**Background:** A step cadence of 100 steps/minute is widely used to define moderate-intensity walking, but its generalizability to older adults is uncertain because the metabolic cost of walking is higher in older populations. This study aimed to investigate moderate-intensity step cadence values during treadmill walking and daily living in older adults.
**Methods:** Data came from the STAND-UP study, a cross-sectional analysis of community-dwelling adults aged ≥60 years recruited from urban venues (2014–2016). Participants wore an activPAL3™ thigh-worn device for 7 days of free-living physical activity monitoring, then attended a laboratory visit. Resting metabolic rate (RMR) was measured via 60-minute supine breath-by-breath indirect calorimetry (Cortex Metalyzer). Participants then completed treadmill walking at 1, 2, 3, 4, and 5 km/h in randomized order, each stage lasting 5 minutes, with expired gas collected between minutes 2–4. Steps were counted by activPAL3™. METS_standard were calculated as VO₂ (mL/kg/min)/3.5; METS_relative were calculated as VO₂ during activity divided by individual RMR VO₂. Moderate-intensity was defined as ≥3 METS_standard. Segmented generalized estimating equations modeled the association between step cadence and MET values with a breakpoint at 100 steps/minute. ROC curve analysis tested classification accuracy.
**Key Results:** 53 participants were included (median age 75 years, BMI 28.0 kg/m², 45.3% women). At 2 km/h, median METS_standard was 3.08 (IQR 2.73, 3.43) and METS_relative was 3.62 (IQR 3.27, 4.30) at a median cadence of 81.0 (IQR 72.0, 88.7) steps/minute. The predicted cadence at 3 METS_standard was 70.3 (95% CI 61.4, 75.8) steps/minute. At 100 steps/minute, predicted METS_standard was 3.77 (95% CI 3.54, 4.01) and METS_relative was 4.74 (95% CI 4.43, 5.04). The gradient was 0.026 METS_standard per 1 step/minute. ROC analysis showed a threshold of 70 steps/minute correctly classified 80% of observations (sensitivity 93.4%, specificity 37.6%, PPV 82.6%, NPV 64.4%), while 100 steps/minute correctly classified 62.4% (sensitivity 54.6%, specificity 87.1%). During free-living, participants took a median of 6988 (IQR 5933, 9211) steps/day, with 2554 (IQR 1297, 4456) steps/day in bouts ≥1 minute. Of bouted steps, 96.4% (IQR 90.7%, 98.9%) occurred at ≥70 steps/minute, and 67.0% (IQR 38.9%, 82.3%) at ≥100 steps/minute. Results were consistent across strata of age, sex, BMI, height, physical function, and physical activity.
**Clinical Implications:** A step cadence threshold of 70 steps/minute may be more appropriate than 100 steps/minute for classifying moderate-intensity walking in older adults. The vast majority of continuous daily stepping already occurs above this threshold, suggesting that simply promoting increased walking time or distance may effectively promote moderate-intensity activity in this population. However, the low specificity at 70 steps/minute (37.6%) indicates a higher risk of misclassifying low-intensity steps as moderate-intensity compared to 100 steps/minute (specificity 87.1%). The study's limitations include a relatively small sample size, treadmill walking in a postprandial state that may have elevated MET values, and lack of metabolic data during free-living conditions. Further research is needed to refine thresholds for older adults and populations with accelerated aging or long-term conditions.