Association between gut microbiome and locomotive syndrome risk in healthy Japanese adults: A cross-sectional study

doi:10.21203/rs.3.rs-4866366/v1

Download PDF

Article

Association between gut microbiome and locomotive syndrome risk in healthy Japanese adults: A cross-sectional study

https://doi.org/10.21203/rs.3.rs-4866366/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Japan faces challenges as a super-aging society, with increasing life expectancy and a prolonged unhealthy period burdening the social security system. Musculoskeletal dysfunction significantly contributes to this unhealthy period. While exercise may influence the gut microbiome, its role in age-related musculoskeletal decline remains unclear. This cross-sectional study used data from the Kanagawa "ME-BYO" Prospective Cohort Study to investigate the association between gut microbiome composition and musculoskeletal function. Participants with a 5-question Geriatric Locomotive Function Scale (GLFS-5) and gut microbiome data were included. Classification tree analysis was performed using GLFS-5 (≥ 6 indicating locomotive syndrome) as the objective variable. Explanatory variables included gut microbiome composition, age, sex, BMI, menopause status, medical history, nutritional intake, alcohol consumption, smoking history, physical activity, and sitting time. Among 568 participants (36.8% male, median age 58.5 years), three terminal nodes were identified as GLFS-5 positive, with one node involving gut microbiome composition. Participants aged ≥ 70.0 and < 78.0 years who did not consume probiotic foods and had ≥ 0.04% relative abundance of the Holdemania genus were classified as at risk for locomotive syndrome. This study suggests a potential association between gut microbiome composition and locomotive syndrome risk in older adults, particularly those with higher Holdemania abundance. However, the cross-sectional design limits causal inference. Further longitudinal and interventional research is needed to clarify the relationship between gut microbiome and musculoskeletal function, and to explore potential preventive strategies targeting the gut microbiome to reduce locomotive syndrome risk and extend healthy life expectancy.

Health sciences/Medical research/Epidemiology

Health sciences/Health care/Geriatrics

microbiota

musculoskeletal diseases

aging

cross-sectional studies

risk factors

Japan was one of the first countries worldwide to face the challenges associated with a super-aging society; the population over 65 years of age (older adults) exceeded 21% of the total population in 2010. Japan’s total population is also declining, decreasing by 510,000 in 2021 compared to the previous year. In contrast, the older adult population in the same year was the largest ever at 36.4 million, corresponding to 29.1% of the total population (1). The proportion of older adults in Japan is projected to increase to 35.3% (38.6 million people) by 2042 (1). This super-aging society is a consequence of a demographic crisis caused by low fertility rates and prolonged life expectancy. Life expectancy in Japan was 81.4 for males and 87.5 for females, with an increase of 1.86 and 1.15 years in males and females, respectively, in nine years (1). In the same year, Japan’s healthy life expectancy was 72.7 years for males and 75.4 years for females, with a difference of 8.7 years for males and 12.1 years for females (1). Healthy life expectancy is defined as the average number of years a person can live in full health, without limitations in daily activities (1). In other words, during the 10 years between these two periods, a person requires nursing care and support, where ADL (Activity of Daily Life) and Quality of Life (QOL) decline, which is referred to as an unhealthy period. The unhealthy period also increases social security burdens, such as medical costs and long-term care benefits; therefore, it is a critical issue that must be solved not only as an individual health issue, but also as a societal issue. Therefore, extending healthy life expectancy is a global health issue, not only in Japan, which has a leading aging population.

In Japan, the government, in cooperation with local governments, has been taking measures to extend healthy life expectancy for at least three years; the life expectancy is expected to reach 75 years and above by 2040 (1). Due to the measures taken from the perspectives of exercise, diet, and smoking cessation, healthy life expectancy has increased between 2010 and 2019 by 2.3 and 1.8 years in males and females, respectively (1). However, improvement in the unhealthy period was only 0.4 and 0.6 in males and females, respectively (1); reducing unhealthy periods might be difficult simply by using measures to extend healthy life expectancy. Thus, preventing conditions requiring care and support, which are significant causes of unhealthy periods, is indispensable to resolving the discrepancy and reducing unhealthy periods.

Japan has a long-term care insurance system that supports people who require long-term care by providing services, such as welfare services, according to the level of required long-term care based on each individual’s mental and physical conditions (1). Individuals certified as requiring support or long-term care under this system can be considered unhealthy. In 2022, the number of individuals certified as requiring support or long-term care under this system reached 7.0 million (about 19.0% of people aged > 65 years) (1). We believe the number of older adults in an unhealthy period should be larger because certification is based on voluntary applications. The major causes of support or long-term care requirement were dementia (18.7%), cerebrovascular disease (CVD, 15.1%), and frailty due to old age (frailty, 13.8%), followed by musculoskeletal diseases, such as joint disease (10.2%) and falls and fractures (12.5%) (1). Musculoskeletal diseases account for approximately 25% of all musculoskeletal diseases. Notably, this proportion increased to 54.6% when we also included diseases such as CVD and frailty, which often impair musculoskeletal function and limit ADL. Therefore, preventing musculoskeletal dysfunction is vital for preventing people from becoming supported and requiring long-term care (long-term care prevention), which will contribute to reducing the unhealthy period by maintaining and even improving ADL and QOL.

