Research investigating the undesirable outcomes of residency training mostly focuses on the effects of stress and long work-hours on patient outcomes (1), as well as physician satisfaction (2), mental health (3–5), and recruitment and retention (6, 7). There is little research, however, examining the physiological and physical health implications of the stress of such training on physicians themselves. Although current research falls short of fully understanding the health effects of medical training, there are some data, mostly from other populations, that our own “health practices” are ruining our health.
With the newer and more restrictive set of the Accreditation Council for Graduate Medical Education (ACGME) residency requirements recently implemented in July 2011, it is an especially salient time to consider further research into the health effects of residency training that might knowledgeably inform policy organizations. Although there may be historical reasons and advantages for the current residency paradigm, such as healthcare demand and training experience, it is nevertheless important that we understand the physiological and neurobiological, or psychoneuroimmunological (PNI), implications of residency training on our young doctors.
The stresses on today’s residents seem perhaps greater than they were several decades ago (8), although there are no direct comparison studies. Indeed, many residents report experiencing much stress and anxiety, with undesirable repercussions extending beyond their work life, although such feelings are rarely discussed (9, 10). For purposes of this article, we will consider “stress” to include sleep deprivation, perceived psychological stress, mood alterations, and social isolation.
The Institute of Medicine issued a report in 2008 setting guidelines for resident duty-hours to enhance sleep, supervision, and safety (11), spurred, in large part, by the findings that adverse events, including death, were the result of sleep-deprived and over-extended interns and residents (12). These recommended changes are undeniably important to curtail such events; however, they fail to recognize the mental and physical health-related effects of such conditions on its trainees. Whereas 45% of outpatients visiting primary care physicians report being excessively bothered by stress (13), we physicians may be less adept at practicing what we preach to our patients. By comparison, residents report a greater number of health complaints than the average worker (14). It is imperative to the integrity of the functioning of our healthcare system that we as physicians care for each other and ourselves as we would our patients. Further knowledge regarding the health effects of residency training could greatly inform healthcare policy creation.
The stress-promoting effects of the practice of medicine as an attending physician are also of great importance. Because of its unique conditions, however, residency exemplifies the definition of effort–reward imbalance; that is, high-cost/low-gain conditions, which have been associated with adverse health outcomes (15). Further discussion of stress effects on attending physicians should be considered separately.
Sleep-loss or deprivation is a stress-provoking experience for most residents, both physically and psychologically. Sleep research is complex and variable, depending upon the population and the definitions used; however, there is evidence that acute sleep-loss results in disrupted affect-regulation and an increase in somatic complaints, anxiety, depression, and paranoia in healthy adults (16). The stress of sleep deprivation is associated with temporary changes in cerebral metabolism, cognition, emotion, and behavior, consistent with mild prefrontal lobe dysfunction (17), as well as decreased reaction-time speed and vigilance (18).
There is a small body of literature investigating the effects of sleep deprivation specifically in medical residents, demonstrating that young doctors are as susceptible to its effects as nonresidents. They exhibit significant slowing in cognitive processing and decline in reaction times on vigilance-testing post-call (19, 20); another study demonstrated 84% scoring on the sleepiness scale within the range for which clinical intervention is indicated (21). These deficits translate into functional and job impairments, as well (22–25).
Sleep deprivation itself is associated with altered immune functioning, including impaired host–defense mechanisms, increased susceptibility to pathogens, decreased natural killer (NK) cell numbers and activity, and increased proinflammatory cytokine profiles (26–29). Indeed, sleep deprivation is a risk factor for human disease; however, objective and subjective measures of the effects of residents’ sleep deprivation differ, making mere resident questioning insufficient for evaluation of sleep (18).
Perceived Stress and Mood Alterations
Sleep, stress, and mood are inextricably linked, with residents repeatedly reporting sleep-loss and fatigue as having a major adverse impact on their mood, relationships, and quality of life (21, 30). Furthermore, interns report higher levels of burnout and depression and lower levels of empathy at the end of their intern year, as compared with the beginning (30). Indeed, there is a conglomeration of detrimental psychological responses that seem to result from medical training, which have their own physiological consequences.
The adverse physiological and health effects of stress and negative mood are well established (31–37). A few of these physiological responses include elevated white blood-cell count, decreased NK cell activity, and altered cortisol secretion (38, 39). Likewise, stress is associated with elevated hormone levels that result in reactivation of viruses, as well as decreased lymphocyte proliferation and antibody production (40). This cascade of neurohormonal changes ultimately significantly affects health by delayed wound-healing, impaired responses to infection and vaccines, and cancer progression, among other processes. Although little research has explicitly focused on altered physiological functioning in medical residents, data from other studies clearly demonstrate that conditions associated with residency training promote poor immune functioning.
