oa Applying Novel Body-Worn Sensors to Measure Stress: Does Stress Affect Wound Healing Rates in the Diabetic Foot?

Background and Aim: In the United States alone, diabetic limb complications and amputations are estimated to cost $17 billion. Significant risk factors that may lead to amputation of the diabetic foot include ineffective wound healing and infection of a wound or ulcer. Previous studies have shown that wound healing is slowed and patient's susceptibility to infection is increased when a patient is under chronic stress. To date, objective measures of stress have not been used to determine if stress affects the rate at which wounds heal. Our study used novel real-time monitoring of patient's heart rate variability to objectively determine the stress levels of patients visiting a surgery clinic for wound dressing changes. The wound healing rates of patients with high stress levels were compared to healing rates of low-stress individuals to assess the effect of stress on rates of wound healing among diabetics with a history of foot ulceration. Methods: Twenty patients (age: 56.7 ±  12.2 years) with diabetic foot ulcers were equipped with a chest-worn sensor (Bioharness 3, Zephyr Technology Corp., Annapolis, MD) during their 45-minute appointments where the patients’ wound was re-dressed. The chest sensor contained a uni-channel ECG recorder, and a novel algorithm was developed to determine heart rate variability from sensor output. Low frequency (0.04 to 0.15 Hz) HRV signals were isolated from high frequency signals (0.15 to 0.40 Hz), and the ratio of their amplitudes was used as a measure of stress. Patients were categorized as low-stress if the ratio of the signals was less than 1, and were otherwise categorized as high-stress individuals. Regardless of classification, each patient's wound size (length, width, depth) was recorded at baseline and in follow-up visits. High and low-stress patients were compared to see if wound sizes decreased more rapidly in either group. Results: Results indicate that patients with low levels of stress reduced their wound size by 79% between baseline and the first follow-up appointment (1.36 mm³ to 0.28 mm³). In contrast, patients with high levels of stress had adverse outcomes, with their wound sizes increasing nearly four times between baseline and follow-up (0.17 mm³ vs 0.84 mm³). Although high stress individuals had smaller wound sizes than low stress individuals initially (0.17 mm³ vs. 1.36 mm³, p <  0.05), the wound sizes of high stress individuals were nearly 3 times larger by the first follow-up (0.84 mm³ vs. 0.28 mm³, p =  0.10). Conclusion: Our research proposes that an individual's stress level can be objectively measured using an algorithm that processes ECG data from a single body-worn sensor that is lightweight and comfortable to wear. The stress levels measured with our algorithm are predictive of positive clinical outcomes. Specifically, individuals with low levels of recorded stress at baseline have faster healing rates and greater reductions in wound size by their second clinical appointment. This indicates that real-time patient stress monitoring using body-worn sensors may help clinicians identify risk factors that prolong wound healing times. In addition, it can be inferred that managing stress in diabetic patients will quicken the pace of wound healing. Surprisingly, however, our results suggest that initial wound sizes are not good indicators of stress levels in patients during initial clinical appointments. In fact, wound sizes of high stress individuals were significantly lower than low-stress individuals at baseline.