Physicians-in-training are particularly susceptible to fatigue and chronic sleep deprivation given their exposure to prolonged work hours and rotating shift-work schedules. In 2003, the Accreditation Council for Graduate Medical Education (ACGME) limited resident duty hours to 80 per week hoping to address stress and time-on-task fatigue in residency as well as improve patient safety.1 Sleep deprivation has been shown to alter performance in this and other populations.2–5. Traditionally, medical professionals have ignored the importance of adequate sleep in their personal well-being. For example, the majority of residents in one program believed they had learned to tolerate sleep deprivation and that it was an expected part of training.6 At the same time, many studies have shown that internal medicine residents are experiencing high levels of stress, increased burnout, and depressive symptoms that may ultimately compromise patient care.7–9 Possible explanations for the mood disturbances in this population include sleep deprivation,10,11 but the link has not been conclusively established. In other populations, periods of sleep loss are associated with mood changes, irritability, and difficulty concentrating.12 Although several models can be used to explain the link between sleep deprivation and depression,13–15 to our knowledge no study has examined the association between acute or chronic sleep deprivation and depression in residents.
Accordingly, in this study we sought to better characterize the relationships between sleep deprivation and the evolution of mood disturbances, empathy, and burnout among a cohort of interns studied over the year prior to institution of duty hour restrictions. We hypothesized that the prevalence of mood disturbances, lowered empathic capacity, and increased burnout would increase from beginning to end of the first postgraduate year, and would be associated with decreases in sleep amounts and increases in sleepiness.
This study was part of a larger protocol looking at the impact of various countermeasures on sleep deprivation during internship. We conducted this study in 2002–03 in the internal medicine residency program at the University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania. The university's institutional review board approved the study.
The internship year consisted of one to three months in the intensive care units with call every third night and every sixth day off, seven months on acute care medicine services with call every fourth night and an average of one day in seven off, two to three months in ambulatory medicine/elective rotations with no required call, and one month of vacation. Interns rotated to three clinical sites.
In addition to providing general demographic information, each intern completed the following instruments, described in detail below, at baseline and at year end: sleep quantities, Epworth Sleepiness Scale (ESS), the Beck Depression Inventory–Short Form (BDI-SF), the Interpersonal Reactivity Index (IRI), and the Maslach Burnout Inventory-Human Services Survey (MBI-HSS).
Acute and chronic sleep deprivation were assessed with single items asking hours of sleep the prior night and prior seven days, respectively. Analytically they were defined as <six hours in the prior night and <42 hours in the prior week.3,16 Subjective sleepiness was measured using the ESS, an eight-item rating scale assessing the propensity to fall asleep in various life situations. Validity studies suggest a score of 11 or greater out of 24 indicates an elevated degree of sleepiness.17 The BDI-SF is a 13-item self-report of specific attitudes and symptoms of depression. A score of 8 or greater out of a possible 39 is indicative of moderate depression. Scores on the shorter version are highly correlated with the well-studied longer version.18–19 The IRI is a 28-item instrument consisting of four different seven-item subscales: perspective taking, empathic concern, fantasy, and personal distress. The perspective-taking subscale assesses the respondent's tendency to spontaneously adopt the psychological viewpoint of others. The empathic concern subscale assesses “other-oriented” feelings of sympathy and concern for unfortunate others. The fantasy subscale measures respondent ability to transpose themselves imaginatively into the feelings and actions of fictitious characters. The personal distress subscale measures “self-oriented” feelings of anxiety and unease in interpersonal settings. Subscale scores are said to be reliable and accurate indicators of social functioning, self-esteem, emotionality, and sensitivity to others.20 No method of combining scores to estimate overall degree of empathy/reactivity is described for the IRI. The MBI-HSS is a 22-item rating scale designed to assess three aspects of the burnout syndrome: emotional exhaustion, depersonalization, and lack of personal accomplishment.21 Burnout was defined as summed scores in the upper tertile of normed scores on only the emotional exhaustion (> 26) and depersonalization (>12) subscales.7
We conducted statistical analyses using STATA Version 7 (StataCorp LP, College Station, Texas). Standard univariate statistics were used to characterize the sample. Beginning to year end data were compared using paired Student's t-test. In addition prevalences of chronic sleep deprivation, acute sleep deprivation, subjective sleepiness, and moderate depression were calculated based on validated cut points. Subscales on the IRI were compared with available data for adult and college student samples.8 Norms based on responses of other professional populations would have been preferable, but these data were not available. We used chi-square and Fisher exact tests to evaluate associations between chronic sleep deprivation and mood, empathy or burnout from the beginning to the end of internship year, focusing on changes over internship rather than specific changes in these variables over time.
