Abstract
Purpose
We evaluated the association between total soy, soy product (natto, miso and tofu) and isoflavone intake and incident disabling dementia in a Japanese population.
Methods
We conducted a population-based prospective study in 18,991 men and 22,456 women. Intake of soy products and isoflavone was calculated using a validated food frequency questionnaire when participants were 45–74 years old (1995 and 1998). Incident disabling dementia was defined by the daily living disability status related to dementia in the long-term care insurance program of Japan from 2006 to 2016. Multivariate hazard ratios (HRs) and 95% confidence intervals (CIs) of disabling dementia were calculated by quintiles of total soy, individual soy product and isoflavone intake, using Cox proportional hazard regression models.
Results
Total soy product intake was not associated with disabling dementia risk in both men and women. By individual soy products, natto intake was marginally inversely associated with disabling dementia in women (trend P = 0.050). When we stratified by age, this inverse association was clearer in women aged under 60 years (multivariate HR for the highest versus lowest quintile was 0.78, 95% CI 0.59–1.04, trend P = 0.020 for those aged under 60 years and 0.90, 95% CI 0.77–1.05, trend P = 0.23 for those aged 60 years and older, respectively). Any soy product or isoflavone intake was not associated with disabling dementia risk in men.
Conclusions
Although total soy product intake was not associated with disabling dementia risk, natto intake may contribute to reducing the risk of disabling dementia in women, especially in those aged under 60 years.
Similar content being viewed by others
Introduction
With the worldwide aging of the population, dementia is a growing global problem. 55 million people currently have dementia, and around 10 million new cases are diagnosed every year [1]. Dementia is related to lifestyle, including eating habits [2], and therefore subject to prevention.
Soy products are widely consumed in Asia [3]. They are rich in isoflavones, which may protect cognitive function [4] or Alzheimer's disease (AD) [5]. A previous meta-analysis study reported that isoflavone intake improved cognitive function and memory in randomized controlled trials [6]. Therefore, we hypothesized that soy product intake may decrease the risk of disabling dementia. So far, several studies have examined the association between soy product intake and risk of cognitive function. High soy product intake was associated with decreased risk of cognitive impairment in prospective cohort studies [7, 8], while another study found no association in cross-sectional study [9]. Regarding dietary patterns, a healthy dietary pattern which included soy products was associated with decreased dementia risk in elderly Japanese [10]. Among individual soy products, high tofu intake increased [11, 12] or decreased [13] the risk of cognitive impairment, whereas fermented soy product intake decreased [13, 14] or was not associated [8] with risk. Regarding fermented soy products eaten in Japan, natto contains bioactive components, such as polyamine [15] and nattokinase [16]: polyamine (spermidine) has been shown to protect against neurodegeneration and cognitive decline [17], and nattokinase protects the hippocampus against β-amyloid-induced neuronal damage [18, 19] in animal experiments. In addition, the Japan Public Health Center-Based Prospective Study (JPHC Study) showed that natto intake was inversely associated with risk of hypertension or stroke [20, 21], which are considered risk factors of dementia. Consumption of soy products, particularly natto, may therefore have protective effects against dementia. The only prospective study from Japan reported that natto intake was not associated with cognitive impairment [8]. However, the investigation area in that study was limited to a single prefecture and the participants were a relatively small number of elderly people.
Accordingly, the association between soy product intake and cognitive function remains inconsistent. In addition, to our knowledge, no study has investigated the association between soy product intake and risk of incident dementia.
We collected a large number of disabling dementia cases from the certification of Japan’s national long-term care insurance (LTCI) system. We defined the outcome of this study as dementia that required care (disabling dementia) under the LTCI system, which provides national compulsory insurance for all individuals aged 40 years or older in Japan. Disabling dementia was assessed not only from cognitive impairment but also by a comprehensive review of health condition, including evaluation by a primary care physician.
Here, we used this information to evaluate the association of total soy and individual soy product intake and disabling dementia in a middle-aged Japanese population, in whom consumption of soy products is high compared to Western countries.
Materials and methods
Study population
The JPHC Study consists of two subcohorts: cohort I [started in 1990, five public health centers (PHCs) of Iwate, Akita, Nagano, Okinawa and Tokyo prefectures, and participants aged 40–59 years], and cohort II (started in 1993, six PHCs of Ibaraki, Niigata, Kochi, Nagasaki, Okinawa and Osaka prefectures, and participants aged 40–69 years). Details of the study design have been described elsewhere [22, 23]. The JPHC study is consisted of 140,420 Japanese participants (68,722 men and 71,698 women). We conducted surveys at baseline, 5- and 10-year follow-up, in which participants were asked to complete self-administered questionnaires about their lifestyle and medical history. Because the 5-year follow-up survey included a more detailed food frequency questionnaire (FFQ) than the baseline survey, we included participants who responded to the 5-year follow-up survey in this analysis. Participants were informed of the objectives of the study, and that completion of the survey questionnaire was regarded as providing consent to participate. The study was approved by the National Cancer Center Japan institutional review board (Number: 2001-021, 2015-085, 2015-312).