Exercise and dietary interventions are recommended for long-term care prevention (2). Exercise directly affects physical function and improves balance and muscle strength. An indirect effect that improves gut microbiome diversity is also known (3). The gut microbiome is endemic to the digestive tract, and previous studies have shown a close relationship with human health, not only gastrointestinal diseases (4, 5). For example, interventions that target modification of the gut microbiome environment are reportedly effective in diabetes mellitus, liver disease, depression, and irritable bowel syndrome (4–8). One study suggested that physical activity might benefit gut microbiome composition, which contributes to healthy aging (9). However, the biological mechanisms by which exercise induces these changes have not yet been clarified. Furthermore, the involvement of the gut microbiome in age-related decline in musculoskeletal function has not yet been elucidated. Thus, we focused on studying the association between the gut microbiome and musculoskeletal function, considering the effect of age. We believe that by clarifying this association, the gut microbiome may serve as a tool for the prevention of long-term care. For example, if the gut microbiome affects musculoskeletal function, changes in the gut microbiome that deteriorate nutrient absorption might be the cause. In this case, supplementation with food or products that would improve the gut microbiome environment and conventional nutritional interventions might be effective. Conversely, if musculoskeletal dysfunction affects the gut microbiome, an altered gut microbiome environment may trigger sequelae that further impair musculoskeletal function. Nevertheless, clarifying the association between the gut microbiome and musculoskeletal function may provide novel measures for long-term care prevention that contribute to reducing unhealthy periods.

We aimed to clarify the association between gut microbiome and musculoskeletal function using cross-sectional data from a cohort of healthy individuals in Kanagawa Prefecture, Japan. To achieve this, we employed decision tree analysis, a method particularly suited for exploring complex, non-linear relationships in biological data. This approach allows for the identification of key variables and their interactions without assuming a predetermined model structure, making it ideal for investigating the potentially intricate associations between gut microbiome composition and locomotive syndrome risk.

Study design and participants

This was a cross-sectional analysis that used the Kanagawa “ME-BYO” Prospective Cohort Study (the ME-BYO cohort) data. The ME-BYO cohort was a collaborative institution of the Japan Multi-Institutional Collaborative Cohort (J-MICC) Study. Details of the J-MICC Study and its association with the ME-BYO cohort have been described elsewhere (10, 11). Briefly, the J-MICC Study was conducted by 13 research institutions located in 12 prefectures of Japan under a standardized protocol. The baseline survey of the ME-BYO cohort started in 2016 and is still ongoing in 2024. The eligibility criteria for the ME-BYO cohort were living or working in Kanagawa Prefecture and aged 18–95 years. Among the 3,409 participants at the end of March 2022, we used data from 577 participants for whom gut microbiome data were available. Written informed consent was obtained from all the participants. All research procedures were approved by the Ethics Committee of Kanagawa Cancer Center (Approval No. 28KEN-36).

Measurements

An examiner performed anthropometric measurements, and body mass index (BMI) was calculated from height and weight. Lifestyle and health condition data were collected using a self-administered questionnaire. We used the age at the time of recruitment. Daily nutritional intake was evaluated using a validated food frequency questionnaire and adjusted by total energy intake using the residual method (12, 13). We selected the following nutrients for analysis that are known to be beneficial to the human gut microbiome: omega-3-unsaturated fatty acids (n3PUFA), omega-6-unsaturated fatty acids (n6PUFA), saturated fatty acids (SFA), soluble dietary fiber (SDF), and insoluble dietary fiber (IDF) (14–16); which facilitate the metabolism of beneficial bacteria and exert proliferative effect (14–16). Nutrients associated with the musculoskeletal system include total energy intake, protein, folate, vitamins (A, B1, B2, E, and D), and calcium (17–24). Alcohol intake (g/day) was also assessed. Smoking status was assessed as currently smoking, coded as 1, and other answers, such as quit or never smoked, as 0. Daily physical activity was evaluated based on the International Physical Activity Questionnaire in metabolic equivalents hours per day (METs-h/day), and activities higher than 3.0 METs as moderate to vigorous physical activity (MVPA) (25). Participants were considered exposed to antibiotics and probiotics (medication) if they took those medications within two weeks and were considered exposed to natto, which contains Bacillus subtilis natto, and probiotics (food), such as yogurt containing Lactobacillus, if they consumed those foods more than three times in the last week. Females who were menopausal and all males were treated as participants without menstruation. Medical histories of stroke, coronary artery disease, diabetes mellitus, and malignant tumors were used.

Assessment of musculoskeletal function

The locomotive syndrome was used to assess musculoskeletal function. The locomotive syndrome is a concept introduced by the Japanese Orthopedic Association (JOA) to prevent a reduction in mobility due to musculoskeletal dysfunction (1, 26). The JOA aims to increase awareness and motivate the public regarding long-term care prevention by disseminating this concept. To evaluate locomotive syndrome, we used the 5-question geriatric locomotive function Scale (GLFS-5), a short-form version of the 25-item original questionnaire (27, 28). The GLFS-5 is a 5-item questionnaire that uses a 5-point Likert scale, with a final score ranging from 0 to 20 points, where a higher score indicates a higher risk and a cut-off value of 6 points indicates the existence of locomotive syndrome. The GLFS-5 was also collected from the baseline questionnaire of the ME-BYO cohort.

Gut microbiome data

At enrollment, we distributed the fecal specimen collection kit, a brush-type kit containing guanidine thiocyanate solution (Techno Suruga Laboratory, Shizuoka, Japan), and the participants were instructed to return the kit via post mail to the laboratory. The methods used to obtain microbiome data have been described elsewhere (29–31). Short-amplicon 16S rRNA gene sequencing targeting the V1-V2 region was performed using the 27F forward and 338R reverse primers and clustered by gene-specific taxonomic classification based on the QIIME2 pipeline (version 2020.8) (31). After clustering, the sequences were classified for a taxonomic assignment using the robust taxonomy simplifier for SILVA (arts-SILVA), which was originally developed from the 16S rRNA taxonomy dataset based on SILVA 138 (32). Data at the phylum to genus taxonomy level were used, and the relative abundances of 469 operational taxonomic units (OTUs) were available.

