Phytochemicals in human health and disease

As I began this body of work, there were no published experimental studies regarding the clinical or biological effects of isoflavones or lignans in humans. Phytochemicals were typically described as ‘plant toxicants’ or ‘non-nutrients’, with a reputation for adverse rather than beneficial effects and generally pertaining to livestock. ‘Phytoestrogens’ (or ‘phyto-oestrogens’) was an obscure term in the medical literature. Following my initial publication in 1990, there was an extraordinary rise in the number of publications in the phyoestrogen field alone, such that by 1996 there were more than 600 publications reported on Medline. The significance of equol, a product of gastrointestinal microbial fermentation of the isoflavones daidzein and formononetin, had long been well established in numerous animal species, particularly sheep. However, though anticipated in our early human studies, it was a minor part of this literature until recently. Overlooking the key role of gut flora in the bioavailability and biotransformation of isoflavones, and the factors permissive to this process, has led to confusion and controversy in the literature regarding inconsistent biological effects in humans. Attempting to address this issue are a series of publications in this thesis arising from a decade of collaboration with colleagues at Victoria University working in the area of food science and biological science. Following demonstration that, in vitro, certain probiotic bifido-bacteria strains could hydrolyse isoflavone glucosides and biotransform daidzein into equol, potentially yielding more bioavailable and bioactive forms, preliminary studies in-vivo suggested more consistent urinary isoflavone recovery with co-ingestion of live bifido-bacteria and soy isoflavones in the form of fermented soy milk. Also suggested was a possible change in equol-producing status in this setting. If confirmed these findings could have clinical implications. Phytoestrogens are, strictly speaking, phytochemicals with oestrogenic activity, variably defined. However, not all the diverse biological properties of these compounds are mediated through oestrogen-like actions. Furthermore some plant components appear to have effects on the reproductive system, in the absence, as yet, of a defined hormonal mechanism, as with Lepidium Meyenii (Maca). The observed effects, on anxiety and depression scores, are likely phytochemical properties independent of oestrogenic activity. Having demonstrated biological activities of certain phytochemicals in humans, as found naturally occurring in foods, and further defined the role of the gut flora, the field has now moved from the perspective of hormone-dependent diet-disease relationships to a greater understanding of the potential to modify disease expression in rare, and as well as more common, conditions. It is on this basis that I am submitting this thesis as an original contribution to the advancement of scientific knowledge. Oestrogenic activity in plants was first described in 1926 (Dohrn, et al. 1926). The subsequent wealth of data regarding such effects in numerous animal species (Price and Fenwick 1985), along with then recent publications reporting the presence of isoflavone phytoestrogens including equol, the microflora metabolite of daidzein, in human urine (Setchell, Borriello, et al. 1984), began a line of enquiry by the author from mid-1986, regarding possible biological effects in humans. Preparing for a third year undergraduate essay, I had been inspired by the concept, raised by Seely, and others, from the University of Manchester, of naturally occurring compounds in plants [phytochemicals], including those with oestrogenic activity [phytoestrogens], as potentially having effects on human health (Seely, et al. 1985). From this, I developed further hypotheses regarding the role of phytochemicals in human health, and the beginning of my research career. The hypotheses that underpin this thesis included the following:- Hypothesis 1: Phytochemicals may be biologically active in humans with potentially beneficial as well as adverse effects on health Hypothesis 2 Phytochemicals may be modified in their properties by intestinal and/or probiotic microflora Hypothesis 3 Individual phytochemicals may have multiple mechanisms of action and individual foods may contain multiple phytochemicals. Therefore observed effects of phytochemical-rich food supplementation may be complex and not predictable by a single mechanism. Postmenopausal women, potentially most sensitive to such effects, supplemented with phytoestrogen-rich foods based on recent literature, were studied and an oestrogenic effect on vaginal cytology was reported in the BMJ (Wilcox et al 1990/Chapter 1). This was the first report of oestrogenic effects of isoflavone and lignan phytoestrogens (then described as plant oestrogens) in humans. Following this publication we were approached by