This study included 62,401 participants from eight areas in which follow-up data on disabling dementia were available, namely Omonogawa and Yokote districts in Yokote city in Akita prefecture; Tomobe district in Kasama city and Iwase district in Sakuragawa city in Ibaraki prefecture; Usuda district in Saku city in Nagano prefecture; Kagami and Noichi districts in Konan city in Kochi prefecture; and Gushikawa district in Uruma city in Okinawa prefecture.
Disabling dementia cases and follow-up
We defined the outcome as dementia that required care (disabling dementia) under the LTCI system, which is administered by the Ministry of Health, Labor and Welfare Bureau of Japan. The LTCI system was initiated in 2000 for residents aged 65 years and older and for those with disabilities aged 40–64 years, and is administered by municipalities. The municipality assesses applications using two documents, namely a comprehensive assessment of the applicant’s functional health condition, and a written opinion about the applicant’s disability status by a primary care physician. We defined disabling dementia as certification of any level of need for long-term care, and a written opinion of the grade of cognitive disability (grade IIa, IIb, IIIa, IIIb, IV or M). Grades of cognitive disability were no dementia; some dementia but almost independent in daily life (level I); dementia with some difficulty in communicating but with independence in daily living with minimal observation (level IIa and IIb); dementia with some difficulty in communicating and a need for partial care (level IIIa and IIIb); severe dementia with difficulty in communicating and a need for complete care (level IV); and significant psychiatric symptoms, behavioral problems or serious physical illness and need for specialized medical care (M) [24]. These criteria of disabling dementia were validated against dementia diagnoses by neuropsychiatrists: the sensitivity was moderate (73%), and the specificity was high (96%) [25]. Additionally, grades of cognitive disability have been well-correlated with the Mini-Mental State Exam (MMSE) score (Spearman rank correlation coefficient − 0.736) [26].
In this study, certification records in the LTCI system were available from 1 January 2006 to 31 December 2016. Therefore, the follow-up start point of this study was defined as January 1, 2006. We followed eligible participants until 31 December 2016.
Soy product intake
We used the JPHC Study food frequency questionnaire (FFQ) data from the 5-year follow-up survey conducted in 1995–1998. The survey included questions about the frequency and portion size of about 138 food and beverage items [27, 28]. Standard portion sizes for each food item were small (50% less than the standard serving size), medium (proportional to standard serving size), and large (50% larger than the standard serving size). The nine frequency categories for each food item were never, 1–3 times/month, 1–2 times/week, 3–4 times/week, 5–6 times/week, 1 time/day, 2–3 times/day, 4–6 times/day, and ≥ 7 times/day; and for drink items were never, 1–2 times/week, 3–4 times/week, 5–6 times/week, 1 time/day, 2–3 times/day, 4–6 times/day, 7–9 times/day, and ≥ 10 times/day. The dietary intake of each food item was calculated by multiplying frequency by the standard or relative portion sizes for each food item, excluding soy milk, green tea and miso: intake of soymilk and green tea was calculated by multiplying frequency by the standard portion, and because miso is a seasoning used in miso soup and accounts for 8% of the amount of miso soup, miso intake was calculated by multiplying the amount of miso soup by 0.08. The Spearman’s correlation coefficients between the energy-adjusted intake of total soy products from the FFQ and those from dietary records were 0.53 for men and 0.49 for women in cohort I [29], and 0.52 and 0.54 in cohort II [30]. The amount of total soy products was the sum of tofu, yushidofu, koya-dofu, abura-age, natto, miso, and soymilk items from the FFQ. Tofu is the sum of tofu, yushidofu and koya-dofu items from the FFQ. The amount of total, individual soy products and isoflavones was adjusted for total energy intake using the residual method [31]. Because soy products are isoflavone-rich, we also calculated total isoflavone intake, including daidzein and genistein. Daidzein and genistein intakes were estimated using the values from each soy product in a specially developed food composition table for isoflavones in Japanese foods [32, 33]. The Spearman’s correlation coefficients between the energy-adjusted intake of isoflavone from the FFQ and those from dietary records were 0.65 for men and 0.56 for women in cohort I, and 0.53 and 0.45 in cohort II, respectively (unpublished data).
Statistical analysis
We included 62,401 participants in this study. We excluded non-Japanese participants (n = 51), as well as those with an incorrect birth date (n = 3), duplicate registration (n = 4), refusal to follow-up (n = 11), and those who moved out or died before the start of follow-up (n = 11,656), leaving 50,676 participants as eligible for analysis. We further excluded those who did not answer the 5-year follow-up questionnaire survey (n = 6637), those who left the frequency of soy products intake unanswered (n = 414), and those who reported total energy intake at the upper or lower 2.5% ends of the range (1030 and 4169 kcal for men and 882 and 3640 kcal for women, respectively). Finally, 41,447 participants (18,991 men and 22,456 women) were included in the analysis (Supplemental Fig. 1).