Statistical analysis

Linear discriminant analysis (LDA) effect size (LEfSe) was used to detect phyla that are associated (p < 0.001 and LDA score > 0.2) with the existence of locomotive syndrome using lefser function in lefser package. This initial screening step allowed us to identify relevant phyla for further investigation. Subsequently, we focused on the OTUs from class to genus level that belonged to these identified phyla. These selected OTUs were then included as potential predictors in our classification tree analysis. This method is used in clinical evaluations, such as disease prediction and screening (33–35). The advantage of classification tree analysis is that no distribution assumption is required, and nonlinear association can be handled, as it is a nonparametric procedure (36, 37). The analysis was not affected by outliers, collinearities, heteroscedasticity, or distributional error structures that affect parametric procedures (36). Furthermore, the classification structure obtained from the analysis was explicit and easy to interpret. Since the microbiome data we used were relative abundances, excess zero values were included; thus, we adopted classification tree analysis. Classification tree analysis was performed using the ‘rpart’ function of the rpart package (38). The predictive variable was the GLFS-5, and age, sex, BMI, lifestyle factors described in the measurement section above, and the gut microbiome were used as inputs. We set the minimum terminal node size to 20 observations and performed cost-complexity pruning (CP). We obtained the CP parameter by five-fold cross-validation with ten times repeat using the train function in the caret package and set it to 0.02 (39). All analyses were performed using the R software (version 4.2.1; R Foundation for Statistical Computing, Vienna, Austria) (40).

Nutritional data were unavailable for three participants, GLFS-5 was unavailable for six participants, and 568 participants (209 males [36.8%]) were finally included in the analysis. The participants’ characteristics are listed in Table 1. The median (inter-quartile range:IQR) age was 58.5 (48.0–71.0) years, and the median (IQR) of GLFS-5 was 0.0 (0.0–3.0). Among the 99/568 (17.4%) GLFS-5 positive group (31/99 males [31.3%]), the median (IQR) age was 74.0 (71.0–78.0), and BMI was 22.6 kg/m² (20.9–24.6).

Table 1

Characteristics of the study participants
	Whole subject (n = 568)	GLFS-5 < 6 (n = 469)	GLFS-5 ≥ 6 (n = 99)
Age (years)	58.5 (48.0–71.0)	56.0 (46.0–67.0)	74.0 (71.0–78.0)
Sex (male, n [%])	209 (36.8)	178 (38.0)	31 (31.3)
BMI (kg/m²)	22.4 (20.4–24.2)	22.3 (20.3–24.2)	22.6 (20.9–24.6)
GLFS-5	0.0 (0.0–3.0)	0.0 (0.0–1.0)	8.0 (6.0–10.0)
Daily nutritional intakes
Total energy (kcal)	1,511.4 (1,302.7–1,727.0)	1,526.0 (1,317.6–1,731.4)	1,474.7 (1,215.5–1,663.1)
Protein (g)	50.7 (46.2–55.8)	49.8 (45.8–54.6)	54.5 (50.7–60.2)
Soluble dietary fiber (g)	2.0 (1.7–2.4)	1.9 (1.6–2.3)	2.2 (2–2.6)
Insoluble dietary fiber (g)	7.7 (6.6–9.3)	7.6 (6.4–9.1)	8.7 (7.4–10.4)
n-6 polyunsaturated fatty acids (g)	10.7 (9.0–12.5)	10.6 (8.9–12.3)	10.8 (9.4–13.0)
n-3 polyunsaturated fatty acids (g)	2.5 (1.8–2.1)	2.1 (1.8–2.4)	2.2 (1.9–2.8)
Folate (µg)	462.9 (357.0–620.3)	443.9 (344.6–606.2)	526.1 (421.7–672.5)
Vitamin B1 (mg)	0.9 (0.8–1.0)	0.9 (0.8–1.0)	0.9 (0.8–1.1)
Vitamin B2 (mg)	1.4 (1.2–1.7)	1.4 (1.2–1.7)	1.6 (1.3–2.0)
Vitamin E (mg)	29.3 (21.5–39.8)	29 (21.7–39.3)	32.5 (21.0–41.9)
Vitamin D (mg)	11.5 (8.2–17.0)	10.7 (7.8–15.5)	16.8 (11.9–22.5)
Vitamin A (mg)	1.9 (1.2–2.7)	1.9 (1.1–2.6)	2.4 (1.4–2,941.1)
Calcium (mg)	509.5 (426.8–636.7)	494.4 (403.5–604.7)	596.1 (489.9–709.9)
Alcohol intake (g/day)	0.0 (0.0–8.2)	0.0 (0.0–9.4)	0.0 (0.0–1.8)
< 20 (n [%])	492 (86.6)	400 (85.3)	92 (92.9)
≥ 20, < 40 (n [%])	47 (8.3)	42 (9)	5 (5.1)
≥ 40, < 60 (n [%])	17 (3)	15 (3.2)	2 (2)
≥ 60 (n [%])	12 (2.1)	12 (2.6)	0 (0)
MVPA (METs-h/day)	8.8 (4.3–15.7)	9.0 (4.4–16.4)	7.8 (3.8–12.6)
Sitting time (hours/day)	4.0 (2.0–6.0)	4.0 (2.0–6.0)	4.0 (2.0–6.0)
Antibiotics (n [%])	3 (0.5)	2 (0.4)	1 (1.0)
Probiotics (medication, n [%])	15 (2.6)	6 (1.3)	9 (9.1)
Natto (n [%])	113 (19.9)	69 (14.7)	44 (44.4)
Probiotics (food, n [%])	99 (17.4)	53 (11.3)	46 (46.5)
Smoking (n [%])	36 (6.3)	34 (7.2)	9 (9.1)
Menopaused (n [%])	232 (40.8)	168 (35.8)	64 (64.6)
Past medical history
Cancer (n [%])	48 (8.5)	30 (6.4)	18 (18.2)
Cerebrovascular disease (n [%])	18 (3.2)	5 (1.1)	13 (13.1)
Coronary artery disease (n [%])	20 (3.5)	13 (2.8)	7 (7.1)
Dyslipidemia (n [%])	95 (16.7)	64 (13.6)	31 (31.3)
Hypertension (n [%])	122 (21.5)	82 (17.5)	40 (40.4)
Data are shown as median (inter-quartile range) unless otherwise specified.
BMI, body mass index; MVPA, moderate-to-vigorous physical activity; MET-h, metabolic equivalents

