seen with the difficulty in increasing fruit and vegetable intake,[15] let alone adherence over the number of years as required to see chronic disease events. Perhaps an exception to this concern is the PREDIMED trial, in which foods were provided that are considered highly desirable, as nuts and olive oil, for the duration of the trial, and adherence was excellent.

In the absence of randomized controlled trial data, the current assessment in the Journal of the association of lignan intake with freedom from CVD will act as a stimulus to consider higher intakes of lignan-containing plant foods and also support current guidelines to consume more plant-based diets, including fruit, vegetables, whole-grain cereals, legumes, nuts, and oils such as olive oil, and all in less processed form, as these are the lignan-containing foods.