Soy

What it is

Soy foods are made from soybeans, such as soya milk and whole soybeans [1]. Trials test whole soy foods, soy protein and isoflavone extracts [1]. Isoflavones are plant compounds with estrogen-like effects [2].

What the research says

For LDL cholesterol, a 46-trial pooled analysis found soy protein lowered LDL by about 4.8 mg/dL, roughly 3 to 4 per cent [3]. A 35-trial analysis found a similar drop, larger in people with high cholesterol and with whole soy foods [1]. Isolated soy isoflavone supplements had no effect on the lipid profile in that analysis [1], and LDL was not significantly lower in a meta-analysis of postmenopausal women [4].

For hot flashes, a 2015 meta-analysis estimated isoflavones cut hot flashes by about 25% after the placebo effect was removed, but the effect is slow [5]. A 2016 analysis linked soy isoflavones to about 0.8 fewer hot flashes a day [6]. The 2013 Cochrane review found no conclusive evidence of benefit [2]. A 2025 meta-analysis of 12 studies found a small overall effect on menopause symptoms but no significant effect on hot flushes [7].

For breast cancer, a study of more than 300,000 Chinese women found moderate soy intake was not linked to higher risk [8].

Doses used in studies

LDL trials used a median of 25 g of soy protein a day for about six weeks [3]. Hot flash trials used 84 to 126 mg of soy germ isoflavones a day [9], 10 to 60 mg of isoflavones a day in a soy drink [10], and 40 g of soy flour a day for six months [11]. These are trial doses, not recommendations.

Safety and interactions

Side effects: the Cochrane review found no evidence of womb or vaginal stimulation, or other adverse effects, in trials lasting up to two years, but most trials were small [2]. Upper limits and toxic doses: the studies we read did not identify a toxic dose, and we found no upper intake limit to quote.

Thyroid: a 60-person trial found overt hypothyroidism in six people after a 16 mg daily soy phytoestrogen dose [12]. A small 12-woman study found a soy isoflavone supplement did not change levothyroxine absorption [13], and a review found any effect likely small [14]. Breast cancer: higher intake was not linked to breast cancer risk [8]. In women already diagnosed, pre-diagnosis intake was linked to a small survival gain after menopause, but the evidence is limited [15].

Pregnancy and breastfeeding: we found no trials of soy supplements in these groups. Kidney, liver and bleeding disorders were not addressed in the studies we read; ask a doctor if you have these conditions.

Caveats and open questions

Results between trials vary widely, with heterogeneity of 92 to 99% in one 2015 analysis [1]. Many menopause trials were small and at high risk of bias, and placebo groups improved by between 1% and 59% [2][6]. Several menopause products mix soy with other herbs, so their results do not show what soy alone does [16][17]. Mood: one trial found no effect on everyday mood [10], while a meta-analysis found a small drop in depression [18]. One 2023 meta-analysis of isoflavone supplements found lower triglycerides and slightly higher HDL in women under 65, with no LDL result in its abstract [19].