LEfSe showed that Actinobacteria (LDA score = 7.1) and Firmicutes (LDA score = -7.3) phyla were associated with the existence of locomotive syndrome. The results of the classification tree analysis are shown in Fig. 1. Of the five terminal nodes, three were classified as GLFS-5 positive. The branching conditions were age, probiotic (food) intake, and the Holdemania genus. In the terminal node with the following condition, 39/50 (78.0%) of the classified participants were positive in GLFS-5; nodes with age ≥ 70.0 years and took probiotics (food). The median (IQR) GLFS-5 was 8.0 (6.5, 9.0) points. Characteristics of the groups classified as GLFS-5 positive and negative are presented in Table 2. The total number of participants classified to the GLFS-5 positive node was 97/568 (17.1%), and 68/97 (70.1%) participants were positive in GLFS-5. Accordingly, sensitivity was 0.69, specificity was 0.94, positive likelihood ratio was 11.1, and negative likelihood ratio was 0.3.

Table 2

Characteristics of participants according to the classified node in the classification analysis
	Negative terminal nodes (n = 471)	Positive terminal nodes (n = 97)
Age (years)	56.0 (46.0–67.0)	76.0 (72.0–79.0)
Sex (male, n [%])	175 (37.2)	34 (35.1)
BMI (kg/m²)	22.2 (20.3–24.2)	22.7 (21.1–24.3)
GLFS-5	0.0 (0.0–1.0)	7.0 (3.0–9.0)
≥ 6 (n [%])	31 (6.6)	68 (70.1)
Daily nutritional intakes
Total energy (kcal)	1,517.4 (1,316.3–1,726.5)	1,498.0 (1,237.0–1,727.0)
Protein (g)	49.8 (45.9–54.6)	54.5 (50.8–60.8)
Soluble dietary fiber (g)	1.9 (1.6–2.3)	2.3 (2.0–2.7)
Insoluble dietary fiber (g)	7.5 (6.4–9.0)	8.9 (7.8–10.9)
n-6 polyunsaturated fatty acids (g)	10.6 (8.9–12.3)	11.1 (9.6–13.0)
n-3 polyunsaturated fatty acids (g)	2.1 (1.8–2.4)	2.3 (2.0–2.8)
Folate (µg)	442.4 (346.4–595.0)	562.0 (431.4–776.9)
Vitamin B1 (mg)	0.9 (0.8–1.0)	0.9 (0.8–1.1)
Vitamin B2 (mg)	1.4 (1.2–1.7)	1.6 (1.4–2)
Vitamin E (mg)	28.8 (21.7–39.3)	32.5 (21.3–42.1)
Vitamin D (mg)	10.7 (7.8–15.5)	16.7 (12.5–22.5)
Vitamin A (mg)	1.9 (1.1–2.6)	2.2 (1.3–3.0)
Calcium (mg)	495 (403.1–608.5)	587.7 (480.5–700.5)
Alcohol intake (g/day)	0.0 (0.0–9.2)	0.0 (0.0–3.7)
< 20 (n [%])	401 (37.2)	91 (93.8)
≥ 20, < 40 (n [%])	43 (38.2)	4 (4.1)
≥ 40, < 60 (n [%])	15 (39.2)	2 (2.1)
≥ 60 (n [%])	12 (40.2)	0 (0)
MVPA (METs-h/day)	8.7 (4–16.4)	9.4 (6.0–13.3)
Sitting time(hours/day)	4.0 (2.0–6.0)	4.0 (2.0–6.0)
Antibiotics (n [%])	3 (0.6)	0 (-)
Probiotics (medication, n [%])	6 (1.3)	9 (9.3)
Natto (n [%])	70 (14.9)	43 (44.3)
Probiotics (food, n [%])	49 (10.4)	50 (51.5)
Smoking (n [%])	33 (7.0)	3 (3.1)
Menopaused (n [%])	171 (36.3)	61 (62.9)
Past medical history
Cancer (n [%])	30 (6.4)	18 (18.6)
Cerebrovascular disease (n [%])	8 (1.7)	10 (10.3)
Coronary artery disease (n [%])	10 (2.1)	10 (10.3)
Dyslipidemia (n [%])	65 (13.8)	30 (30.9)
Hypertension (n [%])	87 (18.5)	35 (36.1)
Data are shown in median (inter-quartile range) unless otherwise specified.
BMI, body mass index; MVPA, moderate-to-vigorous physical activity; MET-h, metabolic equivalents

The terminal node associated with the microbiome was under the condition of age ≥ 70.0 and < 78 years, did not take probiotics (food), and whose relative abundance of Holdemania genus ≥ 0.04%, shown as node 1 in Table 3. MVPA was highest in this node 1 (median [IQR] 10.4 [8.0–18.1]) compared with other groups, including the negative terminal nodes (median [IQR] 8.7 [4.0–16.4]). In the terminal node, under the condition of taking probiotics (food), 37/50 (74.0%) participants were also taking natto, and this proportion was higher than that in the other groups.

