Table of Contents
The popularity of soy foods and soy-based products has been increasing in recent times. This has been particularly the case as a dairy alternative, with people switching to using soy ‘milk’ and soy-based yogurts and cheese. Additionally, soy has become popular as a meat alternative in a variety of dishes for those looking to reduce meat intake. Soy foods such as tofu can be used in recipes in place of meat, and soy-based ‘meat alternatives’ that are vegetarian and vegan friendly have been developed.
With this increased prevalence, there has been some debate about the health effects of consuming soy foods and products. On one side, there have potential benefits highlighted of inclusion of soy in the diet. It contains phytoestrogens, which may have beneficial effects. Additionally, it is low in saturated fat, and so is potentially beneficial when used in place of saturated fat-rich foods.
However, some have claimed that the phytoestrogens (isoflavones specifically) in soy can be a cause for concern due to the ability of these compounds to mimic the effects of the hormone oestrogen. One common claim is that high soy intake is detrimental for men particularly, as it is “feminizing”; causing gynecomastia, loss of libido and erectile dysfunction.
So what is the truth? Is soy a health food? A harmful endocrine disruptor? Or simply neutral? In this episode we dive into the research and look at the evidence to date tells us about these questions. We consider two big health outcomes in particular; cardiovascular disease and breast cancer. And then final discuss what this means practically for our dietary choices.
People in this Episode
- Danny Lennon – Sigma Nutrition, Director
- Alan Flanagan – Sigma Nutrition, Research Commmunication Officer
03:02 – Soy: Nutrient profile & metabolism
20:18 – Soy & men’s health: Is soy “feminizing”?
39:42 – Breast cancer
55:09 – Cardiovascular disease & menopause
1:12:15 – Pragmatic recommendations
01:44 – Soy: Nutrient profile & metabolism
19:00 – Soy & men’s health: Is soy “feminizing”?
38:24 – Breast cancer
53:51 – Cardiovascular disease & menopause
1:10:56 – Pragmatic recommendations
1:17:30 – Danny’s “Key Ideas”
The three key ideas highlighted for this episode were:
- Impact of phytoestrogens and their metabolites
- Meta-point about research appraisal
- Can I eat soy regularly?
1: Impact of phytoestrogens and their metabolites
Perhaps the most interesting aspect to consider about soy as a nutrient source, is that it contains isoflavones, a group of polyphenol compound. Isoflavones are phytoestrogens, meaning they are compounds that can behave like estrogens.
One of the subtypes of isoflavone, daidzein, undergoes metabolism by the gut bacteria, producing metabolites, most importantly S-equol. Equol has an affinity for binding to estrogen receptors that is higher than that of the parent isoflavone, daidzein.
It is important to note how isoflavones can act differently to estrogen:
- There are two oestrogen receptors in the human body: alpha and beta receptors
- The hormone oestrogen binds with equal affinity to both alpha and beta
- Isoflavones do indeed have oestrogen-mimicking actions, but they don’t act in the exact same way.
- Rather isoflavones preferentially bind to the beta receptors, with much less of an affinity for the alpha receptors
- Therefore they classified as SERMs (Selective Estrogen Receptor Modulators)
- The difference in receptor binding between isoflavones and oestrogen is important because the alpha and beta receptors are present in differing amounts in different tissues, and have different effects in the body when one of the compounds binds to them.
- So isoflavones can have some oestrogen-like effects in the body, but it is weaker of an effect, and it also doesn’t mimic all the effects of oestrogen. Affinity for binding to estrogen receptors is higher for equol than for the parent isoflavone
- Importantly, isoflavones they have some oestrogen-like effects, without changing oestrogen concentrations. Maybe why differences in relation to risk predicted by mechanisms.
- Finally, as mentioned in relation to them being a phytonutrient, isoflavones can have effects that are completely separate to an oestrogen-mimicking effect.