We divided participants into quintiles of intake for total soy and individual soy products by sex, except for natto only in men, and used the lowest quintile group as reference. Regarding the 4,208 men who answered that they did not eat natto, we categorized these into the first fifth, and divided the remaining men into quartiles. Sex-specific, mean values and prevalence of selected factors were calculated and compared by total soy product intake using linear or logistic regression analyses. Dietary intake was adjusted for energy intake using the residual method except for total energy and green tea, and P values were calculated from the median value of each food by total soy product intake using the Kruskal–Wallis test.
Models were adjusted for confounding factors. Model 1 was adjusted for age (continuous) and stratified by 8 areas. Model 2 was additionally adjusted for body mass index (< 18.5, 18.5–24.9, 25.0–29.9, or ≥ 30 kg/m2, missing), smoking status (never, former, < 20 cigarettes/day, and ≥ 20 cigarettes/day, missing), alcohol intake (0, 1–149, 150–299, or ≥ 300 ethanol g/week, missing), total metabolic equivalents (hours per day, quartiles, missing), history of cancer (yes or no), history or medication for diabetes (yes or no), medication for hypertension or hypercholesterolemia (yes or no), menopausal status (yes, no, missing; women only), health checkup in the past year (yes or no), job status (employed, no, missing), living status (living with others, alone, missing) and dietary intakes of total energy (quintiles), vegetables (quintiles), fruits (quintiles), fish (quintiles), green tea (quintiles) and sodium (quintiles). Missing values for dietary intake excluding total energy were given a value of 0. We created dummy variables for covariates and missing data. Trends were calculated by assigning median value to categories of each soy product intake. Additionally, as participants in the previous studies from Japan [8, 10] which reported associations between soy products and cognitive function or dementia were aged 60 years and older, we stratified participants by age at the time of the 5-year questionnaire follow-up survey. As a supplementary analysis, to consider types of dementia, disabling dementia cases were classified into two categories: with or without a history of stroke, following systematic stroke registration confirmed according to the criteria of the National Survey of Stroke [34] as well as self-reports at the baseline, 5- and 10-year follow-up questionnaires. The end of the follow-up period was set for the end of 2009 for Yokote, Saku and Uruma areas, and for the end of 2012 for Kasama, Sakuragawa and Konan areas, with consideration to the availability of stroke registry data. In addition, using the dataset of a supplementary analysis, we examined the mediating role of stroke incidence between soy product intake and disabling dementia risk the follow-up period (2009 or 2012). These analyses were carried out using multistate models, an extension of competing risk survival analysis, which allows the analysis of various transition rates from an initial state to intermediate events and subsequently to final states in the same framework [35]. In particular, these models were used to simultaneously estimate the risk associated with soy product intake in three transitions: (1) from healthy state to incident dementia in those free of incident stroke; (2) from healthy state to incident stroke; and (3) from incident stroke to incident dementia. Age was used as the follow-up starting point (2006 or 2009), and the analyses were adjusted for covariates in multivariate-adjusted models.
Person years of follow-up were calculated for each participant from the follow-up start point until the date of diagnosis of disabling dementia, migration from a study area to a non-study area, date of death, or the end of follow-up, whichever occurred first. Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs).
We confirmed the proportional hazards assumption according to total soy product intake × time interactions and found that the assumption was not violated. All probability values for the statistical tests were 2-tailed, and probability values below 0.05 were considered significant. For each set of analyses shown in the tables of results, multiple comparison was accounted for by Bonferroni correction. All statistical analyses were performed using the SAS software, version 9.4. (SAS Institute, Cary, NC, USA) and R software, version × 64 4.1.2; in particular, the multistate model analyses were performed using the “mstate” package.
Results
During a mean follow-up of 9.4 years, 4911 cases of disabling dementia (1996 men and 2915 women, respectively) were observed. The baseline characteristics of participants according to total soy product intake are shown in Table 1. Participants who consumed soy products tended to be older, have a higher proportion of treatment for hypertension or hypercholesterolemia and consume more vegetables in both men and women.
