Medical Students and Stress Reduction Techniques:
An Analysis of Two Experiments
By Jennifer L. May
Psychology of Stress and Coping Dr. Keefe
Why Study Medical Students and Stress Reduction Techniques?
Throughout the semester this course has focused on the negative effects of stress on health and strategies to help alleviate stress. I found myself particularly interested in examining stress and possible reduction strategies for medical students. Because of their competitive and rigorous academic nature, medical schools in general are notorious for the amount of stress that they create for their students. The ever-constant pressures of medical school make it an ideal setting for the effectiveness of stress-reduction techniques. And as Whitehouse et al. (1996) point out, "Academic examination stress has provided a particularly useful research paradigm because it permits the prospective investigation of the impact upon psychological and physiological functioning of a commonplace, predictable and relatively circumscribed stressor."
While the sheer amount of stress encountered by medical students is reason enough to justify including them in a study on stress management, medical students are also a particularly useful group to study for other reasons. Shapiro et al. (1998) suggest that the nature of the patient-doctor relationship can have significant effects on the patient's health. Because doctors are an integral part of the health care system, a general improvement in their ability to manage stress as well as their ability to empathize with patients could tremendously improve the state of our health care system - potentially benefiting many more lives than simply the doctors themselves. While medical students are not yet doctors, they are in the process of learning how to be doctors. During their time in medical school, they are practicing the skills needed for their fields. If the stress management techniques studied are indeed effective in lowering stress, increasing empathy and awareness, and improving the health of medical students, then such studies could potentially have ramifications for the health care system as a whole. Yet before looking at the effects of stress reduction for doctors on the health care system, one must first examine if stress reduction techniques effectively lower the distress of doctors - or their prototypes, medical students.
Study 1 (Shapiro, Shauna, Schwartz, and Bonner) Spiritual Healers
Study 1 was an experiment that included an intervention-treatment group and a waitlist-control group. Both groups included premedical and medical students, and the treatment group participated in an 8-week mindful meditation-based stress reduction program. Through teaching students meditation based stress reduction intervention, experimenters sought to "(1) decrease overall psychological symptomatology… [including] measures of anxiety and depression; (2) reduce both state and trait anxiety…; (3) cultivate empathy and mindful listening skills…; and (4) contribute to an increase in spiritual experience/feelings…" (Shapiro, Shauna, Schwartz, Gary, & Bonner, 1998).
Overall, the design of Study 1 was good. Only students who agreed to participate in either the treatment or in the wait-listed control group were included in the randomization process. If the experimenters had allowed students to participate in the intervention group who were not willing to participate in the control group, then the intervention and control groups would differ across the third variable of willingness to participate in the control group, which in turn could potentially confound the study. Both the treatment and the control groups included equal proportions students in the categories of gender, race, and medical versus premedical status. This randomization with constraints controls for potentially confounding variables, strengthening the internal validity of the experiment. However, the experimenters do not say why they chose to use those particular variables as constraints. While it is not logistically possible to use every variably as a constraint (and this limits external validity), other variables were not looked at like socioeconomic background, religious background, region of origin, etc. I understand that by letting those variables vary randomly (rather than systematically) the authors are, in a sense, controlling for them. However, I would just like to know their rationale in deciding which variables they would use as constraints.
Many areas of strength in Study 1 are only moderate areas of strength, because the experimenters were too simplistic in the nature of their study. For example, their experiment was strengthened both by their decision to use different facilitators to "determine generality across experiments" as well as by their decision to take measurements both before and after the intervention in order to gage its effectiveness. However, their claim that they "rigorously scrutinize the benefits of the intervention" seems overstated. In fact, they only used two facilitators rather than multiple ones, and they did not establish baseline measurements for their participants prior to the stressful exam time-period in which both pre- and post- intervention measures were taken. Furthermore, if they really wanted to "rigorously scrutinize" such benefits, they could have take physiological measures as well as psychological ones.
A few areas of this experiment are in need of clarity. It is unclear whether the use of confidential identification numbers in the experiment protected only the identity of the participant or the participant's group (intervention or control) from the researcher as well. Also, if the possibility exists for "bias [to be] induced by the meditative state of the class" between pre- and post-intervention measures, how did the researchers decide on the length of the break between the intervention and post-measures? And how does this possible bias of a meditative state that they are trying to control for differ from the intervention itself? It is also unclear whether the daily diaries were used only to track meditation outside of the program or if they were also used for emotional disclosure as well. Perhaps the most unclear aspect of the experiment is why the experimenters chose to include a spirituality measure if they were unwilling to offer interpretations of what the results from that measure might mean.
Study 1, however, does have areas of great strength. The measure of empathy is a new area of research according to Shapiro et al (1998), and one that they clearly argue is worth studying. Also their hierarchical variable analysis resulting in their path diagram offers empirical data to support the causal theory that they posited. Overall, their experiment is insightful and adds to the general body of knowledge surrounding the study of stress reduction techniques.
Study 2 (Whitehouse et al.)
Study 2 was an experiment conducted over 19 weeks that included a hypnosis/relaxation intervention group and a control group. 21 subjects were randomly selected to participate in the intervention group and 14 were part of the control group. The intervention group received training in "the use of self-hypnosis as a coping skill" and utilized daily diaries, while the control group only kept daily diaries with no training in stress-reduction. For both groups, "[s]elf-report psychosocial and symptom measure, as well as blood draws, were obtained at four time points: orientation, late semester, examination period, and postsemester recovery" (Whitehouse, et al., 1996).
