Trends of the Incidence and Prevalence of Metabolic Syndrome Among Participants of the Speci c Health Check and Guidance System in Japan

Kunitoshi Iseki (  chihokun_ohra@yahoo.co.jp ) The Japan Speci c Health Checkups study (J-SHC study) Group TSUNEO KONTA The Japan Speci c Health Checkups study (J-SHC study) Group KOICHI ASAHI The Japan Speci c Health Checkups study (J-SHC study) Group SHOUICHI FUJIMOTO The Japan Speci c Health Checkups study (J-SHC study) Group TOSHIKI MORIYAMA The Japan Speci c Health Checkups study (J-SHC study) Group KUNIHIRO YAMAGATA The Japan Speci c Health Checkups study (J-SHC study) Group KAZUHIKO TSURUYA The Japan Speci c Health Checkups study (J-SHC study) Group ICHIEI NARITA The Japan Speci c Health Checkups study (J-SHC study) Group MASAHIDE KONDO The Japan Speci c Health Checkups study (J-SHC study) Group MASATO KASAHARA The Japan Speci c Health Checkups study (J-SHC study) Group YUGO SHIBAGAKI The Japan Speci c Health Checkups study (J-SHC study) Group CHIHO ISEKI Okinawa Renal and Heart Association (OHRA) TSUYOSHI WATANABE The Japan Speci c Health Checkups study (J-SHC study) Group


Introduction
A speci c nationwide health check-up and guidance system, called Tokutei-Kenshin, was initiated in April 2008 in Japan 1 . This project aims to detect metabolic syndrome (MetS) and if con rmed, to provide individual instruction to modify lifestyle and the necessary treatment. The target population comprises Japanese citizens between the ages of 40-74 years. We have been focused mostly on chronic kidney disease (CKD) [2][3][4] , diabetes mellitus (DM) 5,6 , and mortality [7][8][9][10] . Lifestyle is a signi cant modi er of CKD 11 , cardiovascular disease (CVD) 12 , and mortality [12][13] . We recently con rmed that MetS was a signi cant risk factor of mortality 14 .
Intervention through this screening program was shown to be very effective for the reversal of MetS 15 . However, the proportion of people attending the intervention program is as low as 11%. The trends in the incidence and persistent prevalence of MetS have not been well studied. Such information would be helpful for the future modi cation of the protocol of screening among the Japanese population. Lifestyle modi cation if convinced by the screened participants would prevent the incidence of MetS and reduce the persistent prevalence of MetS.
In the present study, we calculated the incidence, prevalence. and persistent prevalence of MetS using the subjects who were screened at two consecutive years (1year) between 2008 and 2014. Also, we examined the total screened subjects as many of them were screened more than twice.

Page 4/14
The baseline prevalence of MetS was summarized in Table 1. As a whole, they were 10.9% (total), 17.0% (men), and 6.8% (women), respectively. Men had more than twice the higher prevalence of MetS than that of women. Figure 1 showed the prevalence of MetS in each year of screening. The incidence of MetS was summarized in Table 2. As a whole, they were 5.7% (total), 9.6% (men), and 3.5% (women), respectively. Men had more than twice the higher incidence of MetS than women. Figure 2 showed the 1-year incidence of MetS from the baseline year of screening. The persistent prevalence of MetS was 47.3% (total), 49.6% (men), and 43.3% (women), respectively (Table 3, Figure 3).
Baseline characteristics of subjects who remained without MetS and developed MetS were summarized in (Supplementary Table 1). Those who developed MetS were a signi cantly higher proportion of men and older age. Other variables except renal failure/ dialysis were also signi cantly different as the number of subjects was large.
Baseline characteristics of subjects who remained MetS and disappeared MetS were summarized in (Supplementary Table 2). Those who remained MetS were a signi cantly higher proportion of men and older age. Other variables except for current smokers and history were also signi cantly different.
Baseline characteristics by gender were shown in Supplementary Table 3. In men, the prevalence of smokers, DM, and history of CVD and stroke were more than twice as high compared to women. However, the current smoker was less in the elder-age group (≥65 years) than in men (Supplementary Table 4). In women, the current smoker was also higher among the younger age group (<65 years) (Supplementary Table 5).

Discussion
This study examined the 1-year incidence and persistent prevalence of MetS among the nationwide screening program during the scal year of 2008 to 2014 (6 years). As a whole, the 1-year incidence was 9.6% in men and 3.5% in women, and the persistent prevalence of MetS was 49.6% in men and 43.3% in women, respectively. Signi cant differences in sex and age group were evident in both the 1-year incidence and persistent prevalence of MetS. Men and the elderly group (age≥65) population were at risk of higher 1-year incidence and persistent prevalence of MetS. Subjects with MetS have a signi cant impact on the incidence of CKD 2, 3, 4 , diabetes mellitus (DM) 5,6 , and mortality [7][8][9][10] . Accordingly, when diagnosed MetS, they are entitled to further examination and lifestyle intervention. Lifestyle per se is a signi cant modi er of CKD 11 , CVD 12 , and mortality [12][13] . Recently, we con rmed that MetS was a signi cant risk factor of mortality 14 . Therefore, early intervention in overweight/obese adults, namely MetS subjects is necessary to prevent the progression of CKD 16, 17 and death.
The dipstick proteinuria test for CKD detection was shown to be cost-effective 18, 19 , however not yet shown for the diagnosis of MetS at the general screening. The key strategy for the prevention of MetS is to keep body weight within the normal range by nutritional management and adequate exercise. in particular aged populations. Intervention through this screening program was shown to be very effective for the reversal of MetS 15 . However, the proportion of people attending the program as low as 11%. The trends in the 1-year incidence and persistent prevalence of MetS have not been well studied. Such information would be helpful for the future modi cation of the protocol of screening among the Japanese population.
Lifestyle modi cation if convinced by the screened participants would prevent the overall incidence of MetS and reduce the prevalence of MetS. Other than the weight reduction in overweight and obese subjects, excess alcohol intake, in particular, men, is frequently observed with MetS. Alkerwi A et al recommended restricting alcohol consumption of less than 20 g/day among women, and of less than 40 g/day among men 20 . Other lifestyles such as depression 21 and self-reported sleep duration 22 are reported to the associated with MetS. These observations need to be con rmed among the Japanese but are suggesting the importance of further questionnaires among apparently healthy people.
MetS were de ned as waist circumference (men≥85cm, women≥90cm) plus two or three abnormal values in blood sugar metabolism, lipid, and blood pressure 14 . Waist circumference is a surrogate of central obesity but is often variable by body size, gender, and race. We reported the signi cance of "a body shape index (ABSI)" on all-cause mortality among screened subjects 10 . ABSI seemed to be a better predictor of death than that of the body mass index (BMI). However, the presence of CKD affected differently on mortality between men and women.