We observed no association between total soy product intake and disabling dementia in both men and women: the multivariate HRs (95% CI) for the highest versus lowest quintile were 1.07 (95% CI 0.92–1.24, P for trend = 0.25) for men and 1.04 (95% CI 0.92–1.18, P for trend = 0.44) for women (Table 2). Among individual soy products, natto intake was marginally inversely associated with risk of disabling dementia in women: the multivariate HR (95% CI) for the highest versus lowest quintile was 0.89 (95% CI 0.78–1.02, P for trend = 0.050). The highest versus lowest quintile of miso intake was associated with increased risk of disabling dementia: the multivariate HR for the highest versus lowest quintile was 1.16 (95% CI 1.02–1.33, P for trend = 0.16). On multiple hypothesis comparison for the interaction term of exposures, natto intake was not significant in women. No soy products were associated with risk of disabling dementia in men. Isoflavone intake was not associated with risk of disabling dementia in both men and women. When we stratified by age, participants aged under 60 years showed a stronger association between natto intake and risk of disabling dementia than those aged 60 years and older in women, but not in men (Tables 3, 4). In women, the multivariate HRs for the highest versus lowest quintile of natto intake were 0.78 (0.59–1.04, P for trend = 0.020) for women aged under 60 years and 0.90 (0.77–1.05, P for trend = 0.23) for those aged 60 years and older. Other associations between intake of soy products or isoflavones and disabling dementia were similar to the main results in both men and women.
The association between intake of soy products or isoflavones and disabling dementia was similarly observed when we analyzed the data for cases of disabling dementia with or without a history of stroke in both men and women (Supplemental Tables 1 and 2). The multivariate HRs for the highest versus lowest quintile of natto intake were 0.75 (0.50–1.11, P for trend = 0.21) for cases of disabling dementia with a history of stroke and 0.81 (0.65–1.01, P for trend = 0.017) for those without a history of stroke in women (Supplemental Table 2). In multistate model analyses assessing the effect of natto intake on the transitions between the various states, the multivariate-adjusted HRs associated with the highest quintile of natto intake in women were 0.78 (0.64–0.96) for dementia occurrence in healthy individuals; 1.01 (0.69–1.47) for stroke occurrence in healthy individuals; and 0.54 (0.23–1.27) for dementia occurrence in individuals who previously suffered an incident stroke event (Supplemental Fig. 2).
Discussion
In this study, we found that intake of total soy products and isoflavones was not associated with the risk of disabling dementia in both men and women. For individual soy products, natto intake was marginal inversely associated with risk of disabling dementia in women, whereas no soy product was associated with this risk in men. This is the first cohort study to reveal an association between intake of total soy, individual soy products or isoflavones and risk of incident disabling dementia in middle-aged Japanese men and women.
We showed a marginal inverse association between natto intake and disabling dementia in women. Natto contains isoflavone, whose intake has shown protective effects on cognitive function and AD via a reduction in the level of inflammation and oxidative stress [4]. In addition, isoflavone intake induced the degradation of amyloid-β protein via peroxisome proliferator-activated receptor gamma/apolipoprotein E activation, and decreased tau phosphorylation, which causes the formation of neurofibrillary tangles [5]. However, given the lack of any association between other soy products and isoflavone and disabling dementia in this study, the mechanism likely relates to the unique components and characteristics of natto. Nattokinase, an enzyme contained in the sticky component of natto, shows amyloid-degrading ability in vitro [18]. In addition, oral administration of nattokinase in a rat model of AD showed a significant decrease in brain levels of apoptotic and inflammatory factors, and an increase in brain-derived neurotrophic factor and insulin-like growth factor-1 levels when compared with untreated AD-induced rats [19]. In Alzheimer disease-induced rats, a high dose of nattokinase resulted in the disappearance of amyloid plaques in the hippocampus [19]. Natto is also rich in polyamine [15]. Drosophila fed with polyamine were protected against age-induced memory impairment [17]. Natto intake may protect the hippocampus against β-amyloid-induced neuronal damage. In previous cohort studies [20, 21], furthermore, natto intake was associated with a decreased risk of hypertension and stroke, and may have prevented vascular dementia. Regarding other reasons, participants who responded that they did not or only rarely consumed natto may have had poor masticatory function. Poor masticatory function has been associated with risk of dementia in elderly people [36]. Most soy products have a soft texture, including tofu (soybean curd) and miso (soybean paste), but natto is made by fermenting steamed whole soybeans using Bacillus subtilis [37]. A previous study reported that masticatory function increased regional cerebral blood flow [38, 39]. Moriya et al. reported that poor self-assessed masticatory status was associated with poor verbal and visual immediate explicit memory function [40]. To date, only one prospective study from Japan reported no association between natto intake and cognitive function in 403 men and 373 women aged 60–81 years [8]. When we stratified by age, natto intake was significantly inversely associated with disabling dementia in women aged under 60 years, but not in those aged 60 years and older (Tables 3, 4). We considered that this study design allowed us to eliminate the effect of aging or cognitive impairment on dietary intake, because exposure of soy product intake was obtained in middle-aged participants, especially for under aged 60 years. In this study, there was a higher percentage of medical history in participants aged 60 years and older than in those under 60 years in women: respective proportions were 32.4% and 12.8% for medication for hypertension; 13.9% and 5.2% for medication for hypercholesterolemia; and 5.3% and 2.6% for a history of diabetes mellitus. Although we adjusted for these medical histories, other risk factors affected disabling dementia risk more strongly than these variables, and may have attenuated the association between natto intake and dementia risk in older participants. Although the cause of this difference is not clear, the effect of natto intake in reducing risk of disabling dementia may be stronger in younger than older people.