Table 3

Characteristics of participants according to the classified node in the classification analysis
		Positive terminal nodes
	Negative terminal nodes (n = 471)	Node 1 (n = 24)	Node 2 (n = 23)	Node 3 (n = 50)
Age (years)	56.0 (46.0–67.0)	73.0 (71.0–74.0)	80.0 (79.0–82.0)	76.0 (72.0–78.0)
Sex (male, n [%])	175 (37.2)	9 (37.5)	14 (60.9)	11 (22.0)
BMI (kg/m²)	22.2 (20.3–24.2)	23.0 (22.4–24.1)	22.8 (21.8–24.2)	22.3 (20.6–24.3)
GLFS-5	0.0 (0.0–1.0)	6 (1.0–9.3)	8 (3.5–9.5)	7.3 (6.0–9.0)
≥ 6 (n [%])	31 (6.6)	13 (54.2)	16 (69.6)	39 (78.0)
Daily nutritional intakes
Total energy (kcal)	1,517.4 (1,316.3–1,726.5)	1,395.0 (1,219.0–1,565.0)	1,609.0 (1,362.0–1,865.0)	1,500.5 (1,211.4–1,635.0)
Protein (g)	49.8 (45.9–54.6)	54.2 (51.3–57.1)	51.6 (46.7–55.5)	57.4 (52.8–65.3)
Soluble dietary fiber (g)	1.9 (1.6–2.3)	2.0 (1.9–2.4)	2.2 (2.0–2.5)	2.5 (2.1–2.9)
Insoluble dietary fiber (g)	7.5 (6.4–9.0)	7.9 (7.2–9.2)	8.3 (7.6–10.3)	9.8 (8.1–11.2)
n-6 polyunsaturated fatty acids (g)	10.6 (8.9–12.3)	10.2 (9.7–12.7)	10.5 (8.0–12.3)	11.7(10.1–13.4)
n-3 polyunsaturated fatty acids (g)	2.1 (1.8–2.4)	2.3 (2.0–2.6)	2.1(1.8–2.7)	2.4 (2.0–2.9)
Folate (µg)	442.4 (346.4–595.0)	455.3 (404.0–572.0)	541.7 (436.5–671.0)	637.6 (508.9–809.1)
Vitamin B1 (mg)	0.9 (0.8–1.0)	0.9 (0.8–1.0)	0.8 (0.7–0.9)	1.0 (0.8–1.1)
Vitamin B2 (mg)	1.4 (1.2–1.7)	1.5 (1.1–1.8)	1.6 (1.3–1.8)	1.9 (1.4–2.2)
Vitamin E (mg)	28.8 (21.7–39.3)	25.7 (20.6–33.3)	24.7(16.7–36.4)	35.0 (27.8–45.3)
Vitamin D (mg)	10.7 (7.8–15.5)	16.7 (12.0–20.6)	14.3 (9.2–19.7)	18.7 (13.5–23.9)
Vitamin A (mg)	1.9 (1.1–2.6)	1.8 (1.0–2.6)	2.0 (1.4–2.6)	2.7 (1.8–3.3)
Calcium (mg)	495 (403.1–608.5)	534.4 (461.2–615.0)	511.6 (459.5–684.5)	633.5 (544.5–758.8)
Alcohol intake (g/day)	0.0 (0.0–9.2)	0 (0.0–3.9)	0.0 (0.0–9.7)	0.0 (0.0–0.9)
< 20 (n [%])	401 (85.1)	20 (83.3)	21 (91.3)	50 (100.0)
≥ 20, < 40 (n [%])	43 (9.3)	3 (12.5)	1 (4.3)	0 (-)
≥ 40, < 60 (n [%])	15 (3.2)	1 (4.2)	1 (4.3)	0 (-)
≥ 60 (n [%])	12 (2.6)	0 (-)	0 (-)	0 (-)
MVPA (METs-h/day)	8.7 (4.0–16.4)	10.4 (8.0–18.1)	8.2 (3.8–14.8)	9 (6.0–13.0)
Sitting time(hours/day)	4.0 (2.0–6.0)	4 (3.5–6.0)	4 (3.0–6.0)	4 (2.0–6.0)
Antibiotics (n [%])	3 (0.6)	0 (-)	0 (-)	0 (-)
Probiotics (medication, n [%])	6 (1.3)	1 (4.2)	0 (-)	8 (16.0)
Natto (n [%])	70 (14.9)	4 (16.7)	2 (8.7)	37 (74.0)
Probiotics (food, n [%])	49 (10.4)	0 (-)	0 (-)	50 (100.0)
Smoking (n [%])	33 (7.0)	2 (8.3)	1 (4.3)	0 (-)
Menopaused (n [%])	171 (36.3)	14 (58.3)	8 (34.8)	39 (78.0)
Medical history
Cancer (n [%])	30 (6.37)	5 (20.8)	5 (21.7)	8 (16.0)
Cerebrovascular disease (n [%])	8 (1.7)	4 (16.7)	3 (13.0)	3 (6.0)
Coronary artery disease (n [%])	10 (2.1)	3 (12.5)	3 (13.0)	4 (8.0)
Diabetes (n [%])	20 (4.3)	1 (4.2)	3 (13.0)	4 (8.0)
Dyslipidemia (n [%])	65 (13.8)	4 (16.7)	10 (43.5)	16 (32.0)
Hypertension (n [%])	87 (18.5)	9 (37.5)	10 (43.5)	16 (32.0)
Data are shown as median (inter-quartile range) unless otherwise specified.
BMI, body mass index; MVPA, moderate-to-vigorous physical activity; MET-h, metabolic equivalents

This study examined the association between locomotive syndrome (an indicator of musculoskeletal dysfunction) and the gut microbiome in healthy Japanese adults. The relative abundance of Holdemania among the 449 genera was associated with positive screening results for locomotive syndrome under certain conditions.

