LAUSR

Introduction Pre-operative exercise testing is widely used to assess perioperative risk. ‘Maximal’ tests, such as Cardiopulmonary Exercise Testing, are not always well tolerated. To overcome this sub-maximal exercise tests (SET) are increasingly being utilised. Though potentially better tolerated, results may be dependent on patient motivation and effort. Assessment of heart rate recovery (HRR), as a marker of underlying cardiac vagal activity, following SETs could potentially add objectivity. Several authors have unsuccessfully attempted to model HRR using a variety of non-linear functions(1). We hypothesised that a) an individual's HRR profile could be modelled using non-linear mixed effects modelling (NLME) and b) that the kinetics of an individual's HRR profile were the same, regardless of effort level. Methods Thirty-four healthy volunteers underwent three, six-minute SETs on a cycle ergometer. Individuals on beta-blockers or with contraindications to exercise testing were excluded. The first test was used to familiarise the volunteer with the test protocol and was undertaken at 20% of predicted maximum wattage (Wmax). The following two tests were delivered in a randomised order at 40% or 60% Wmax. Data on HRR was collected for 5-minutes on test cessation. HRR was modelled using the asymptotic regression function; Asymptote + (Maximum HR - Asymptote)*e-RateConstant*Time. Results The median age of the study population was 39 (range; 22-72) with 15% of the population possessing chronic co-morbidities. Figure 1 demonstrates the NLME model using a fit by maximum likelihood. Residual inspection revealed a homoscedastic distribution. The root mean square error (RMSE) for the model fitted values for both 40% and 60% Wmax was estimated at ±7bpm, demonstrating a good fit independent of work load. Modelling assumed that the asymptote and rate constant were fixed regardless of effort, however the maximum heart rate achieved on exercise cessation could vary between 40% and 60% Wmax. Discussion We suggest the rate constant (obtained from NLME modelling) could provide an index for an individual's fitness independent of patient work load. Further work is required to assess the minimum increase in HR required during SETs in order to establish a valid HRR profile and the influence of different types of exercise on HRR. SETs are easier to conduct, better tolerated and a cheaper more widely available alternative to maximal exercise testing. After further validation, we plan to begin clinical testing of this novel concept which has the potential to provide widespread access to objective exercise testing which can be applied easily in any pre-operative clinic scenario.