We also found that the inverse association between natto intake and risk of disabling dementia was found only among women. Although the prevalence of smokers and drinkers was higher among men than women, we found no change in the results by smoking and heavy drinking status when we stratified by former/current or never smokers and over or under 300 ethanol g/week drinkers in men. Moreover, although we adjusted for smoking or drinking status, we considered that other residual confounding remained. The results of this study are similar to those of a previous study in which soy isoflavone intake was associated with decreased cognitive impairment in women only, although it was not associated with natto [8]. Accordingly, the meaning of the sex difference in the results remains unclear and further studies are needed.
In this study, only the highest quintile of miso intake was associated with an increased risk of disabling dementia, and no significant trend was observed among women. The salt content of miso is high compared to other soy products, particularly when considered with regard to frequency of intake. While high salt intake increases the risk of stroke [41], considered a risk factor of dementia [42], our previous prospective study reported that miso intake was not associated with incidence of high blood pressure [20] or cardiovascular disease [21]. The reason for the tendency toward a positive association between miso intake and risk of disabling dementia is unclear, and further studies are needed.
A prospective study showed that intake of total soy product or isoflavone was not associated with cognitive function in Chinese men and women [9]. Consistent with that study, we also found no association between intake of total soy products or isoflavones and risk of disabling dementia in either men or women. Two prospective studies reported that soy product intake was associated with decreased risk of cognitive function [7, 8]. An et al. found that soy product intake was inversely associated with risk of cognitive impairment in a cognitively intact Chinese population aged 80 years and older at baseline [7]; however, this finding considered attributable to selection bias. Another study from Japan reported that intake of total soy products, soybeans and isoflavones was associated with a significantly decreased risk of cognitive impairment in women [8]. According to Ozawa et al. [10], a healthy dietary pattern which included soy products was associated with decreased dementia risk in participants aged 60 years and older, but this result might have been affected by soy and other healthy foods. Regarding isoflavone intake, we did not include soybeans, which are richer in isoflavones than tofu and miso [32], in total soy product intake, nor glycitin (an isoflavone) in isoflavone intake, despite its inclusion in a previous study [8]. Although a previous meta-analysis study reported that isoflavone improved cognitive function in randomized control studies, only two of the included studies were from Asia [6]. In addition, because the dietary intake of isoflavone is higher in Asia than in Western countries [43,44,45], further studies are needed to confirm its effects on cognitive function in Asian people.
The strengths of this study include its use of middle-aged participants at the survey point of exposure, its large sample size, long follow-up, use of a validated FFQ and adjustment for potentially important confounding factors, including smoking status, alcohol intake, and other food items. In addition, disabling dementia was validated by a clinical assessment for dementia. The long follow-up period is likely to allow estimation of a robust association between soy product intake and risk of disabling dementia. Despite these strengths, several limitations warrant discussion. First, disabling dementia was originally established for the Japanese system, and this outcome is not diagnosed by psychiatrists. However, the validation of disabling dementia criteria was confirmed by comparison with the clinical diagnoses of neuropsychiatrists with moderate sensitivity and with high specificity [25], and grades of cognitive disability were well-correlated with the MMSE score [26]. Second, we could not classify disabling dementia into Alzheimer's or vascular types. Although the number of cases was relatively small, the association of total soy, individual soy product and isoflavone intake was similarly observed with disabling dementia regardless of a history of stroke in both men and women (Supplemental Tables 1 and 2). Therefore, natto intake may protect against disabling dementia regardless of dementia type in women. Based on the multistate model analyses, we found no evidence to suggest that the effect of natto intake on dementia occurrence could be partly due to an effect of natto intake on the occurrence of stroke in women (Supplemental Fig. 2). Third, although participants