Skip to main content
. 2022 Apr 15;19:46. doi: 10.1186/s12966-022-01284-2

Fig. 2.

Fig. 2

The association of steps/day and percent of intense steps (> 100 steps/min) with incident diabetes, HCHS/SOL cohort (2008–2017). Compared to adults who accumulated 3400 steps/day and had 1% of intense steps (referent), a lower risk of diabetes was observed with greater steps/day, and at a given steps/day, a lower risk was found with greater percent of intense steps. a Diabetes based on self-reported diagnosis, medications, labs (3 criteria definition, n = 6634). b Diabetes based on medications and labs (2 criteria definition) n = 6633. Abbreviations: CI = confidence interval. Steps/day predicted estimates at the 2nd percentile (2000 steps/day), 25th percentile (5000 steps/day), 50th percentile (7000 steps/day), 75th percentile (10,000 steps/day), and 90th percentile (14,000 steps/day). Percent of intense steps (> 100 steps/min) predicted estimates at 50th percentile (10%), 75th percentile (20%), and 90th percentile (30%). All comparisons made to the referent of the 10th percentile of steps/day and percent of intense steps (3400 steps/day and 1%). Models adjusted for age (continuous), quadratic term for age, sex (male, female), Latino background by HCHS/SOL field center (17 level categorical variable), education (< high school/no GED, high school/GED, > high school), married/partner status (yes, no), employment (yes, no), years in the US (born in US, < 10 years, > 10 years), self-rated general health (excellent/very good, good, fair/poor), mobility limitations (yes, no), cigarette pack years (continuous), alcoholic drinks per week (continuous), energy intake (continuous), AHEI-2010 (continuous) and accelerometer wear time (continuous, hours per day)