Mother's heart rate: a valid way to measure physical activity during pregnancy? Data from the FitMum RCT

Objectives We aimed to compare Garmin-estimated daily minutes of moderate-to-vigorous-intensity physical activity (MVPA) with daily minutes of MVPA assessed using heart rate (HR) thresholds based on maximum HR and HR reserve (HRR), respectively. Furthermore, we explored pregnancy-related HR changes. Methods The FitMum study included 220 healthy pregnant women before gestational age week 15+0 wearing Garmin vívosport activity trackers continuously until delivery. The activity trackers measured MVPA based on a proprietary algorithm including HR and accelerometry (model 1). We assessed MVPA as time above 64% of maximum HR with (model 2a) and without (model 2b) a minimum of 10 consecutive minutes of MVPA. In model 3, we assessed MVPA as time above 40% of HRR. Results Model 2a assigned less MVPA compared with model 1 (mean bias: -3.66 min/day). Model 2b assigned more MVPA compared with model 1 (mean bias: 16.82 min/day) and Model 2a (mean bias: 20.48 min/day). Model 3 assigned more MVPA compared with model 2b (mean bias: 12.00 min/day). Resting and average HR increased during pregnancy. Conclusion Using HR for assessing MVPA offers transparency. The Garmin algorithm (model 1) assigned more MVPA compared with the HR model (model 2a). Removing the 10-min bout criterion resulted in more MVPA (Model 2a vs 2b). Using the HRR for assessing MVPA (model 3) resulted in more MVPA compared with using the maximum HR (model 2b). Resting and average HR resembled findings from previous studies.