Reference values for accelerometer metrics and associations with cardiorespiratory fitness: a prospective cohort study of healthy adults and patients with heart failure

Abstract

Background

Accelerometry has gained increasing popularity and yields numerous physical activity (PA) outcomes (Rowlands et al., 2019). These include traditional cut-point-based (i.e. light, moderate, and vigorous PA) and cut-point-free metrics (i.e. intensity gradient [IG] and average acceleration [AvAcc]). IG reflects the intensity distribution of PA across the day (Rowlands et al., 2018; Fairclough et al., 2019). AvAcc is a proxy for the daily volume of PA ( Rowlands et al., 2018; Fairclough et al., 2019). Cut-point-based metrics are commonly expressed in minutes per day, making their interpretation simple (Troiano et al., 2014). Yet, the measured acceleration needs to be categorised by setting population- and device-dependent cut-points to obtain these metrics (Troiano et al., 2014). Cut-point-free metrics, on the other hand, are comparable across studies, accelerometer brands (Migueles et al., 2022), and diverse populations (Rowlands et al., 2018). However, their interpretation is not easy. Besides, it is unknown how cut-point-free metrics are associated with cardiorespiratory fitness (CRF), an important health indicator in healthy individuals and patient populations with impaired CRF (Kodama et al., 2009). We thus aimed to 1) compare the association of CRF with cut-point-free metrics to that with cut-point-based metrics in a prospective cohort of healthy adults aged 20 to 89 years and patients with heart failure, and 2) provide age-, sex-, and CRF-related reference values for healthy adults.

Methods

The COmPLETE study was cross-sectional. Healthy individuals were recruited via unaddressed letters sent to randomly selected postal districts in the Basel area (Wagner et al., 2019). Patients with heart failure were approached as described elsewhere (Wagner et al., 2019). Subjects were asked to wear GENEActiv accelerometers on their non-dominant wrist for up to 14 days and undergo cardiopulmonary exercise testing on a cycle ergometer to determine CRF. Raw accelerometer data were processed using the R-package GGIR (Migueles et al., 2019; van Hees et al., 2013). Associations between CRF and accelerometer metrics were examined using multiple linear regression models adjusted for sex, age, and body mass index. Percentile curves were generated with Generalised Additive Models for Location, Scale, and Shape (Stasinopoulos & Rigby, 2008).

Results

Four hundred and sixty-three healthy adults and 67 patients with heart failure were included in the analyses. IG and AvAcc provide complementary information on PA. Both metrics were independently associated with CRF in healthy individuals. The best cut-point-free regression model (AvAcc+IG) performed similar to the best cut-point-based model (vigorous activity) and explained 73.9% and 74.2% of the variance in CRF, respectively. In patients with heart failure, IG was associated with CRF, independent of AvAcc. Cut-point-free models (IG+AvAcc, IG alone) had comparable predictive value for CRF as the best cut-point-based metric (moderate-to-vigorous activity). We produced age-, sex-, and CRF-related reference values for IG, AvAcc, moderate-to-vigorous, and vigorous activity for healthy adults. Moreover, we developed a web-based application (rawacceleration) facilitating the interpretation of cut-point-free metrics.

Conclusions

Cut-point-free metrics are not only more robust than cut-point-based metrics, but also have similar predictive value for CRF and, in turn, indirectly for the risk of mortality and longevity (Kodama et al., 2009; Mok et al., 2019). This may be the case in both healthy individuals and patients with heart failure. Our findings together with those of previous studies (Rowlands et al., 2018; Fairclough et al., 2019), therefore, provide a rationale that cut-point-free metrics facilitate the capture of the volume and intensity distribution of the PA profile across populations, and thus may be a viable alternative to cut-point-based metrics in describing PA. Our reference values will enhance the utility of IG and AvAcc and facilitate their interpretation. Finally, our web-based application will simplify this process and also support the translation of cut-point-free metrics into meaningful outcomes.