may have changed dietary habits during follow-up, soy product intake was assessed only one time. Any such dietary change might have attenuated the association shown in the present study. However, Spearman rank correlation coefficients of the estimated energy-adjusted intake of soy products between two FFQs administered at a one-year interval were moderate: 0.64 for men and 0.67 for women in cohort I; and 0.57 and 0.44 in cohort II [30, 46]. Fourth, we did not have information on cognitive function or dementia status among participants at the follow-up start point. However, exposure information was collected around 10 years before the incidence of disabling dementia. In addition, similar results were obtained when analysis was limited to data collected around 5 years before the incidence of disabling dementia, and even after the exclusion of subjects who needed caregivers in activities of daily living, natto intake was still inversely associated with the risk of disabling dementia in women: the multivariate HR (95% CI) for the highest versus lowest quintile of natto intake was 0.84 (0.72–0.98, P for trend = 0.031) (data not shown in Table). Moreover, considering that the onset of disabling dementia occurs over a long time, the first year of registration may include participants who already had dementia. The results did not substantially change when cases diagnosed within the first year of follow-up were excluded in men. The marginal inverse association between natto intake and disabling dementia risk was attenuated: the multivariate HR (95% CI) for the highest versus lowest quintile was 0.93 (0.81–1.07, P for trend = 0.19), but women aged under 60 years still showed a marginal inverse association with risk, with a multivariate HR (95% CI) for the highest versus lowest quintile of 0.82 (0.62–1.10, P for trend = 0.054) in women aged under 60 years (data not shown in Table). Furthermore, education level also may be a confounding factor, but we did not adjust this for all subjects, because we only had information about education level in cohort I (n = 19,148, Yokote, Saku and Uruma areas). Nevertheless, when we adjusted for education level (junior high school or less, high school, and university or higher education, missing) in cohort I participants only, the inverse association between natto intake and disabling dementia risk was not changed (data not shown in Table). Fifth, the risk of disabling dementia with the lowest quintile of natto intake might be overestimated because some participants were prescribed warfarin and were prohibited from eating natto because of its high concentration of vitamin K [47]. These participants were likely to have had atrial fibrillation or valvular disease which in turn likely enhanced the risk of stroke and vascular dementia. However, when we excluded participants with a history of cardiovascular disease from 5-year follow-up questionnaires (stroke, myocardial infarction or angina, n = 460 excluded), natto intake still showed an inverse association with the risk of disabling dementia in women: the multivariate HR (95% CI) for the highest versus lowest quintile was 0.86 (0.75–0.99, P for trend = 0.011).
In conclusion, although further studies are warranted to confirm the generalization of these results, natto intake may contribute to reducing the risk of disabling dementia in women.
Abbreviations
- PHC:
-
Public health center
- FFQ:
-
Food frequency questionnaire
- HR:
-
Hazard ratio
- CI:
-
Confidence interval
References
World Health Organization (2021) Dementia Key facts. https://www.who.int/news-room/fact-sheets/detail/dementia. Accessed 12 Oct 2021
Solfrizzi V, Custodero C, Lozupone M et al (2017) Relationships of dietary patterns, foods, and micro- and macronutrients with alzheimer’s disease and late-life cognitive disorders: a systematic review. J Alzheimers Dis 59:815–849. https://doi.org/10.3233/JAD-170248
Messina M, Nagata C, Wu A (2006) Estimated Asian adult soy protein and isoflavone intakes. Nutr Cancer 55:1–12. https://doi.org/10.1207/s15327914nc5501_1
Soni M, Rahardjo TB, Soekardi R et al (2014) Phytoestrogens and cognitive function: a review. Maturitas 77:209–220. https://doi.org/10.1016/j.maturitas.2013.12.010
Lu Y, An Y, Lv C et al (2018) Dietary soybean isoflavones in Alzheimer’s disease prevention. Asia Pac J Clin Nutr 27:946–954. https://doi.org/10.6133/apjcn.052018.01
Cui C, Birru RL, Snitz BE et al (2020) Effects of soy isoflavones on cognitive function: a systematic review and meta-analysis of randomized controlled trials. Nutr Rev 78:134–144. https://doi.org/10.1093/nutrit/nuz050
An R, Liu G, Khan N et al (2019) Dietary habits and cognitive impairment risk among oldest-old Chinese. J Gerontol B Psychol Sci Soc Sci 74:474–483. https://doi.org/10.1093/geronb/gbw170
Nakamoto M, Otsuka R, Nishita Y et al (2018) Soy food and isoflavone intake reduces the risk of cognitive impairment in elderly Japanese women. Eur J Clin Nutr 72:1458–1462. https://doi.org/10.1038/s41430-017-0061-2