The results of the classification tree analysis showed that participants possessing a relative abundance of Holdemania 0.04% or greater under branches of age (69.5–77.5 years) and not taking probiotics (food) were associated with the risk of locomotive syndrome (Fig. 1). A negative association has been reported between the relative abundance of Holdemania and an increase in skeletal muscle mass and prolongation of one-leg standing time with eyes closed following exercise intervention (41). The highest MVPA observed in the terminal node in Holdemania does not contradict this evidence. Decreased skeletal muscle mass is important in locomotive syndrome (1). Together, these findings suggest that participants with a relative abundance of the Holdemania ≥ 0.04% may have lower skeletal muscle mass and increased locomotive syndrome risk.

A previous study has suggested an association between frailty, which is related to a decline in musculoskeletal function in the older adults, and the gut microbiome (42). This study reported a negative association between frailty and the relative abundances of Corpobacillus and Veillonella. Holdemania and Corpobacillus belong to the common order Erysipelotrichales, with Veillonela in the common class Negativicutes. The concepts of frailty and locomotive syndrome are similar in that frailty is a state in which age-related decline in motor and cognitive function impairs the vitality of the body and mind and is also affected by comorbid illness, but its decline can be reversed with appropriate intervention and care (43). As Holdemania is classified in the same bacterial lineage as previously reported frailty-related microbiomes, our results suggest the involvement of the gut microbiome in locomotive syndrome and musculoskeletal function.

The conditions for the branch in the classification tree other than the microbiome were age and probiotics (food). The association with age is consistent with existing evidence that age is a risk factor for locomotive syndrome (1). A survey of the prevalence of locomotive syndrome in Japan revealed that the prevalence of locomotive syndrome tended to increase sharply from the age of 70 years (44). Specifically, the prevalence of locomotive syndrome was 8.4% for those aged in their 40s, 9.2% for those in their 50s, 8.3% for those in their 60s, and 16.3% for those in their 70s. The first branching condition in the results of the classification tree analysis is age 70 or older, and our results are consistent with existing evidence. However, the positive association between probiotic (food) consumption frequency and the risk of locomotive syndrome contradicts existing evidence; in fact, consuming probiotic food is recommended to prevent osteoporosis (45). We considered that this was due to reverse causation, reflecting that participants who were aware of their reduced mobility tended to be more careful in consuming probiotic foods to maintain or improve their musculoskeletal function or prevent the development of locomotive syndrome. The same reason could explain the high number of participants in this group taking natto.

The strength of this study is that it was conducted to develop a new prevention method for locomotive syndrome. Classification tree analysis allowed us to identify the population subgroup that would be classified as at risk for locomotive syndrome based on its interpretable results. The conditions of the subgroups identified in this study can be used to stratify the population for preventive intervention targets. As mentioned earlier, age is a risk factor for locomotive syndrome, and current preventive interventions for the older adult population must be continued. Furthermore, in the present study, we found that, in a population aged 70–78 years, a high relative abundance of Holdemania was associated with a high percentage of locomotive syndrome positivity. This suggests the potential of gut microbiota-based interventions in this population. Practically, interventions aimed at preventing locomotive syndrome in the entire population aged > 70 years are not feasible. Although the causal relationship between gut microbiota and musculoskeletal dysfunction has not been elucidated in this study, the use of a test kit to determine the prevalence of Holdemania among 70–78 year old could provide a preventive intervention for those at high risk of developing locomotive syndrome, as well as promote self-education in this population. Success in these preventive interventions is expected to reduce social security costs and improve QOL by extending healthy life expectancy and reducing unhealthy periods (46).

While our findings suggest a potential role for gut microbiome analysis in assessing locomotive syndrome risk, it is important to consider the practical implications of implementing such measurements in clinical practice. The use of gut microbiome data for risk assessment would require consideration of several factors, including the cost of microbiome sequencing and the time required for sample processing and data analysis. These factors must be weighed against the potential benefits of identifying a subgroup of individuals at risk for locomotive syndrome. Additionally, the cost-effectiveness of this approach compared to current screening methods needs to be evaluated. Future research should focus on improving the efficiency of the process of gut microbiome analysis, reducing associated costs, and conducting comprehensive cost-benefit analyses to determine the feasibility and value of incorporating gut microbiome measurements into routine health screenings for locomotive syndrome risk. Such studies will be crucial in translating our findings into practical, cost-effective interventions for preventing and managing locomotive syndrome.

This study had several limitations. First, we could not directly analyze the association between the gut microbiome and skeletal muscle mass due to a lack of data. Second, because this was a cross-sectional study, causal relationships could not be inferred. Further evaluations using longitudinal or interventional studies are required. Third, the assessment of locomotive syndrome was undertaken using the GLFS-5, a self-administered questionnaire, which may not reflect the actual degree of musculoskeletal function due to bias caused by its self-reported nature, and people with early musculoskeletal dysfunction may be unaware of their potential risk, which may lead to misclassification. Future studies should include objective physical performance measures such as gait speed and balance tests. Fourth, the association between Holdemania and the locomotive syndrome was observed under other branching conditions; thus, this study could not assess independent effects. Other limitations include selection bias caused by using data from a genomic cohort study of healthy individuals. I n this study, the prevalence of locomotive syndrome was 17.4%, much lower than the 34.1% prevalence of locomotive syndrome among 60 years and older in Japan, and individuals who already had reduced musculoskeletal function at a certain level were not included. However, as the study aimed to elucidate evidence for primary or secondary prevention, the analyzed population was similar to the population in which we aimed to generalize the results (47).

