**Background:** Metabolic syndrome (MetS) is a global public health concern affecting approximately one billion people worldwide. It is a cluster of conditions including obesity, dyslipidemia, hypertension, hyperglycemia, and insulin resistance that increases risk for cardiovascular disease and type 2 diabetes. Physical activity (PA) is known to play an important role in preventing MetS, but most prior evidence relies on self-reported questionnaires subject to recall bias. Consumer wearable devices offer the potential for objective PA measurement in free-living conditions. This study aimed to examine the association between MetS status (classified into three groups: no-MetS, pre-MetS, and MetS) and objectively measured daily PA using a wearable device in Japanese office workers, and to investigate whether day of the week modifies this association.
**Methods:** This was a secondary analysis of data from the intervention group of a 90-day randomized controlled trial conducted between December 3, 2018 and March 2, 2019. Participants were full-time professionals, managerial, or clerical workers from five companies with >1000 employees in the Tokyo metropolitan area. Of 7437 eligible employees, 272 were enrolled and randomized; 179 were in the intervention group. After exclusions (declined participation, missing final questionnaire, unclassifiable MetS status, missing or unreasonable PA data, missing sleep or alcohol data), 163 participants (151 men, 12 women) with 8265 observations were included. Participants wore a Fitbit Versa 24 hours/day and completed questionnaires via a smartphone application. MetS status was classified using annual health check-up data according to Japanese Society of Internal Medicine 2005 criteria: obligatory criterion of waist circumference (≥85 cm men, ≥90 cm women) plus ≥2 of: triglycerides ≥150 mg/dL and/or HDL-C <40 mg/dL; systolic BP ≥130 mmHg and/or diastolic BP ≥85 mmHg; fasting blood glucose ≥110 mg/dL. Pre-MetS was defined as meeting the waist circumference criterion but only 1 non-obligatory criterion. PA outcomes were daily step count and active minutes (fairly active [3-6 METs] plus very active [>6 METs] minutes combined). Binary variables were created based on EPAR reference values (9000 steps/day men, 8500 women; ≥60 min MVPA/day). Multilevel mixed-effects logistic regression models were used, with Model 1 adjusting for age and sex; Model 2 additionally adjusting for day of week, eating habits, smoking, overtime hours, living arrangement, socioeconomic status, alcohol consumption, and total sleep time; and Model 3 further adjusting for health awareness composite score and self-rated health. Sensitivity analysis stratified by weekday vs weekend/holiday with interaction testing.
**Key Results:** Median age was 44.2 years (IQR 12.0). MetS prevalence was 31.8% in men (48/151) and 25.0% in women (3/12). The MetS group had the highest median step count (11,195 [IQR 5262] steps/day, p<0.001) and active minutes (56 [IQR 60.2] mins/day, p<0.001) among groups. Compared to no-MetS, the MetS group showed non-significant inverse associations with step count (Model 3: OR=0.81; 95% CI: 0.47, 1.40) and active minutes (Model 3: OR=0.94; 95% CI: 0.59, 1.49). The pre-MetS group showed significant inverse associations with step count (Model 3: OR=0.60; 95% CI: 0.36, 0.99) and active minutes (Model 3: OR=0.62; 95% CI: 0.40, 0.96), representing approximately 40% lower odds of meeting daily PA recommendations. In sensitivity analysis, day of week was a significant effect modifier for both PA outcomes (p<0.001). On weekdays, pre-MetS remained inversely associated with step count (Model 3: OR=0.57; 95% CI: 0.34, 0.95) and active minutes (Model 3: OR=0.61; 95% CI: 0.39, 0.95). On weekends/holidays, all MetS groups showed significantly lower step counts compared to weekday no-MetS (no-MetS Model 3: OR=0.35; 95% CI: 0.27, 0.44; MetS: OR=0.21; 95% CI: 0.12, 0.36; pre-MetS: OR=0.23; 95% CI: 0.13, 0.38). For active minutes on weekends/holidays, no-MetS showed non-significant positive association (OR=1.12; 95% CI: 0.91, 1.38), while MetS (OR=0.76; 95% CI: 0.47, 1.24) and pre-MetS (OR=0.71; 95% CI: 0.45, 1.12) showed non-significant inverse associations.
**Clinical Implications:** The finding that pre-MetS individuals had significantly lower odds of meeting PA recommendations suggests this group may be an important target for early intervention before progression to full MetS. The non-significant association in the MetS group may reflect that these individuals were already receiving mandatory health guidance under Japan's national screening program, potentially increasing their health awareness and PA levels. The significant effect modification by day of week indicates that PA patterns differ substantially between weekdays and weekends, with all groups showing reduced step counts on weekends. This suggests that interventions should promote consistent PA throughout the week, particularly for those with or at risk of MetS. The study's limitations include its cross-sectional design preventing causal inference, predominantly male sample (92.6%) limiting generalizability, and the winter study period (December-March) potentially introducing seasonal effects. Further longitudinal research with larger, more diverse samples is needed to confirm these findings.