Talaei M, Feng L, Yuan JM et al (2020) Dairy, soy, and calcium consumption and risk of cognitive impairment: the Singapore Chinese Health Study. Eur J Nutr 59:1541–1552. https://doi.org/10.1007/s00394-019-02010-8
Ozawa M, Ninomiya T, Ohara T et al (2013) Dietary patterns and risk of dementia in an elderly Japanese population: the Hisayama Study. Am J Clin Nutr 97:1076–1082. https://doi.org/10.3945/ajcn.112.045575
Xu X, Xiao S, Rahardjo TB et al (2015) Tofu intake is associated with poor cognitive performance among community-dwelling elderly in China. J Alzheimers Dis 43:669–675. https://doi.org/10.3233/JAD-141593
White L, Petrovitch H, Ross G et al (2000) Brain aging and midlife tofu consumption. J Am Coll Nutr 19:242–255. https://doi.org/10.1080/07315724.2000.10718923
Hogervorst E, Mursjid F, Priandini D et al (2011) Borobudur revisited: soy consumption may be associated with better recall in younger, but not in older, rural Indonesian elderly. Brain Res 1379:206–212. https://doi.org/10.1016/j.brainres.2010.10.083
Hogervorst E, Sadjimim T, Yesufu A et al (2008) High tofu intake is associated with worse memory in elderly Indonesian men and women. Dement Geriatr Cogn Disord 26:50–57. https://doi.org/10.1159/000141484
Nishimura K, Shiina R, Kashiwagi K et al (2006) Decrease in polyamines with aging and their ingestion from food and drink. J Biochem 139:81–90. https://doi.org/10.1093/jb/mvj003
Sumi H, Hamada H, Tsushima H et al (1987) A novel fibrinolytic enzyme (nattokinase) in the vegetable cheese Natto; a typical and popular soybean food in the Japanese diet. Experientia 43:1110–1111. https://doi.org/10.1007/BF01956052
Gupta VK, Scheunemann L, Eisenberg T et al (2013) Restoring polyamines protects from age-induced memory impairment in an autophagy-dependent manner. Nat Neurosci 16:1453–1460. https://doi.org/10.1038/nn.3512
Hsu R, Lee K, Wang J et al (2009) Amyloid-degrading ability of nattokinase from Bacillus subtilis Natto. J Agric Food Chem 57:503–508. https://doi.org/10.1021/jf803072r
Fadl NN, Ahmed HH, Booles HF et al (2013) Serrapeptase and nattokinase intervention for relieving Alzheimer’s disease pathophysiology in rat model. Hum Exp Toxicol 32:721–735. https://doi.org/10.1177/0960327112467040
Nozue M, Shimazu T, Sasazuki S et al (2017) Fermented soy product intake is inversely associated with the development of high blood pressure: the Japan Public Health center-based prospective study. J Nutr 147:1749–1756. https://doi.org/10.3945/jn.117.250282
Nozue M, Shimazu T, Charvat H et al (2021) Fermented soy products intake and risk of cardiovascular disease and total cancer incidence: the Japan Public Health Center-based Prospective study. Eur J Clin Nutr 75:954–968. https://doi.org/10.1038/s41430-020-00732-1
Tsugane S, Sobue T (2001) Baseline survey of JPHC study–design and participation rate. Japan Public Health Center-based Prospective Study on Cancer and Cardiovascular Diseases. J Epidemiol 11:S24–S29. https://doi.org/10.2188/jea.11.6sup_24
Tsugane S, Sawada N (2014) The JPHC study: design and some findings on the typical Japanese diet. Jpn J Clin Oncol 44:777–782. https://doi.org/10.1093/jjco/hyu096
Ministry of Health, Labour and Welfare Bureau of Japan (1993) Standardized assessment manual for the grade of activity daily living related to dementia among elderly with dementia. https://www.mhlw.go.jp/stf/shingi/2r9852000001hi4o-att/2r9852000001hi8n.pdf. Accessed 31 Jan 2022
Noda H, Yamagishi K, Ikeda A et al (2018) Identification of dementia using standard clinical assessments by primary care physicians in Japan. Geriatr Gerontol Int 18:738–744. https://doi.org/10.1111/ggi.13243
Hisano S (2009) The relationship between Revised Hasegawa Dementia Scale (HDS-R), Mini-Mental State Examination (MMSE) and Bed-fast Scale, Dementia Scale. Jpn J Geriatr Psychiatry 20:883–891
Tsubono Y, Takamori S, Kobayashi M et al (1996) A data-based approach for designing a semiquantitative food frequency questionnaire for a population-based prospective study in Japan. J Epidemiol 6:45–53. https://doi.org/10.2188/jea.6.45
Tsugane S, Sasaki S, Kobayashi M et al (2003) Validity and reproducibility of the self-administered food frequency questionnaire in the JPHC Study Cohort I: study design, conduct and participant profiles. J Epidemiol 13:S2–S12. https://doi.org/10.2188/jea.13.1sup_2
Sasaki S, Kobayashi M, Tsugane S (2003) Validity of a self-administered food frequency questionnaire used in the 5-year follow-up survey of the JPHC Study Cohort I: comparison with dietary records for food groups. J Epidemiol 13:S57–S63. https://doi.org/10.2188/jea.13.1sup_57
Ishihara J, Sobue T, Yamamoto S et al (2003) Validity and reproducibility of a self-administered food frequency questionnaire in the JPHC Study Cohort II: study design, participant profile and results in comparison with Cohort I. J Epidemiol 13:S134–S147. https://doi.org/10.2188/jea.13.1sup_134
