The answer as to whether soy is good or bad for somebody depends on both the individual and how they consume the soy food. There is no one statement that can apply to everybody.
Soy will affect you differently based on:
The form of soy ingested; be it soybean, soymilk, tofu, or soy protein powder
Your metabolic condition; healthy or unhealthy
Hormonal status and gender
This article will try to clarify a few points.
Editor's Note: The next section and the following few discuss Soy Processing techniques. Although valuable to know, they do not answer the question that the FAQ title poses. Please go to the section called Soy and Health (can click the side menu directly) and whatever follows for a direct answer to your question.
Soybeans are where all soy products begin. In their natural bean-y state, they are usually (by weight):
40% dietary protein
20% dietary fat
35% dietary carbohydrate (13% soluble, 22% insoluble)
5% Ash content (dietary and non-dietary minerals)
They also tend to contain
220mg sterols per 100g oil (the fat component)
1-10mg isoflavones per 100g of soybean, which are associated with the protein fragment of soy.
The soy isoflavones are a subset of Flavones called isoflavones. In particular they are Genistein and Daidzein, although some consideration should be given to the compound Equol (potent metabolite of daidzein). They can have biological significance in the doses found in some soy foods.
Interestingly, only some people produce Equol. It is made from Daidzein in the intestines by colonic bacteria, but not all persons have the strains of bacteria that can produce Equol. Some studies note that the 'benefits' or 'harms' associated with soy only apply to Equol-producers which consist of 30-50% of persons (in Western countries, higher percentage in those with Asian Lineage).
Soybeans also contain various anti-nutrients such as:
Trypsin inhibitors (inhibit protein digestion in the stomach and small intestine)
Phytates (mineral chelating agents; can prevent mineral uptake)
Eating soybeans straight involve ingesting all of the above compounds approximately those amounts (when listed). It is not advised due to Trypsin Inhibitors.
Soy can be directly processed into various food products that resemble or directly mention Soy. These are slightly different than products there merely use soy byproducts.
Soybeans can turn into their milk counterpart by a process of:
Soaking (in water or 0.5% sodium bicarbonate)
Rinsing (with water)
Filtrating (with water) (removes okara)
Cooking the remainder
Voila, soy milk.
The soaking and rinsing stages are fairly standard among soybean processing, and the soaking tends to deactivate some trypsin inhibitors but eliminate a good deal of the anti-nutrient phytate. If the soybeans have lectin content, it can also be reduced by soaking. Unfortunately, vitamin and mineral content decreases slightly during soaking.
Filtrating is a step in which the Okara is removed from the milk. Okara is a term for the soybean residues. It is mostly fiber, cellulose, hemicellulose, and some lignan, with about 25% protein by weight.
What is left after separating the Okara will be made into Soy Milk. Cooking the milk will always involve some manner of heat treatment. If soy producers use Ultra-High Temperate (UHT) processing techniques then this can destroy the trypsin inhibitors in soy. Thus any cooked form of soy or raw soy product made from soy milk (such as tofu) is free of trypsin inhibitors and the protein can be digested.
Alternatively, high-pressure processing can achieve the same desired effect.
Tofu can be made from Soy Milk in a process similar to making cheese from dairy milk. The milk:
Is coagulated with the addition of a coagulant
The 'wheys' (liquid fragment) are strained out during pressing of the newly formed 'curd'
Tofu is the result, with the 'wheys' as a byproduct
Soy undergoes various different processing techniques which may or may not affect its nutritional content. Below is an adapted chart from a review article which demonstrates the possible routes a raw soybean can undergo. The small ASCII guy with his raised hand points to a processing technique that applies to soy protein concentrate that is notable, as it removes most isoflavone content and makes supplementing soy protein non-estrogenic.
Full-fat soy products are divided into either 'active' products or 'deactivated' products. The difference is in regards to enzymes in soybeans (such as trypsin inhibitors).
After cleaning and preparing the beans, full-fat active flour is merely taking those beans and processing them into a powder (via milling). The trypsin inhibitors remain active.
'Deactivated' full-fat flour and grits undergo heat treatment at 85C, or an alternate but equally effective method of enzyme deactivation, in order to deactivate (destroy the effects of) the trypsin inhibitors.
Both full-fat options retain similar nutritional profiles to the raw soy bean (40% protein, 20% fat, 5% ash, 35% carbohydrate) and have the full isoflavone and saponin content.
Soy Oil, one of the main manufacturing products of soy beans, is the process of taking raw beans (not either full-fat option) and separating the oil from the rest of the bean.
This results in Soy Oil, a commodity, and de-fatted 'white flakes' that can be used in further processing.
The defatted flour and grits are what remains of the full-fat versions after the soy oil is removed; they are typically referred to as 'white flakes'. These flakes may either:
Be milled, to produce active defatted soy flour/grits
Be toasted and then milled, to produce deactivated defatted soy flour/grits
Note that the previously mentioned 'deactivation' of trypsin inhibitors was a branch off the path, and did not apply to Soy Oil. It is possible to get white flakes with active trypsin inhibitors.
If not used directly for soy flour or grits, they will be used to make a soy protein powder. Either for supplemental usage, or to use in food processing.
Soy protein concentrates are when the 'white flakes' of defatted soy is treated to an ethanol (alcohol) extract, has its pH neutralized, and is then ready to go.
Due to the aforementioned ethanol (alcohol) extraction, the protein supplements tend to have very low isoflavone content (as the isoflavones remain in the alcohol solution). Typically around 0.05mg/g or 4% of the initial value of the soybean.