In June 2002, 58 of 59 incoming first-year residents provided baseline data. In 2003, 47 (80%) also completed the year-end instrument. Of the original 58 residents, 30 (51%) were women; 19 (32%) were married, and four (7%) had children. Demographics of the 47 residents who completed the year-end instrument were not significantly different from the entire cohort (p values > .05). Table 1 shows changes in mean scores for sleep amounts, sleepiness, depression, interpersonal reactivity, and burnout from the beginning to end of the academic year. Except for reported sleep quantities in the prior 24 hours (p = .76), there were significant changes in mean scores for sleep quantities in the prior seven days, sleepiness, and nearly all mood measures. Additionally, scores that were originally higher (i.e., more favorable) than general population norms (p < .001) approached norms at the end of the year for empathic concern (p = .15) and perspective taking (p = .069). Fantasy scores started out the same (p = .53) but were lower by the end of the year (p = .001). The personal distress level of the respondents was significantly lower than population norms at both the beginning (p = .001) and end of the year (p = .002).
Prevalence of “high” scores changed for chronic sleep deprivation (9% to 43%, p = .0001) and sleepiness (11% to 36%, p = .0036). (see Figure 1) While the prevalence of moderate depression increased from 4.3% to 29.8% (p = .0002), no one had scores indicating severe depression. Overall, 4.3% reported a high level of burnout at baseline compared with 55.3% at year end (p < .0001). Specifically, prevalence of elevated depersonalization subscale scores increased from 12.8% to 68% (p < .001) and elevated emotional exhaustion subscale scores increased from 8.5% to 68% (p < .001). There was an absolute negative change in the mean scores for personal accomplishments over the course of the year. (p < .001). However, this did not reflect a change in the prevalence of a sense of high personal accomplishment in the intern class; all 47 respondents scored in the extreme highest third of norms (range 0 to 31) both at the beginning and end of the year.
If one was not chronically sleep-deprived prior to internship, there was an association between becoming chronically sleep-deprived and becoming depressed (OR = 7, p = .014, Fisher exact). We found no other associations between the development of chronic sleep deprivation and any of the interpersonal reactivity or burnout subscales.
Residency training, particularly internship, is generally acknowledged to be stressful. Reasons for supranormal levels of stress may include long work hours, sleep deprivation, challenging workloads, and limited time for personal pursuits.10,11 Although several researchers have linked these stressors to the high prevalence of depression,8,11 decreased empathy,8 and increased burnout,7 to our knowledge no prior work has looked at the relationships among sleep deprivation, sleepiness, depression, burnout, and mood as each changes over the course of internship year.
Overall, our study reaffirmed findings from prior work on the changes in mood and empathy8 and prevalence of sleep deprivation and sleepiness.6 It adds to existing literature by looking at covariations in scores over time, and provides data to support prior notions that sleep deprivation is related to depression.22 Interns arrived well-rested and alert, neither depressed nor burned out, and, not surprisingly given the pool of bright and ambitious individuals who choose medicine as a career, scored much higher than did the general population in terms of empathy and perspective taking. However, by the end of internship year, sleep quantities were dramatically reduced and chronic sleep deprivation was present in epidemic proportions. This is a troublesome finding given the contention that performance is impaired and errors occur more frequently following acute and chronic sleep loss than when participants are well-rested.2 Also of concern is the marked increase in the prevalence of moderate depression and burnout by the end of internship year. A recent review of burnout in residency training found four studies that refuted a relationship between burnout and sleep deprivation and two that described a relationship.23 The latter association was noted in studies examining work hours, not sleep time. To our knowledge, our study is the first to describe a relationship between burnout and sleep amounts. Finally, our study demonstrated a strong association between the development of chronic sleep deprivation and the development of moderate depression in those who were neither chronically sleep deprived nor moderately depressed at the start of internship. Moreover, multiple indicators suggest interns are experiencing a switch from a feeling of engagement with their work to a high level of burnout.
Our findings mirror the well-documented pattern that subjective sleepiness plateaus and may decrease even as objective sleepiness increases16 and performance declines including reaction times,24 psychomotor vigilance testing,3 and short-term memory impairment.25
Several conceptual models can be used to explain the link between sleep deprivation and depression. One of the traditional behavioral models describes how learned helplessness15 can lead to depressed moods. Interns are repeatedly placed in undesirable situations in which they have no control over their time and sleep amounts. Even when they have the opportunity to sleep on call or maximize recovery sleep, reversal of state is very rare. Additionally, sleep loss could lead to depressive disorders as proposed by the contrasting, empirically-based hierarchical model described by Parker13 in which fatigue and irritability can lead to depression. Alternatively, the association between sleep deprivation and depression may be explained by Patten's conceptual integrative model14 in which mathematical modeling details the possibility of conversion from an adaptive biological response pattern to psychosocial stress because of an accumulation of stressful circumstances and experiences. In sum, depressive disorders are heterogeneous entities and no one mechanism may be suitable to explain all the manifestations.
Our study had some limitations. First, a single-institution study of a single residency program limits generalizability. However, given the generic nature of medicine internships, our residency program is likely similar to many medicine internships across the United States. Second, global estimates of sleep duration are usually less reliable than are daily diaries or the combination of daily diary and actigraphy. Third, testing participants immediately before the start of internship may have actually produced different results than if they had been sampled a few weeks prior. The effects of relocation and anticipation of new job responsibilities may produce higher than average stress and sleep disturbances, which would actually serve to minimize the differences noted from the beginning of the year to the year end data. Lastly, the prevalence of depression was not high enough in the beginning of the year data to assess the effects of sleep deprivation on those who already had established depression.