2: Meta-point about research appraisal
Evidence used to suggest various harms of soy consumption typically comes from:
- laboratory or animal studies, that…
- use pure isoflavones or high doses
But animals metabolise isoflavones in a different way to humans; they are much more efficient.
Animal trials can be useful for initial investigation of a hypothesis, or to establish mechanisms. Particularly in cases where there is yet to be any human data.
But… this is an area where there is human intervention trials and epidemiological trials. So why would we put the findings of animals trials ahead of that?
Whilst animal studies have raised concern that soy may cause untoward effects, the totality of human data refutes this.
3: Can I eat soy regularly?
- Isoflavones’ potential to have bad side effects in humans is based solely on data from laboratory or animal studies using pure isoflavones or high doses.
- In addition to being a good source of protein and micronutrients, soy may have some unique benefits as it contains naturally occurring bioactive components such as isoflavones
- Comprehensive reviews by the European Food Safety Authority, World Cancer Research Fund and the World Health Organisation all conclude that soy foods as part of a healthy balanced diet are safe.
- In totality it seems soy can be consumed daily from a number of foods, as part of an overall healthy dietary pattern, with 1-2 servings per day of soy foods/drinks included in most guidelines.
Detailed Study Notes
What is Soy? The Nutrient Profile
You may see it referreed to as soy (US) or soya (UK and others), but the terms mean the same thing.
Increasing popularity worldwide now, particularly as a dairy alternative (soy milks or soy-based yogurts and cheeses). As well as a meat alternative for those looking to reduce meat intake, as soy foods and products often serve as a meat-alternative in a variety of vegetarian dishes.
Soybeans are a legume of Asian origin. But… compared to most other beans and pulses, soy beans:
- have a much lower carbohydrate content
- are higher in protein quality and content
- are significantly higher in total fat (mainly polyunsaturated fat)
As an example, consider the % of calories from these types of cooked beans (ref):
- Soy beans: 12% CHO, 36% PRO, 42% fat
- Other beans: 70% CHO, 27% PRO, 3% fat
Soybeans come in different varieties, differing in colour:
- Yellow soybeans: The main type used to make soy milk, tofu, tempeh, and soy flour.
- Green soybeans: Young green soybeans = edamame. Eaten as appetizer or in salads, stir-fries, and soups.
- Black soybeans: A number of Asian food cultures use fermented black soybeans in their dishes.
We also distinguish between unfermented soy foods (e.g. soy milk, tofu, soy-based meat alternative) and fermented soy foods (e.g. tempeh, natto)
- Great source of high-quality plant protein
- Contains all nine essential amino acids
- Soy milks (3g/100ml) have more protein than many other plant milks
- Notably, soy drinks are the only plant-based substitutes seen as equivalent to dairy by dietary guidelines
- Tofu is very high in protein per serving, often why it’s used as a meat substitute, particularly for those looking to keep protein intake high
- Contains a range of micronutrients, notably folate and vitamin C
- Soy beans are the main dietary source of isoflavones
- Soy food and drink products provide on average 3-4mg of isoflavones per 1g soy protein
- British Dietetic Association (BDA): Some processing methods can decrease isoflavone content by 80-90% (e.g. isolated soy protein)
- In Japan, the average isoflavone intake is approximately 40 mg/d compared to Europe and US at less than <3mg/d
Isoflavone Metabolism & Effects
Soy contains isoflavones, a compound that is central to most of the health impacts discussed in relation to soy.
Specifically, they are a sub-class of flavanoid:
Isoflavones differ from other flavonoids subclasses in one important way: they are phytoestrogens, meaning they are compounds that behave like estrogens (or “oestrogens” for us in Europe. Note: I’ll use the US spelling for these notes, as the US has our biggest audience share).