In conclusion, classification tree analysis revealed an association between Holdemania and locomotive syndrome, which reflects musculoskeletal function. Prospective studies that can clarify causal relationships should be conducted for the future use of gut microbiome information in preventive interventions.

Author contributions

MN, and SN, conceptualized, investigated, and was a major contributor in writing the manuscript. MN performed formal analysis. SN contributed in data curation, validated and visualized the analysis, and contributed in project administration. SN, and HN contributed in provision of data, and funding acquisition. TM, and HN reviewed and edited the manuscript. All authors read and approved the final manuscript.

Acknowledgments

We would like to thank Editage (www.editage.jp) for English language editing.

Competing interests

All authors declare no financial or non-financial competing interests.

Funding details

This study was funded by Grants-in-Aid for Scientific Research for Priority Areas of Cancer (No. 17015018) and Innovative Areas (No. 221S0001) and by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant (No. 16H06277, 22H04923 [CoBiA] and 17K17614) from the Japanese Ministry of Education, Culture, Sports, Science and Technology. The funder played no role in study design, data collection, analysis and interpretation of data, or the writing of this manuscript.

Data availability statement

The datasets analysed during the current study are not publicly available due to ethical restrictions but are available upon application and approval from the Kanagawa Prostective “ME-BYO” Cohort Study office (contact via the corresponding author) for researchers who meet the criteria for data sharing.

Code availability statement

Unreported custom computer code or scripts were not used in this study.

