Endocrine Abstracts

Objectives: Graves’ disease (GD) is a common cause of hyperthyroidism. An increased heart rate (HR) is a hallmark of elevated thyroid hormones (TH). In a prospective observational study we monitored HR continuously with wearable fitness trackers to evaluate the correlation between HR and TH levels and the relevance of HR monitoring for clinical management of patients with GD.

Methods: Seventeen outpatients with newly diagnosed or relapsed GD starting anti-thyroid drugs (ATD) were recruited for this study. Patients were invited to the endocrine outpatient clinic at the University Hospital Basel for five visits three to five weeks apart. At inclusion, they received a wearable fitness tracker for continuous HR monitoring during the study. At each visit we measured clinical parameters, downloaded data from the device, performed an ECG and determined serum TH levels (TSH, fT3, fT4). We assessed hyperthyroid symptoms with a questionnaire and adjusted ATD dose.

Results: Fifteen of the seventeen patients were female. Mean age was 37.1±12.9 years. At baseline, all participants were hyperthyroid with a suppressed TSH. The peripheral TH were elevated with a mean fT4 of 38.3±21.2 pmol/l and a mean fT3 of 15.7±9.3 pmol/l. The mean resting HR was 82.0±11.2 bpm. With ATD treatment TH decreased significantly over the course of the study, fT4 dropped to 14.8±6.3 pmol/l and fT3 decreased to 5.2±1.9 pmol/l. In parallel, the resting HR declined to 67.1±9.2 bpm (P < 0.0001 and p for trend <0.0001 for all parameters). Levels of fT4 and fT3 correlated significantly with resting HR when analyzed over all participants (R²=0.67 for both fT4 and fT3). Many physiological factors influence HR and the individual normal range of TH. To account for these inter-individual differences and to assess the relation between TH level and HR in more detail, we analyzed the individual HR and TH data using a nonlinear model based on an exponential growth equation. For most patients this function achieved a very good fit between the predicted and the actual TH values for any given HR. The median coefficient of determination (R²) for fT4 was 0.81 (IQR 0.66 to 0.94) and 0.87 (IQR 0.70 to 0.95) for fT3.

Conclusion: We can demonstrate a strong relation between TH levels and continuously monitored resting HR in patients with GD. The course of HR and TH levels can be modelled in individual patients. Wearable devices could potentially be used to tailor treatment of GD to the needs of the individual patient.