There are then 3 subtypes of isoflavone:
- Daidzein (~50%)
- Undergoes metabolism by the gut bacteria
- Produces metabolites, most importantly S-equol
- Genistein (~40%)
- Glycitein (~10%)
Equol: The metabolism of daidzein by gut bacteria leads to the production of s-equol. Affinity for binding to estrogen receptors is higher for equol than for the parent isoflavone.
“Equol producer” = people have specific gut bacteria who are specialised at metatobolising daidzein into equol
Soy Phytoestrogens: Impact of Background Diet and Ethnicity
Alan mentioned a 2009 RCT by Vergne et al. that had two trials set up to evaluate whether soy phytoestrogen bioavailability was influenced by:
- chronic ingestions
- ethnic origin
- dietary context
They had people consume a soy-based cheese (containing ~45mg isoflavones) daily for 10 days. In order to compare the impact in a Western diet context versus an Asian dietary context, the first trial examined twelve healthy young Asians immersed for 2 months in France, with the participants randomised in a cross-over design to both a Western and Asian diet. The second trial compared Asians under the Western diet to twelve healthy young male Caucasians under the same diet.
The results of the trials seemed to show that after an acute intake of soy cheese, Asians absorb soy phytoestrogens better than Caucasians, regardless of whether the background diet is Western or Asian. On chronic ingestions, Caucasians had an adapatation and saw increased plasma concentrations of daidzein and genistein.
How Isoflavones Act Differently to Estrogen
- There are two estrogen receptors in the human body: alpha and beta receptors
- The hormone estrogen binds with equal affinity to both alpha and beta. (Affinity describes how ‘strongly’ a hormone binds with a receptor, i.e. how well it “fits” the receptor, and how likely it is to bind)
- Isoflavones do indeed have estrogen-mimicking actions, but they don’t act in the exact same way.
- Rather isoflavones preferentially bind to the beta receptors, with much less of an affinity for the alpha receptors
- Therefore they classified as SERMs (Selective Estrogen Receptor Modulators)
- The difference in receptor binding between isoflavones and estrogen is important because the alpha and beta receptors are present in differing amounts in different tissues, and have different effects in the body when one of the compounds binds to them.
- So isoflavones can have some estrogen-like effects in the body, but it is weaker of an effect, and it also doesn’t mimic all the effects of estrogen.
- Importantly, isoflavones don’t really affect estrogen levels. So they have some estrogen-like effects, without changing estrogen concentrations. Maybe why differences in relation to risk predicted by mechanisms.
- Finally, as they are a phytonutrient, isoflavones can have effects that are completely separate to an estrogen-mimicking effect.
Soy & Men’s Health: Is Soy “Feminizing”?
- As noted aboce, soy contains isoflavones, which are phytoestrogens – meaning they are compounds that can behave like estrogens.
- Estrogen is also present in small amounts in the male body, and if testosterone drops for any reason (e.g. due to certain medications), then the balance of testosterone:estrogen can be negatively skewed. This could lead to man experiencing symptoms such as breast tissue tenderness, swelling or growth.
- Gynecomastia = a condition of overdevelopment or enlargement of the breast tissue in men
- Given that male-to-female transgender patients take estrogens to stimulate breast growth, there has been a hypothesis that maybe high soy intake could have similar effects.
- One published case study of a 60 year old man drinking nearly 3 litres of soy milk per day. oestrone and oestradiol (2 of the 3 major endogenous estrogens, with the other being oestriol) concentrations to be 4-fold increased above the upper limit of the reference range. Resolved on stopping drinking.
- No good evidence to suggest it lowers testosterone to any clinically meaningful degree, and zero evidence outside of that one case study, that it causes gynecomastia.
- A 2007 trial on 12 men examined the effect of taking 2 scoops of soy isolate powder per day for 28 days, showing that it decreased serum testosterone. However, a meta-analysis three years later showed that neither soy foods or supplements caused a change in serum testosterone.