Cabinet Office, Ministry of Health, Labour and Welfare of Japan. Annual Reports on the Ageing Society, Health and Welfare, and Long-Term Care Insurance System. https://www8.cao.go.jp/kourei/whitepaper/index-w.html; https://www.mhlw.go.jp/english/wp/index.html (2014–2022).
Ishibashi, H. Locomotive syndrome in Japan. Osteoporos Sarcopenia. 4, 86–94 (2018).
Ortiz-Alvarez, L., Xu, H., & Martinez-Tellez, B. Influence of Exercise on the Human Gut Microbiota of Healthy Adults: A Systematic Review. Clin Transl Gastroenterol. 11, e00126; 10.14309/ctg.0000000000000126 (2020).
Wen, L., et al. Innate immunity and intestinal microbiota in the development of Type 1 diabetes. Nature.455, 1109–1113 (2008).
Sampson, T.R., et al. Gut Microbiota Regulate Motor Deficits and Neuroinflammation in a Model of Parkinson's Disease. Cell.167, 1469–1480.e12 (2016).
Kelishadi, R., Farajian, S., & Mirlohi, M. Probiotics as a novel treatment for non-alcoholic Fatty liver disease; a systematic review on the current evidences. Hepat Mon.13, e7233 (2013).
Huang, R., Wang, K., Hu, J. Effect of Probiotics on Depression: A Systematic Review and Meta-Analysis of Randomized Controlled Trials. Nutrients. 8, 483; 10.3390/nu8080483. (2016).
Dale, H. F., Rasmussen, S. H., Asiller, Ö.Ö., Lied, G.A. Probiotics in Irritable Bowel Syndrome: An Up-to-Date Systematic Review. Nutrients. 11, 2048; 10.3390/nu11092048. (2019).
Campaniello, D., et al. How Diet and Physical Activity Modulate Gut Microbiota: Evidence, and Perspectives. Nutrients.14, 2456; 10.3390/nu14122456 (2022).
Nakamura, S., et al. The ME-BYO index: A development and validation project of a novel comprehensive health index. Front Public Health.11, 1142281; 10.3389/fpubh.2023.1142281 (2023).
Takeuchi, K., et al. Study profile of the Japan Multi-institutional Collaborative Cohort (J-MICC) Study. J Epidemiol. 31, 660–668 (2021).
Cade, J. E., Burley, V. J., Warm, D. L., & Thompson, R. L., Margetts, B. M. Food-frequency questionnaires: a review of their design, validation and utilisation. Nutr Res Rev. 17, 5–22 (2004).
Nanri, A., et al. Development, Relative Validity, and Reproducibility of a Short Food Frequency Questionnaire for the Japanese. Nutrients. 14, 4394; 10.3390/nu14204394 (2022).
Costantini, L., Molinari, R., Farinon, B., & Merendino, N. Impact of Omega-3 Fatty Acids on the Gut Microbiota. Int J Mol Sci. 18, 2645 (2017).
Ghosh, S., et al. Fish oil attenuates omega-6 polyunsaturated fatty acid-induced dysbiosis and infectious colitis but impairs LPS dephosphorylation activity causing sepsis. PLoS One. 8, e55468 (2013).
Simpson, H. L., & Campbell, B. J. Review article: dietary fibre-microbiota interactions. Aliment Pharmacol Ther. 42, 158–179 (2015).
Cho, Y. J., et al. Sex- and age-specific effects of energy intake and physical activity on sarcopenia. Sci Rep.10, 9822; 10.1038/s41598-020-66249-6 (2020).
Volkert, D., et al. ESPEN guideline on clinical nutrition and hydration in geriatrics. Clin Nutr. 38, 10–47 (2019).
Schoufour, J. D., et al. The association between dietary protein intake, energy intake and physical frailty: results from the Rotterdam Study. Br J Nutr. 121, 393–401 (2019).
Tagawa, R., et al. Dose-response relationship between protein intake and muscle mass increase: a systematic review and meta-analysis of randomized controlled trials. Nutr Rev. 79, 66–75 (2020).
Tang, B. M., et al. Use of calcium or calcium in combination with vitamin D supplementation to prevent fractures and bone loss in people aged 50 years and older: a meta-analysis. Lancet. 370, 657–666 (2007).
Fratoni, V., & Brandi, M. L. B vitamins, homocysteine and bone health. Nutrients. 7, 2176–2192; 10.3390/nu7042176 (2015).
Yee, M. M. F, Chin, K. Y., Ima-Nirwana, S., & Wong, S. K. Vitamin A and Bone Health: A Review on Current Evidence. Molecules. 26, 1757 (2021).
Gana, W., et al. Analysis of the Impact of Selected Vitamins Deficiencies on the Risk of Disability in Older People. Nutrients. 13, 3163; 10.3390/nu13093163 (2021).
Craig, C. L., et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc. 35, 1381–1395 (2003).
Nakamura, K., & Ogata T. Locomotive Syndrome: Definition and Management. Clin Rev Bone Miner Metab. 2016;14(2):56–67.
Seichi, A., et al. Development of a screening tool for risk of locomotive syndrome in the elderly: the 25-question Geriatric Locomotive Function Scale. J Orthop Sci. 17, 163–172 (2012).
Kobayashi, T., et al. Development of a simple screening tool based on the 5-question geriatric locomotive function scale for locomotive syndrome. J Orthop Sci. 27, 913–920 (2022).
Yoshida, N., et al. Average gut flora in healthy Japanese subjects stratified by age and body mass index. Biosci Microbiota Food Health. 41, 45–53 (2022).
Watanabe, S., et al. A cross-sectional analysis from the Mykinso Cohort Study: establishing reference ranges for Japanese gut microbial indices. Biosci Microbiota Food Health. 40, 123–134 (2021).
Bolyen, E., et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol. 37, 852–857 (2019).
Quast, C., et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 41, D590-D596 (2013).
Waked, J.P., et al. Model for Predicting Temporomandibular Dysfunction: Use of Classification Tree Analysis. Braz Dent J. 31, 360–367 (2020).
Takahashi, S., et al. Classification Tree Analysis Based On Machine Learning for Predicting Linezolid-Induced Thrombocytopenia. J Pharm Sci. 110, 2295–2300 (2021).
Ishikawa, Y., et al. Classification tree analysis to enhance targeting for follow-up exam of colorectal cancer screening. BMC Cancer. 13, 470 (2013).
Mubayi, A. Computational Modeling Approaches Linking Health and Social Sciences: Sensitivity of Social Determinants on the Patterns of Health Risk Behaviors and Diseases. Handbook of Statistics. Chapter 10, 249–304 (Elsevier, 2017).
Podgorelec, V., Kokol, P., Stiglic, B., & Rozman, I. Decision trees: an overview and their use in medicine. J Med Syst. 26, 445–463 (2002).
Therneau, T., & Atkinson, B. rpart: Recursive Partitioning and Regression Trees. R package version 4.1–15. https://CRAN.R-project.org/package=rpart (2019).
Kuhn, M. caret: Classification and Regression Training. R package version 6.0–92. https://CRAN.R-project.org/package=caret (2022).
R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/ (2021).
Zhong, F., et al. Effects of combined aerobic and resistance training on gut microbiota and cardiovascular risk factors in physically active elderly women: A randomized controlled trial. Front Physiol. 13 1004863; 10.3389/fphys.2022.1004863 (2022).
Ticinesi, A., et al. Gut Microbiota, Muscle Mass and Function in Aging: A Focus on Physical Frailty and Sarcopenia. Nutrients. 11, 1633; 10.3390/nu11071633 (2019).
Morley, J. E., et al. Frailty consensus: a call to action. J Am Med Dir Assoc. 14, 392–397 (2013).
Kimura, A., et al. Prevalence of locomotive syndrome in Japan: a nationwide, cross-sectional Internet survey. J Orthop Sci. 19, 792–797 (2014).
Nakamura, K. The concept and treatment of locomotive syndrome: its acceptance and spread in Japan. J Orthop Sci. 16, 489–491 (2011).
Cabinet office. Minutes of the 7th Meeting of the Council on Economic and Fiscal Policy: Annual Report on the Ageing Society. https://www5.cao.go.jp/keizai-shimon/kaigi/minutes/2015/0526/agenda.html (2015).
Yoshimura, N., et al. Prevalence and co-existence of locomotive syndrome, sarcopenia, and frailty: the third survey of Research on Osteoarthritis/Osteoporosis Against Disability (ROAD) study. J Bone Miner Metab. 37, 1058–1066 (2019).

(Not answered)

Download PDF

Editorial decision: revise
26 Sep, 2024
Review #2 received at journal
17 Sep, 2024
Review #1 received at journal
12 Sep, 2024
Reviewer #2 agreed at journal
02 Sep, 2024
Reviewer #1 agreed at journal
02 Sep, 2024
Reviewers invited by journal
22 Aug, 2024
Submission checks completed at journal
13 Aug, 2024
Editor assigned by journal
06 Aug, 2024
First submitted to journal
06 Aug, 2024

You are reading this latest preprint version

Association between gut microbiome and locomotive syndrome risk in healthy Japanese adults: A cross-sectional study

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Study design and participants

Measurements

Assessment of musculoskeletal function

Gut microbiome data

Statistical analysis

Results

Discussion

Declarations

References

Additional Declarations

Status:

Version 1