Strengths and limitations
The strength of the present study is that we have followed a large number of participants of the nationwide screening program. We believe that this cohort represents the currently available database for the analysis of the changes in MetS status in Japan.
There are several limitations in the present study. First, participants in this analysis were those who had an interest in lifestyle and their health condition. The participation rate was 38.9% (2008) and 51.4% (2016) of the target population. (Ministry of Health, Labour and Welfare) Therefore, it would not be representative of the whole Japanese population. We have no data on whether the MetS (+) individuals had attended the intervention program or not. A previous study showed that men and relatively younger (age<65 years) had a lower participation rate compared to their counterparts 19 . Second, other socioeconomic factors related to the incidence and prevalence of MetS are unknown in this screening program. The number of family members, the presence of a spouse, and the location of residence might in uence the lifestyle. Third, factors other than the differences in lifestyles and history of CVD, stroke, and renal failure at baseline may explain the results of the present study. Long-term lifestyle would be di cult to change by single intervention, in a particular elderly population. Fourth, we have no data of those aged 75 and over. The medical care system for the elderly in the later stage of life has also started in 2009 in Japan. Further studies on transition to this program may be warranted. Finally, other limitations on the diagnosis of MetS have been discussed in the previous paper 14 .

Conclusions
This study showed the prevalence, 1-year incidence, and the persistent prevalence of MetS and found that they remained high level. More provocative strategies are needed for subjects of aged (≥65) men.

Methods
Details of the dataset of the nationwide screening program of the Speci c Health Check-up and Guidance System (Tokutei-Kensin) in Japan have been previously published [2][3][4][5][6][7][8][9][10][11][12][13][14] . From 2008 to 2014, we collected individual records of 3,809,853 participants from 192 municipals of 27 prefectures. The process for the database construction is summarized in Figure 4. From the total number of subjects who participated (about 3.8 million), we selected those who visited at least twice during the study visit. Finally, we used the dataset of 933,490 participants for the analysis. During the study period, the mean number of visits was 3.4 times per subject. MetS was de ned as waist circumference (men≥85cm, women≥90cm) plus two or three abnormal values in blood sugar metabolism (fasting blood glucose≥100mg/dL or HbA1c≥5.2% by 2012 (JDS), HbA1c≥5.6% by NGSP since 2013), lipid (triglyceride≥150mg/dL, or HDL cholesterol<40mg/dL), and blood pressure (systolic≥130mmHg, or diastolic≥85mmHg) 14 . From 2008 to 2013, the number of participants who had enough information for the diagnosis of MetS was 894,628 (more than twice), 752,809 (more than thrice), and 682,548 (1-year later), respectively. The original database was solely used and managed by Okinawa Heart and Renal Association (OHRA). Furthermore, the preliminary dataset was veri ed and con rmed independently by Dr. Tsuneo Konta. Afterward, further analyses were done by using a standard analysis le (SAF) without any personal identi er.
The prevalence of MetS was calculated as the denominator was the number of people at the year of screening and the numerator was the number of MetS among them. The 1-year incidence of MetS was calculated as the denominator was the number of people with MetS (-) at the rst checkup and the numerator was the number of developed MetS among them at the next year's screening. For the calculation of the 1-year persistent rate of MetS, we used the denominator was the number of people who are MetS (+) at the rst checkup and the numerator was the number of people who remained MetS (+) at the next year screening.

Statistical analysis
Data were analyzed with SAS/STAT software (version 6.03, SAS Institute, Tokyo, Japan). The student's ttest and the Chi-squared test were performed to compare the signi cance of discrete variables. Multivariate Cox regression analysis was performed to evaluate the risks for the changes in the trend of the 1-year incidence and persistent prevalence of MetS. Factors used for the adjustment were body mass index, systolic and diastolic blood pressure, fasting blood glucose, HbA1c, triglyceride, HDL cholesterol, LDL cholesterol, eGFR, proteinuria, alcohol intake, smoking, history of stroke, acute myocardial infarction renal failure, dialysis, and drug use for hypertension, diabetes mellitus, and hyperlipidemia and were based on self-reported information in the medical questionnaire. Hazard ratio and 95% con dence interval were obtained. A P value of less than 0.05 was considered statistically signi cant in all analyses.

Declarations
The authors state they have no Con ict of Interest (COI).     1-year incidence of metabolic syndrome among those who were free of metabolic syndrome at the rst screening.
Page 14/14 Figure 4 1-year persistent prevalence of metabolic syndrome among subjects who were diagnosed with metabolic syndrome at the rst screening.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.