Willet WC (2013) Nutritional epidemiology, 3rd edn. Oxford University Press, New York
Kimira M, Arai Y, Shimoi K et al (1998) Japanese intake of flavonoids and isoflavonoids from foods. J Epidemiol 8:168–175. https://doi.org/10.2188/jea.8.168
Arai Y, Watanabe S, Kimira M et al (2000) Dietary intakes of flavonols, flavones and isoflavones by Japanese women and the inverse correlation between quercetin intake and plasma LDL cholesterol concentration. J Nutr 130:2243–2250. https://doi.org/10.1093/jn/130.9.2243
Walker AE, Robins M, Weinfeld FD (1981) The National Survey of Stroke. Clinical findings. Stroke 12:I13–I44
de Wreede LC, Fiocco M, Putter H (2011) mstate: an R package for the analysis of competing risks and multi-state models. J Stat Softw 38:1–30
Tada A, Miura H (2017) Association between mastication and cognitive status: a systematic review. Arch Gerontol Geriatr 70:44–53. https://doi.org/10.1016/j.archger.2016.12.006
Golbitz P (1995) Traditional soyfoods: processing and products. J Nutr 125:570S-572S. https://doi.org/10.1093/jn/125.suppl_3.570S
Momose T, Nishikawa J, Watanabe T et al (1997) Effect of mastication on regional cerebral blood flow in humans examined by positron-emission tomography with 15O-labelled water and magnetic resonance imaging. Arch Oral Biol 42:57–61. https://doi.org/10.1016/s0003-9969(96)00081-7
Onozuka M, Fujita M, Watanabe K et al (2002) Mapping brain region activity during chewing: a functional magnetic resonance imaging study. J Dent Res 81:743–746. https://doi.org/10.1177/0810743
Moriya S, Tei K, Murata A et al (2011) Associations between self-assessed masticatory ability and higher brain function among the elderly. J Oral Rehabil 38:746–753. https://doi.org/10.1111/j.1365-2842.2011.02214.x
Strazzullo P, D’Elia L, Kandala NB et al (2009) Salt intake, stroke, and cardiovascular disease: meta-analysis of prospective studies. BMJ 339:b4567. https://doi.org/10.1136/bmj.b4567
Kuzma E, Lourida I, Moore SF et al (2018) Stroke and dementia risk: a systematic review and meta-analysis. Alzheimers Dement 14:1416–1426. https://doi.org/10.1016/j.jalz.2018.06.3061
van Erp-Baart MA, Brants HA, Kiely M et al (2003) Isoflavone intake in four different European countries: the VENUS approach. Br J Nutr 89:S25–S30. https://doi.org/10.1079/BJN2002793
de Kleijn MJ, van der Schouw YT, Wilson PW et al (2001) Intake of dietary phytoestrogens is low in postmenopausal women in the United States: the Framingham study (1–4). J Nutr 131:1826–1832. https://doi.org/10.1093/jn/131.6.1826
Nagata C, Shimizu H, Takami R et al (2003) Association of blood pressure with intake of soy products and other food groups in Japanese men and women. Prev Med 36:692–697. https://doi.org/10.1016/s0091-7435(03)00052-5
Sasaki S, Ishihara J, Tsugane S, JPHC (2003) Reproducibility of a self-administered food frequency questionnaire used in the 5-year follow-up survey of the JPHC Study Cohort I to assess food and nutrient intake. J Epidemiol 13:S115–S124. https://doi.org/10.2188/jea.13.1sup_115
Schurgers L, Vermeer C (2000) Determination of phylloquinone and menaquinones in food. Effect of food matrix on circulating vitamin K concentrations. Haemostasis 30:298–307. https://doi.org/10.1159/000054147
Acknowledgements
We would like to thank Dr. Sarah Krull Abe, Dr. Hikaru Ihira and Dr. Yoko Shimizu for their helpful comments on this analysis and also all the staff members for their contributions and efforts in conducting the survey. Members of the JPHC Study Group are listed at https://epi.ncc.go.jp/en/jphc/781/8510.html
Funding
This study was supported by the National Cancer Center Research and Development Fund (23-A-31 [toku], 26-A-2 and 29-A-4, since 2011), a Grant-in-Aid for Cancer Research from the Ministry of Health, Labour and Welfare in Japan (19shi-2, from 1989 to 2010), Grant-in-Aid for Scientific Research (B) (16H05246, from 2016 to 2019, and 21H03194, since 2021), and a grant from the commissioned project study, Ministry of Agriculture, Fishery and Forestry, Japan (JPJ005336).
Author information
Authors and Affiliations
Consortia
Contributions
ST designed the research and was the principal investigator; NS, MI and ST conducted the survey; NY and KY created the dementia dataset; UM analyzed the data and drafted the manuscript; HC supported the analyses and drafting of the manuscript; NS provided study oversight; and all authors interpreted the data, provided critical input, and read and approved the final manuscript. UM and NS had primary responsibility for the final content.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Murai, U., Sawada, N., Charvat, H. et al. Soy product intake and risk of incident disabling dementia: the JPHC Disabling Dementia Study. Eur J Nutr 61, 4045–4057 (2022). https://doi.org/10.1007/s00394-022-02937-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00394-022-02937-5