- Reed et al., 2021 meta-analysis: ‘Neither soy nor isoflavone intake affects male reproductive hormones’
There is no strong evidence that consuming soy foods and beverages leads to a clinically meaningful reduction in testosterone or that it causes gynecomastia, erectile dysfunction or other negative side effects for men’s health.
Are Isoflavones “Endocrine Disruptors”?
Technical review by Messina et al., 2021:
- Included 417 reports (229 observational studies, 157 clinical studies and 32 systematic reviews & meta-analyses) that examined the relationship between soy and/or isoflavone intake and endocrine-related endpoints
- Found isoflavone intake does not adversely affect thyroid function.
- No adverse effects on breast or endometrial tissue
- No adverse effects on estrogen levels
- Suggestion that menstrual cycle length may be slightly increased, however ovulation is not prevented
- No adverse effects on testosterone or estrogen levels
- No adverse effects on sperm or semen parameters
EFSA, 2010: Endocrine active substances are not necessarily endocrine disrupting chemicals
Soy Infant Formula (SIF) & Soy for Children
Soy milk not recommend in first 6 months of life. Perhaps even first 12 months.
Beyond that, there is no strong human evidence showing SIF is problematic for children, except perhaps children with congenital hypothyroidism (Vesta et al., 2018).
- Setchell et al., 1998 – Isoflavone content of infant formulas and the metabolic fate of these phytoestrogens in early life
- Bernbaum et al., 2008 – Pilot Studies of Estrogen-Related Physical Findings in Infants
- Zung et al., 2008 – Breast development in the first 2 years of life: an association with soy-based infant formulas
Estimates for the prevalence of soy allergy range from 0.1% to 0.6% (1 in 1,000 to 1 in 165). By comparison, milk allergy is much higher prevalence.
On average, most data suggest slightly higher in children; with US data suggesting 0.5%. US data shows decreased by 60% when comparing the average prevalence for all ages with the prevalence for adolescents 14 to 17 years of age. But it’s estimated 70% of children with soy allergy grow out of it by the age of 10.
Soy & Hypothyroidism
Those with thyroid issues may be on the medication levothyroxine. Soy can interfere with the absorption of levothyroxine (see here and here). So it’s recommended that those taking this medication do not consume any soy within 30-60 minutes of taking their medication. But outside of that window, soy should be fine to take (of course, this is not advice. Talk to your doctor if you’ve questions).
Based on mechanistic reasoning related to estrogen’s effects it has been hypothesised that soy isoflavones could increase breast cancer risk, increase risk of reoccurance, or be a problem for those with breast cancer.
Original data on animals in the 1990s indicated the potential for increased risk. However, this has not been replicated in humans.
Differences in ER-positive vs. ER-negative breast cancer
Breast cancers that have estrogen receptors are called ER-positive (or ER+) cancers. While those without receptors are ER-negative.
Of the studies included in the Nachvak meta-analysis, three studies examined the association between soy isoflavones intake and breast cancer mortality based on estrogen receptor (ER) status in patients with breast cancer. Soy isoflavones intake was inversely related to death from breast cancer among women with ER-negative breast cancer (i.e. higher intake = lower risk) but not in those with ER-positive breast cancer.
Stronger binding of phytoestrogens to beta receptors.
The 2017 (revised 2018) World Cancer Research Fund report on ‘Diet, Nutrition, Physical Activity & Breast Cancer’ classified the evidence on soy and soy products as “Limited – No Conclusion” in relation to breast cancer risk. However, as noted in the podcast, trying to reach a threshold of evidence on any individually dietary component or nutrient in relation to significantly changing cancer risk is incredibly difficult, if not impossible.
Symptoms of Menopause
Soy has been hypothesised to help with hot flushes in menopausal women, although it’s unclear exactly why this occurs. Changes in estrogen levels are suspected to play a role in hot flushes, and so this may be the potential way in which isoflavones could be beneficial.
