Soy Supplements and Cognitive Function

By Mary L. Hardy, MD

This year, an estimated 1.35 million U.S. women will become menopausal. The most common symptom experienced by these women besides cessation of menses is hot flashes.1 Much less attention is paid to the cognitive symptoms of menopause, but for many women changes in memory and "ability to think straight" can be debilitating. More than half of mid-life women (62% in one study) report subjective changes in memory, which were worse in perimenopausal than postmenopausal women.2 Although positive correlations were seen in the Seattle Midlife Women’s Health Study between severity of cognitive complaints and stress/depressed mood, estrogen decline itself is thought to have a negative effect on cognitive function.

A number of studies have shown a protective effect of hormone replacement therapy (HRT) on cognitive decline and developing Alzheimer’s disease.3 Maki’s systematic review of clinical and observational trials examining HRT’s effect on the cognitive function of perimenopausal and menopausal women found the strongest data for protection against the development of Alzheimer’s dementia (AD). No benefit was demonstrated for women already suffering from AD. The data in women without prior cognitive defects are more difficult to analyze due to variability in subjects and testing methods, but many of the individual trials cited in this review are encouraging.

Experts recommend soy foods and products, especially those rich in isoflavones, for menopausal patients primarily for soy’s effects on vasomotor symptoms. Isoflavones, also called phytoestrogens, are thought to have weak estrogenic activity.4 Unfortunately, data to support soy’s use for the relief of the cognitive symptoms of menopause were not considered strong enough for some experts to make a recommendation.5 However, the observational data in women taking HRT suggest that there may be a possible mechanism for the isoflavones in soy to exert an estrogen-like protective effect on cognitive dysfunction, especially the development of AD. In fact, one primate model shows decreases in an abnormal brain protein, hyperphosphorylated protein tau, which is implicated in the development of AD.6 Thus, the known activity of isoflavones, the epidemiologic data suggesting that estrogen can protect cognitive function in menopausal women, and animal data suggesting a biologic mechanism of action, all sustain interest in the use of soy products for support of cognitive function in mid-life women. Few studies have been conducted in this area. However, two recent articles, described below, were performed very carefully.

In the first study, Duffy et al tested a soy isoflavone supplement (60 mg/d isoflavones) in 33 postmenopausal women.7 The study was blinded, randomized, and placebo-controlled—so the methodology was sound. Patients were tested not only for their cognitive function, but also for things that could confound cognitive performance such as mood, anxiety, or sleepiness. Groups had equivalent IQs and educational attainment. There were minor differences in the two groups at baseline based on assessment of their diets. The treatment group ate slightly more refined sugar, took more calcium, and drank less alcohol. The alcohol consumption in the placebo group also was quite low, and none of these differences were thought to be clinically significant. Both groups were asked to refrain from eating soy-containing foods during the trial.

Although the trial duration was short (12 weeks), significant improvements were seen in the active treatment groups. Treated subjects did better in tasks requiring sustained attention, recall of pictures, and planning a task. Learning new rules seemed to be equal in both groups, suggesting that new learning was less affected, but the manipulation of that new learning (learning rule reversals) was better in the treated group. This improvement was independent of any changes in menopausal symptoms, mood, or sleepiness as none of these parameters changed. The differences in the two groups were not always statistically significant and varied even within groups of related tests. It could be argued that the improvements noted were the result of "learning to take the test," but the treatment group in general tended to do better. The kinds of functions that demonstrated improvement included memory recall and task planning. So, even in a short intervention with a moderate dose of isoflavones, a difference can be demonstrated in postmenopausal women’s cognitive function.

Kritz-Silverstein and her colleagues tested a similar population but used a higher dose of isoflavones (110 mg/d) for a longer time (6 months).8 Both groups were equivalent at baseline, as above. Mood was assessed, but neither sleepiness nor menopausal symptoms were assessed. Food patterns were checked, but most women did not consume significant amounts of soy foods during the trial. Compliance was checked carefully in this study and was very good. The trial was a randomized, double-blind, placebo-controlled trial, so again the methodology was sound. The memory tests used were not the same as the ones used previously, but similar types of functions were tested.