For men worried about isoflavone content and possible estrogenic side effects, soy protein concentrate is the option to seek. Caution should be taken to ensure that the selected brand of soy protein concentrate has been heat treated to eliminate trypsin inhibitors, as they may still be active.
Soy protein isolate is a process where the 'white flakes' are subject to centrifugation and pH neutralization in order to extract the highest percentage protein by weight.
This process does not undergo further ethanol extraction, and thus still has isoflavone content. Due to the first hexane extraction where oil was removed from the soybean (prior to the formation of 'white flakes'), the Soy Protein Isolate does not have 100% of the isoflavone content of soybeans, but around 38-46% (or 0.5-0.6mg/g).
Again, there is no assurance that trypsin inhibitors have been controlled for via heat treatment. If buying soy protein isolate, be cautious of whether or not the protein has had its trypsin profile deactivated.
Soybeans (raw or roasted) and texturized soy protein tend to preserve all the isoflavone content.
Products that undergo some processing or are otherwise made from processed soy (excluding soy protein concentrate) have a variable isoflavone content of 20-45%, or 0.03-0.06mg/g weight.
Secondary soy proteins (products made from soy protein concentrate) tend to have about 2% of the initial isoflavone content of soybeans, as they are made through soy protein concentrates which are 'cleansed' of isoflavones with an alcohol extract. These tend to be products that are made with soy but do not resemble soy such as soy burgers, soy milk, and protein bars made with soy protein concentrate.
Soy, as isoflavones in protein, have been investigated numerous times for their lipoprotein lowering abilities. They seem to have an ability to lower LDL and Total Cholesterol when consumed at around 25g soy protein isolate daily. For the most part these studies are conducted in post-menopausal women with or without high blood lipids, although the same effects are seen in pre-menopausal women with normal lipid levels and in men. Those with higher cholesterol levels to start tend to see more drastic drops in cholesterol relative to those with normal cholesterol. These effects seem to be placed causatively on the isoflavones, primarily genistein and daidzein, as they can be replicated with supplementation.
It should be noted that a very high standard deviation (variation between people) exists. As noted in two reviews many studies find clinically significant results, but they are too variable to reach statistical significance. This means that although benefit was seen in most persons, it was very different between persons and the researchers could not use statistical analysis to prove that it was actually the soy intake that was reducing cholesterol and LDL.
In postmenopausal women, an elevation of LDL (commonly referred to as the bad cholesterol) and total cholesterol occurs due to less estrogen. In this population, soy isoflavones can mimic the effects estrogen should be doing and reduce the otherwise expected increase. Unfortunately, this is not a reliable action and varies significantly from one person to another
The 25g of soy protein a day is consistent with the claim that "Soy protein can reduce occurrence of cardiovascular disease" which is approved by the FDA for products with more than 6.25g soy protein. This claim may only have validity when applied to true soy foods however, as foods made from soy protein concentrate have minimal isoflavone content and soy derivatives a lesser content.
Basically, it seems to be that there is benefit to heart health (via reducing LDL) and that this benefit is most studied and drastic in post-menopausal women. It also seems that there is high individual variation, and that intervention may not be as reliable as one would like. There does seem to a beneficial trend mixed between the insignificant and significant results.
This variation may come from differences in colonic bacteria, which cleave a good portion of genistein and daidzein from their sugars (which keep them inactive until needed) and further metabolism of daidzein to equol, which is more estrogenic. Only about 33-50% of people previously studied have been found to possess the strain of bacteria that does the latter conversion. For those that lack the strain, a pharmacokinetic problem exists. You get less of a dose of the active ingredient than people who do have this strain, and your results would be less significant.
Although most cholesterol lowering effects appear to be vicarious through isoflavones acting like estrogen, it may not be the only mechanism. Some peptides found in soy proteins have been shown to regulate cholesterol homeostasis in vitro and may be a viable alternate hypothesis.
Most of the benefits of soy and a reduced risk of Heart Disease come from Equol, which doesn't get produced in everybody. It still does appear to work in people that do not produce equol, but in these persons the magnitude of benefit is less
Either way, regular check-ups with a physician to assess LDL and Total cholesterol levels (and see if it's actually working) would be good.
When soy is used as a food product, there do not appear to be large reductions in testosterone for the most part. Supplementation with Soy Protein shares this same notion, and the trend suggests that does not seem to adversely affect testosterone. It should also be noted that many studies on whey or casein protein supplements will use soy protein as a control; many of the studies cited on those pages which use soy protein as control can be used as indirect evidence for the inefficacy of soy protein supplements in lowering test.
This isn't to say that soy is not at all related to lower testosterone levels; some case studies note that drastic overconsumption of soy can result in reduced androgen levels. There is some evidence that for those who harbor the bacteria that can produce equol, that they can experience reductions in testosterone. A study that did not specifically control for Equol but noted its production also noted decreases in testosterone.
One study found decreases in testosterone independent of equol but has been criticized for its statistical analysis for including a large outlier. Another study noted decreases at high isoflavone contents, but minimal decreases at lower isoflavone contents (and did not control for equol production).
Soy does not appear to significantly affect testosterone levels in moderation (1-2 servings of soy food daily, less than 25g of soy protein from non-concentrate sources), and seems to be able to adversely affect testosterone levels when superloaded (enough so that the equol issue is moot; around 100mg isoflavones daily). Between moderation and excess, there is a grey area that is dependent on whether or not one is able to produce equol from daidzein. These numbers do not apply to soy protein concentrate, which has negligible isoflavone content from ethanol extraction.