A 2012 meta-analysis by Taku et al. found that a media intake of 54mg of isoflavones consumed daily (which equates to about 2 servings of soy-based foods/drinks per day) is associated with reduced frequency and severity of hot flushes. And a 2018 RCT also suggested 50 mg/d of soy isoflavones could be beneficial, although previous trials have had conflicting results.
See: Episode #279: Avrum Bluming, MD & Carol Tavris, PhD – Estrogen, Menopause & Misconceptions About Hormone Replacement
There is commonly an increase in blood lipids during menopause. There are also a number of proposed mechanisms as to why the decrease in estrogen during menopause may increase CVD risk. More on lipids and CVD below.
Soy & Cardiovascular Disease
The use of a Portfolio diet (developed by David Jenkins at the University of Toronto) has been shown to be able to decrease LDL-C to a greater extent than just advice for a low saturated fat diet.
This diet is a lipid-lowering dietary pattern; i.e. it includes foods/nutrients in amounts that have known lipid-lowering effects. Four fundamental components of the diet are:
- Plant sterols
- Viscous fiber
- Soy protein
A typical Portfolio diet has about 50 g/d of soy protein. And a meta-analysis by Anderson & Bush (2011) showed that soy protein consumption (~ 30g/d) had a LDL-C lowering effect.
A cumulative meta-analysis by Jenkins et al., (2019) stated that: “we assessed, using a cumulative meta‐analysis, whether at any time point since 1999 had soy foods failed to lower serum cholesterol and found that LDL cholesterol reductions for soy protein have consistently been between −4.2 and −6.7 mg/dL (P<0.006), with no loss of significance at any time point, so justifying the continued use of soy for health and therapeutic purposes as part of cholesterol‐lowering diets.”
Richardson et al., 2015 found that systolic blood pressure was 4.3 mm Hg lower among patients consuming the highest amount of isoflavones. The results were modified by race, with a stronger benefit in African Americans.
- “The mechanism behind the BP-lowering effect seen here that might be attributed to isoflavones could be through the activation of endothelial nitric oxide (NO) synthase (eNOS) and stimulation of NO production. Endothelial cells generate the potent vasodilator NO from L-arginine using NO synthases. Genistein, one of the primary isoflavones found in soy, has been demonstrated to have direct nongenomic effects on eNOS activity in human aortic endothelial cells, leading to eNOS activation and NO synthesis.” – Richardson et al., 2015
Intake is accounted for in dietary guidelines:
- The UK Eatwell Guide:
- Fortified soy drinks included in the ‘Dairy and alternatives’ food group
- Soy proteins such as tofu and soy mince are included in the ‘Beans, pulses, fish, eggs, meat and other proteins’ group
- US Dietary Guidelines 2020-2025:
- Dairy group: fortified soy milk and yogurt alternatives.
- Protein group: tofu, tempeh, products made from soya flour, isolate and concentrate
- Vegetable group: edamame or soya beans
- Oils: soybean oil (in the US is ‘vegetable’ oil)
Guidelines suggest that 1-2 servings of soy-based foods and/or drinks (as part of an overall healthy dietary pattern) is potentially beneficial for both human health and planetary health.
One serving would equate to:
- 50g soy mince
- Quarter block of tempeh
- 90g firm tofu
- 50g edamame
- a large glass soy milk (250ml)
- 200g of plain soy yogurt
Isoflavones are calculated based on 3.5 mg isoflavone content per gram of protein.
Soy products (e.g. soy milks) are now usually fortified with calcium to match that of dairy equivalents. And a number of these products are also fortified with vitamin D (which enhances calcium uptake), vitamin B12 and iodine (seafood can be a rich source of iodine, but dairy is the main source of iodine in the diet in many Western populations due to how much it’s consumed. Dairy contains iodine due to cattle feed being fortified, and the fact iodine is used in sterilising teats at milking).
For those interested in environmental sustainability, the evidence shows that compared to dairy milk, soy milks have a significantly lower carbon footprint, require less water and land.