References

Fairclough, S. J., Taylor, S., Rowlands, A. V., Boddy, L. M., & Noonan, R. J. (2019) Average acceleration and intensity gradient of primary school children and associations with indicators of health and well-being. Journal of Sports Sciences, 37(18), 2159-2167. https://doi.org/10.1080/02640414.2019.1624313

Kodama, S., Saito, K., Tanaka, S., Maki, M., Yachi, Y., Asumi, M., Sugawara, A., Totsuka, K., Shimano, H., Ohashi, Y., Yamada, N., & Sone, H. (2009). Cardiorespiratory fitness as a quantitative predictor of all-cause mortality and cardiovascular events in healthy men and women: A meta-analysis. JAMA, 301(19), 2024-35.https://doi.org/10.1001/jama.2009.681

Migueles, J. H., Molina-Garcia, P., Torres-Lopez, L. V., Cadenas-Sanchez, C., Rowlands, A. V., Ebner-Priemer, U. W., Koch, E. D., Reif, A., & Ortega, F. B. (2022). Equivalency of four research-grade movement sensors to assess movement behaviors and its implications for population surveillance. Science Reports, 12, Article 5525. https://doi.org/10.1038/s41598-022-09469-2

Migueles, J. H., Rowlands, A. V., Huber, F., Sabia, S., & van Hees, V. T. (2019). GGIR: A research community–driven open source R package for generating physical activity and sleep outcomes from multi-day raw accelerometer data. Journal for the Measurement of Physical Behaviour, 2(3),188-96. https://doi.org/10.1123/jmpb.2018-0063

Mok, A., Khaw, K.-T., Luben, R., Wareham, N., & Brage, S. (2019). Physical activity trajectories and mortality: Population based cohort study. BMJ, 365, l2323. https://doi.org/10.1136/bmj.l2323

Rowlands, A. V., Edwardson, C. L., Davies, M. J., Khunti, K., Harrington, D. M., & Yates, T. (2018). Beyond cut points: Accelerometer metrics that capture the physical activity profile. Medicine & Science in Sports & Exercise, 50(6), 1323-32. https://doi.org/10.1249/MSS.0000000000001561

Rowlands, A. V., Fairclough, S. J., Yates, T., Edwardson, C. L., Davies, M., Munir, F., Khunti, K., & Stiles, V. H. (2019). Activity intensity, volume, and norms: Utility and interpretation of accelerometer metrics. Medicine & Science in Sports & Exercise, 51(11), 2410-2422. https://doi.org/10.1249/MSS.0000000000002047

Stasinopoulos, D. M., & Rigby, R. A. (2008). Generalized additive models for location scale and shape (GAMLSS) in R. Journal of Statistical Software, 23(7), 1 - 46. https://doi.org/10.18637/jss.v023.i07

Troiano, R. P., McClain, J. J., Brychta, R. J., & Chen, K. Y. (2014). Evolution of accelerometer methods for physical activity research. British Journal of Sports Medicine, 48(13), 1019-1023. https://doi.org/10.1136/bjsports-2014-093546

van Hees, V. T., Gorzelniak, L., Dean León, E. C., Eder, M., Pias, M., Taherian, S., Ekelung, U., Renström, F., Franks, P. W., Horsch, A., & Brage, S. (2013). Separating movement and gravity components in an acceleration signal and implications for the assessment of human daily physical activity. PloS one, 8(4), Article e61691. https://doi.org/10.1371/journal.pone.0061691

Wagner, J., Knaier, R., Infanger, D., Arbeev, K., Briel, M., Dieterle, T., Hanssen, H., Faude, O., Roth, R., Hinrichs, T., & Schmidt-Trucksäss, A. (2019). Functional aging in health and heart failure: The COmPLETE Study. BMC Cardiovascular Disorders, 19, Article 180. https://doi.org/10.1186/s12872-019-1164-6