This experiment was quite thorough. By taking both psychosocial as well as physiological measures and examining both in detail, the experimenters offer an in-depth, soundly scientific study design for the alternative medical technique of hypnosis. By taking measurements at four different points in time throughout the semester, the researchers were able to examine how the participants' reaction to stress (both psychologically as well as physiologically) differed over time. Not only did the researchers take these measurements at four points of time, but they offer clear rationale as to why they chose such points in time. Orientation and examination times are moments of acute stress; late semester is a moment of weaker but prolonged stress; and postsemester is a time of relaxation. Other areas of strength for the study included a series of complex and well thought-out statistical analyses as well as within subject and between subject analyses. Perhaps the most striking strength of this study was the fact that it offered a behavioral placebo by giving the control group a behavioral intervention believed to be ineffective without the training in hypnosis and relaxation.
This behavioral placebo, however, is a complication if writing in diary is in and of itself a method of stress reduction. It does not, however, confound the experiment because both groups maintain a daily diary. A potential weakness of the study is that its participants were not blind. "All volunteers were informed at the outset of the study that they might be randomly selected to receive self-hypnosis training as a means of coping with stress" (Whitehouse et al., 1996). While this is not necessarily problematic in the intervention group (because awareness is part of the intervention), it seems rather problematic in the control group. If the participants were aware that they were in the control group, then perhaps their measurements were affected by their desire to please the experimenters. However, it is unclear whether or not the experimenters were required to inform all participants about the possibility of hypnosis (perhaps by an institutional review board.) If they were not required to disclose such information, their experiment would be stronger by at least the control group being blind.
Overall, however, this experiment is very strong. It provides the scientific community a controlled, meticulously analyzed set of data for a field that has a scarcity of scientifically sound empirical data. It has taken into account a multitude of variables by measuring and analyzing them. But perhaps one of the most striking strengths of this paper is that it had yielded results; it is an empirical study with results that indicate a positive relationship between hypnosis/relaxation and the reduction of stress in psychosocial measures.
One of the most interesting differences between Study 1 and Study 2 is the use of placebo or lack thereof. Study 1 does not use a placebo, a potential weakness for the study. Perhaps the stress program itself is simply a placebo. Whether or not this is a weakness of the study depends on the aim of the study. If the aim of the study is to isolate a variable that effectively works to reduce stress, then not controlling for a placebo effect weakens the study. However, if the aim of the study is simply to find something that works to reduce stress, then regardless of how the meditation techniques in Study 1 are functioning, the fact that they are functioning (placebo or not) means that study is successful.
Unlike Study 1, Study 2 uses a placebo, but the argument regarding the use of placebo becomes even more complicated than in Study 1. According to Dr. Keefe, it is incredibly difficult to design a "good" behavioral placebo. In assessing whether or not the behavioral placebo in Study 2 is good, one must again determine what the aim of the study is. If the aim of the study is to isolate the variable that reduces stress, ironically, the use of their placebo in this study weakens it, because the placebo decreases the researcher's ability to show statistical significance for hypnosis/relaxation. "For example, the completion of daily diaries, providing a potential medium for disclosing emotionally upsetting experiences, could have health promoting effects" (Whitehead et al., 1996). Perhaps this is a "bad" placebo design, because writing in a diary may not be a null treatment. The authors suggest, "Considerations such as these may require greater attention to choice of methodology… in future research on interventions designed to mitigate the adverse of stress on psychological and immune functioning." However, it is unclear in this particular experiment whether writing in a diary helped people simply because they believed they were taking part in the true intervention or if the act of writing in a diary, apart from belief, did in fact improve the health of the control participants. If the aim of this study, however, is to find an effective behavioral intervention for coping with stress, then perhaps Whitehouse and his research team designed a "good" placebo.
The issues raised by placebos highlight the difficulty in isolating a treatment from a participant's belief in it. This raises the question: is a "true" treatment (as opposed to a placebo) effective even if the participant doesn't believe in its effectiveness? And if it is not effective unless the participant believes in it, doesn't that make it a placebo? Let's look at the context for such questions. Study 2 is an investigation of the extent to which typical academic stressors "can be behaviorally mitigated by use of cognitive strategies" (Whitehouse, et al, 1996). If the intervention depends on the use of cognitive strategies, then it seems that a participant's beliefs are an integral part of the intervention strategy itself and thus cannot be isolated from the strategy. This renders the preceding question extremely difficult to answer. Suffice it to say that such difficult questions demonstrate a need for further research regarding placebos, as well as a need to think about the subjective nature of placebos - a subjectivity that is difficult to control for in an objective experiment.
As long as I am on the topic of the role of subjectivity in "objective" experiments, I would like to briefly mention other ways in which these subjective judgements play a role in these experiments. Throughout both of these experiments, many subjective decisions were made. The choice of measurement scales, the types of measurement (psychological or physiological), the duration of interventions (or breaks between interventions and measurements) were all decided on based on the researchers' best judgement. In fact, in assessing the psychological health of participants, researchers rely on definitions of health that are subjectively voted on by the American Psychological Association. Furthermore, in deciding whether or not results are "significant," studies pick a cutoff such as p< 5% that has arbitrarily been set as a margin of error that the scientific community has generally agreed is an acceptable margin of error. Obviously, this significance cutoff can be set larger or smaller, but the point is that it is in fact a subjective cutoff point.
Debate regarding the use of placebos and regarding the role of subjective belief in experiments is a debate which scientists must authentically address. While I certainly cannot claim to have the answers to these questions that challenge scientists, I do believe that it is valid to raise such questions. Questions that lack clear answers provide new fields of inquiry and study for scientists.
The two studies examined in this paper are brave studies, in the sense that they are examining fields that American culture has deemed "unscientific" through scientific experimentation. These studies both provide evidence for the efficacy of particular "alternative" intervention programs in stress management. And both studies are important because of the larger issues they raise regarding health and stress, placebos and treatment, control and subjective belief.
Special thanks to
CD data recovery software