Despite these limitations, our study raises several concerns. First, the prevalence of chronic sleep deprivation, sometimes without paralleled subjective sleepiness, is particularly problematic in a group of interns with critical job performance requirements at all hours of the day and night. Secondly, the prevalence of moderate depression in our study should remind program directors and medical educators to remain sensitive to the signs and symptoms of depression in house officers. Moreover, the strong association between the development of sleep deprivation and depression suggests job stressors and demands on time may relate to mood alterations. Additionally, the changes in burnout and empathy seen in this group are concerning as issues of professionalism come to the forefront of medical education. The better-than-average scores on multiple attributes erode within the first few months of internship. It remains to be seen if duty-hour reform will translate into more hours slept and thus begin to address the problem of chronic sleep deprivation or lead only to fewer hours worked. Overall, our findings strongly support the need for a complete assessment of sleep amounts among medical trainees under the new work hour limitations. Such ongoing assessment will be important to ensure the professionalism and health of young physicians, the safety of their patients, and the overall integrity of the medical education system.
2 Veasey S, Rosen R, Barzansky B, Rosen I, Owens J. Sleep loss and fatigue in residency training: a reappraisal. JAMA. 2002;228:1116–24.
3 Dinges DF, Powell JW, Pack AI, et al. Cumulative sleepiness, mood disturbance, and psychomotor vigilance performance decrements during a week of sleep restricted to 4-5 hours per night. Sleep. 1997;20:267–77.
4 Lockley SW, Cronin JW, Evans EE, et al. Effect of reducing interns' weekly work hours on sleep and attentional failures. N Engl J Med. 2004;351:1829–37.
5 Landrigan CP, Rothschild JM, Cronin JW, et al. Effect of reducing interns' work hours on serious medical errors in intensive care units. N Engl J Med. 2004;351:1838–48.
6 Rosen IM, Bellini LM, Shea JA. Sleep behaviors and attitudes among internal medicine housestaff in a U.S. university-based residency program. Acad Med. 2004;79:407–16.
7 Shanafelt TD, Bradley KA, Wipf JE, Back AL. Burnout and self-reported patient care in an internal medicine residency program. Ann Intern Med. 2002;136:358–67.
8 Bellini LM, Baime M, Shea JA. Variation in mood and empathy during internship. JAMA. 2002;287:3143–46.
9 Bellini LM, Shea JA. Mood change and empathy decline persist during three years of internal medicine training. Acad Med. 2005;80:164–67.
10 Shapiro SL, Shapiro DE, Schwartz GE. Stress management in medical education: a review of the literature. Acad Med. 2000;75:748–59.
12 Bonnett MH. Sleep deprivation. In: Kryger MH, Roth T, Dement WC (eds). Principles and Practice of Sleep Medicine. 3rd ed. Philadelphia: WB Saunders, 2000:53–71.
13 Parker G. Classifying depression: should paradigms lost be regained? Am J Psychiatry. 2000;157:1195–203.
14 Patten SB. Depressive symptoms and disorders, levels of functioning and psychosocial stress: an integrative hypothesis. Med Hypotheses. 1999;53:210–16.
15 Akiskal HS, McKinney WT Jr. Overview of recent research in depression: integration of ten conceptual models into a comprehensive clinical frame. Arch Gen Psychiatry. 1975;32:285–305.
16 Carskadon M, Dement WC. Cumulative effects of sleep restriction on daytime sleepiness. Psychophysiology. 1981;18:107–13.
17 Johns MW. A new method for measuring daytime sleepiness: the Epworth Sleepiness Scale. Sleep. 1991;14:540–45.
18 Beck AT, Rial WY, Rickels K. Short form of depression inventory: cross-validation. Psychol Rep. 1974;34:1184–86.
19 Beck A, Ward CH, Medelson M, et al. An inventory for measuring depression. Arch Gen Psychiatry. 1961;4:561–71.
20 Davis M. Measuring individual differences in empathy: evidence for a multidimensional approach. J Pers Soc Psychol. 1983;44:113–26.
21 Maslach C, Jackson SE, Leiter MP. Maslach Burnout Inventory Manual. 3rd ed. Palo Alto, CA: Consulting Psychologists Press, 1996.
22 Friedman RC, Kornfeld DS, Bigger JT. Psychological problems associated with sleep deprivation in interns. J Med Educ. 1973;48:436–41.
23 Thomas NK. Resident Burnout. JAMA. 2004;292:2880–89.
24 Doran SM, Van Dongen HP, Dinges DF. Sustained attention performance during sleep deprivation: evidence of state instability. Arch Ital Biol. 2001;139:253–67.
25 Williams HL, Gieseking CF, Lubin A. Some effects of sleep loss on memory. Percept Mot Skills. 1966;23:1287–93.