Most studies examining residents’ stress physiology center around being on-call, although there are certainly other stressors, such as novel exposure to traumatic experiences as a new intern. Increased on-call stress is inherently confounded with sleep-deprivation; nevertheless, significantly elevated or altered levels of adrenocorticotropic hormone, cortisol, testosterone, and luteinizing hormone have been observed post-call and are associated with increased levels of stress, anxiety, and depressive symptoms (41, 42). On-call surgical residents achieve stress levels of tachycardia and exhibit a significant increase in their level of circulating white blood cells (43). Recently, Brant et al. (44) found that although residents did not report a change in stress levels over the course of a rotation, they exhibited evidence of neuro-endocrine changes consistent with hypothalamic–pituitary–adrenal axis dysregulation, suggesting that residents may be under-reporting physiologically meaningful psychological stressors.
Residents report suffering from shrinking social networks and reduced marital satisfaction during their time of training (45). The psychophysiological implications of social isolation and stress are well established among the general population (46). Social isolation predicts morbidity and mortality from multiple disease processes, possibly via excessive stress reactivity and impaired physiological repair or maintenance processes, as well as health behaviors (47). Collegial social support and job control are important factors contributing to the stress-related health of residents (14). Research is needed focusing on the physiological and health effects, both short- and long-term, of social isolation or support in residents.
Most of the existing studies examining the physiological and neurobiological consequences of medical training suffer from methodological limitations, including small sample sizes, confounding variables, and lack of adequate controls. Furthermore, interpretation of findings in this area is variable. Clearly, more research is needed to investigate the physiological effects and their associated clinically-relevant health outcomes of medical training, given that other relevant studies provide evidence that resident conditions promote poor health via multiple mechanisms and pathways.
To my knowledge, there are no studies evaluating long-term health sequelae of residency or healthcare practice-related stress. There is indication from other literature, however, that there are, indeed, long-term effects. A meta-analysis showed that short sleep duration of less than 5–6 hours per night is associated with greater risk of developing or dying from cardiovascular disease (CVD) (48). Short sleep duration is associated with multiple physiologic mechanisms that may implicate it as a cause of CVD, including changes in circulating levels of leptin and ghrelin (49), which facilitate the development of obesity (50) and insulin-resistance (51). Other candidate mechanisms relating decreased sleep and undesired health outcomes include increased cortisol secretion and altered growth hormone (52). Short sleep duration activates an inflammatory response with implications not only for CVD, but for other chronic diseases, including cancer (53).
Other residency-associated stressors are also associated with mechanisms implicated in adverse longer-term health outcomes. Burnout is associated with elevated cortisol and development of CVD in other populations (54), and there is an enormous literature on psychological stress and its short- and long-term impact on CVD (55). Psychologically adverse work environments, in general, may accelerate immuno-senescence (56).
With the new residency requirements implemented in July 2011, it is anticipated that these new guidelines will be the necessary impetus to thoroughly investigating the effectiveness of these restrictions in mitigating the psychophysiological repercussions of residency-associated work, sleep, and stress conditions. This will be important to determine, as, even with night-float systems in place, objective sleepiness among residents can still be an issue (57). Some purport that perhaps the increased on-call workload and intensity actually undermines resident well-being (58), whereas others demonstrate a quality-of-life improvement after the 2003 reforms (59). The efficacy of work-hour restriction on sleep and performance seems to be dependent not only on the number of sleeping hours, but it also involves the pattern of sleep loss and whether residents are actually sleeping more or differently (57, 60). Even modest sleep loss is associated with altered T and NK cell functioning (61), as well as significant sleepiness, psychomotor impairment, and increased secretion of proinflammatory cytokines (62). Vgontzas et al. conclude that “given the potential association of these behavioral and physical alterations with health, well-being, and public safety, the idea that sleep or parts of it are optional should be regarded with caution” (62).
Given both the emphasis on preventive care to better control the leading causes of morbidity and mortality (46, 63), as well as the newly-revised duty-hour restrictions, it is an especially salient time to consider that although our youngest physicians train to become the next cohort of healthcare providers, they routinely endure potentially health-depleting conditions. Data from the PNI literature provide us with much insight into some of the physiological and neurobiological outcomes that residency training may incite, and little has been done to investigate this population, in particular. Future research could also investigate potentially restorative interventions, such as resident support groups and other stress-relieving practices. Although residency is a distinct and somewhat extreme situation, in which trainees have little control over their environment, research investigating potential pathophysiological outcomes in this population may be extrapolated to attending physicians to improve the health of our healthcare providers overall.
The author thanks Jan A. Moynihan, Ph.D., and Jeffrey M. Lyness, M.D., of the University of Rochester Department of Psychiatry, for their input and editorial comments during manuscript preparation.
This work was supported by the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant UL1RR031973, as well as NIH Grant T32 073452. The content is solely the responsibility of the author and does not necessarily represent the official views of the NIH.