Tenovus Institute of Cancer Research in Cardiff, leading to collaboration with Morton and others; our findings confirmed retrospectively the presence of these compounds in the plasma of the study subjects during the intervention period (Morton et al 1994/Chapter 2). In late 1991 I was also approached by Murkies who was seeking to explore the potential effect of isoflavone-rich foods in modifying symptoms of hot flushes in postmenopausal women. I had a major role in presenting the background literature, designing the clinical study protocol, data interpretation and critical input into the manuscript as well as bringing together other key members of the research team (Murkies et al 1995/Chapter 3). Six weeks’ supplementation with soy flour, rich in isoflavones, reduced hot flushes, compared with the refined wheat flour placebo lacking significant isoflavones or lignans as confirmed on biochemical analysis. Interestingly the effects on hormonal cytology were not seen in this study. This raised the question regarding possible differences between subjects recruited for the respective studies. In my original clinical study healthy volunteers from the wider community were recruited whereas symptomatic post-menopausal women were recruited in this subsequent study. A later, longer-term study in asymptomatic post-menopausal women confirmed a positive effect on vaginal cytology (Chiechi, et al. 2003). One potential explanation was the possibility of a difference in gut microflora, and consequently the capacity to biotransform daidzein into the more potent metabolite equol, between the two groups of women. In the original study, 4 out of 12 women (33%) tested were equol producers (Chapter 2) compared with the later study with Murkies, where only 5 of a total 58 (8.6%) women were equol producers (Chapter 3). In a review of the effects of soy on menopausal symptoms, presented as an invited lecture at the 8th International Congress on the Menopause in 1996, I summarized the variable findings of the studies reported to that date (Wilcox 1997/Chapter 4). A further review written jointly with Murkies, with additional critical input by Davis, summarized the literature regarding phytoestrogens in a broader clinical context (Murkies et al 1998/Chapter 5). The pivotal role of equol, anticipated from animal data and early pre-clinical studies, was confirmed in subsequent reports by others. (Setchell, Brown and Lydeking-Olsen, The clinical importance of the metabolite equol-a clue to the effectiveness of soy and its isoflavones. 2002) In 1997, I was approached by Stojanovska, leading to subsequent collaboration together with Shah, an academic and applied food scientist with a strong track record in dairy fermentation. In 1999, Shah and I obtained Australian Research Council (in partnership with Sanitarium Health Foods) for 2001-2004: ARC-Spirt Grant (AUD $79,116, combined Scholarship & cash contribution) to develop in-vitro fermentation of soy milk with live bifidobacteria with the aim of enhancing the bioavailability of ingested isoflavones (Tsangalis et al 2004/Chapter 6). In vitro hydrolysis of isoflavone glucosides and biotransformation of daidzein to equol in fermented soy-milk containing live bifido-bacteria was associated with greater urinary isoflavone recovery in postmenopausal women (Tsangalis et al 2005/Chapter 7) and a trend to increased prevalence of equol production after 2 weeks’ supplementation (Tsangalis et al 2007/Chapter 8). These findings have formed the basis of a clinical study (to be reported) comparing the effects of fermented soy milk, non-fermented soymilk and casein over a 12 week period involving 40 postmenopausal women in a parallel arm study. The study has been designed to look at markers of oestrogen action, as well as androgen profiles and SHBG. The latter may be both influenced by oestrogenic activity as well as hepatic effects of systemic insulin resistance; Jayagopal et al (Jayagopal, et al. 2002) reported a reduction in insulin resistance in 32 postmenopausal women with type 2 diabetes supplemented with 30g daily soy protein (Jayagopal, et al. 2002), though this may have been confounded by differences in protein intake. Although early population studies (Adlercreutz H 1987) reported an association between dietary phytoestrogen intake, particularly dietary lignans, and SHBG levels, the majority of early human intervention studies (Chapter 4) failed to confirm this. One study that did (Brzezinski A 1997)used a combination of flaxseed (linseed) and soy as a phytoestrogen supplement. However Pino et al (Pino AM 2000) subsequently reported a positive effect of soy supplementation in raising SHBG in postmenopausal women that was essentially dependent on the baseline SHBG level: this effect was only observed where the baseline SHBG level was less than 55 nmol/L. Reduced SHBG is commonly seen in states of insulin resistance, as associated with central adiposity, as well with as androgen excess. (...)