As demonstrated in the shorter study, treated subjects performed better both with respect to their own baseline and to the performance of the placebo patients. Specifically, statistically significant results were obtained in tests of verbal memory (category fluency), and non- statistically significant improvements were found for tests of visual motor tracking and attention. A cruder measure of cognitive function used in routine clinical practice, the Mini-Mental Status Examination (MMSE), did not change during the trial. This screening tool is designed to reveal the presence of dementia, not milder cognitive dysfunction. When the results were analyzed comparing the results of younger vs. older patients, the older patients showed a larger benefit. Again, some of the improvement observed in both groups could be attributed to learning effects (the fact that subjects perform better the second time they do a task than the first), but still differences that exceeded this learning effect were demonstrated between the groups.

In terms of applying these studies to our own patients, a few caveats. First, most of the women tested were fairly well-educated and in the second trial represented a very high socioeconomic status. These are the kind of patients who generally use supplements, but we don’t know if the effects of the supplements will translate fully to different groups. Soy supplementation will not likely correct cognitive difficulties caused by depression, dementia, or other medical conditions, so these conditions should be ruled out before soy is tried. Also, all the women tested here were postmenopausal and there may be differences in the results for perimenopausal women.

The studies used pills that contained isolated isoflavones in moderate to fairly high ranges. These were obviously active and the isoflavones are more likely to contribute to cognitive protection. However, it is always beneficial to try to use soy in a manner that also includes the protein component, as soy protein is a heart healthy food and has been shown to lower cholesterol. Some concern has been raised about using high doses of isolated isoflavones, but for the average patient, dosages in the ranges quoted here should be fine. More caution should be exercised in patients who are at high risk of adverse effects from increased exposure to the estrogenic effects of soy isoflavones. No risk has been directly proven, but this is another reason for the use of whole soy foods as opposed to isolated supplements.

Finally, it is important to make sure that our patients have accurate expectations for their soy intervention. Effects on menopausal symptoms are likely to be mild at best. Memory should improve, but the effects are subtle. Patients should be clear about having an endpoint to "test." Ask them to rate some activity they do regularly so that they can see the benefit over time. Remind them that these interventions work best in the context of a healthy lifestyle and take time to see maximal benefit. If the patient is eating soy as a food, then the additional beneficial effects on heart and bone health can accrue as well. In this context, it seems that the addition of isoflavone-rich supplements and/or foods could be a useful treatment for a nagging concern of women during menopausal transitions and should be discussed with women complaining of memory problems.


1. North American Menopause Society. Menopause Guidebook. Available at Accessed Oct. 10, 2003

2. Mitchell E, Woods N. Midlife women’s attributions about perceived memory changes: Observations from the Seattle Midlife Women’s Health Study. J Wom Health Gender Based Med 2001;10:351-362.

3. Maki P. HRT and cognitive decline. Best Pract Res 2003;17:105-122.

4. Use of botanicals for management of menopausal symptoms. ACOG Practice Bulletin. June 2001. Available at Accessed Oct. 8, 2003.

5. Vincent A, Fitzpatrick LA. Soy isoflavones: Are they useful in menopause? Mayo Clin Proc 2000;75:1174-1184.

6. Kim H, et al. Attenuation of neurodegeneration-relevant modifications of brain proteins by dietary soy. Biofactors 2000;12:243-250.

7. Duffy R, et al. Improved cognitive function in postmenopausal women after 12 weeks of consumption of a soya extract containing isoflavones. Pharmcol Biol Behavior 2003;75:721-729.

8. Kritz-Silverstein D, et al. Isoflavones and cognitive function in older women: The soy and postmenopausal health in aging (SOPHIA) study. Menopause 2003;10:196-202.