The main purpose of this study was to examine the relationship between life satisfaction among medical students and a basic model of personality, stress and coping [15,16]. Personality traits, such as neuroticism and extroversion, have been found to be predictors of life satisfaction [17]. As indicated above, research on the psychological adaptation of medical students has focused on distress. Perceived medical school stress has been linked to current mental distress [18] and to forthcoming mental health problems [19], and is therefore assumed to affect life satisfaction. According to the resilience model, the way that students cope with stress factors may influence their mental health in medical school [15,20]. We anticipated that active or problem focused coping and seeking social support would promote higher satisfaction with life, whereas passive or emotion focused coping would have a negative impact on life satisfaction. In particular, we wanted to study the characteristics of students who sustained high levels of life satisfaction during medical school, and compare them with the characteristics of their peers to find factors of resilience that may be used to make positive changes, and hence help future medical students [21]. We assumed that stress factors pertaining to medical school might have a negative impact on students' life satisfaction, and wanted to examine if this effect can be counteracted with efficient coping strategies. We also anticipated that particular personality traits would affect students' susceptibility to the educational challenges.

A recent study from our group [22] followed a cohort of Norwegian physicians from their final term in medical school to the ninth postgraduate year and found lower levels of life satisfaction in the fourth and ninth postgraduate years compared to comparable population samples. The study also found that the level of life satisfaction was even lower at the end of medical school, but it did not include data from the undergraduate years. There are a number of studies which have compared levels of distress among medical students with that of other students [23-25], but to our knowledge there has been no comparative study focussed on positive mental health.

How do we explain the decrease in life satisfaction during the first years in medical school? We do not know the level of life satisfaction among the students before they embarked on their studies of medicine. It is possible that they reported higher levels of satisfaction than they normally would due to the fact that they had recently been accepted to medical school, which has Norway's highest admission requirements and that they simply returned to their normal level of satisfaction by the third year in medical school. However, the fact that medical students were as satisfied as other students when entering medical school, but less satisfied at graduation, contradicts this explanation, and indicates that the reduction in life satisfaction may have been caused by factors specific to medical school. It may be that the medical students did not adapt to their studies as well as other students. Although some studies have found that pharmacy students [23], graduate science students, [24,25] and law students [32] experience even more distress than medical students, the presumption that medical school has an impact on medical students' mental health is supported by several studies, which have shown that the medical students' well-being deteriorates during medical school. A longitudinal study of depression among medical students found that upon entering medical school the students' emotional status resembled that of the general population, but that depression scores rose and remained high during medical school [33]. This is consistent with findings of lower levels of life satisfaction among Canadian medical residents than among the general population [27]. Similar results have also been found in our group when comparing Norwegian medical graduates and physicians to comparable population samples in Norway [22]. Although the mean levels of life satisfaction among the medical students were quite high throughout medical school, our study supports the notion of the medical education and career as having an unfavourable effect on general life satisfaction among students and physicians. According to two review articles by Dyrbye et al, the impact of this effect has not been thoroughly studied, but some results indicate that student distress may correlate with impaired academic performance, cynicism, academic dishonesty, substance abuse and suicide [8,34].

The main aim of this study was to identify resilient medical students and find out if these students differed from their peers in personality, perceived stress, and coping strategies. We succeeded in finding a subgroup of students with stable high levels of life satisfaction. Compared to the group with fluctuating levels of life satisfaction, the stable students differed with respect to both perceived stress and coping abilities.

Our study found that low scores on the PMSS – social and personal renunciations item predicted stable high life satisfaction, even when susceptible personality traits were controlled for. This indicates that students in the stable group perceived medical school as interfering far less with their social and personal life than did students with fluctuating levels of life satisfaction. The ability to find enough energy to spend on several of life's domains, while maintaining a balance between them may be crucial for experiencing stable high life satisfaction. This finding is supported by previous studies which have shown that socializing decreased in medical school [35], that inadequate social activity was linked to impaired psychological health among medical students [36], and that leisure activities can reduce stress in medical school [37]. Another study revealed that many medical students felt guilty for spending time on social activities and personal well-being, although they recognised the importance of doing so [13]. Medical school should encourage students to maintain their outside interests and leisure activities and to make time for friends and recreation.

In the univariate analysis, low levels of academic worries corresponded to stable high life satisfaction. Although this effect did not remain statistically significant in the adjusted model, this finding may still indicate that perceived cognitive abilities are important for life satisfaction among medical students.

As expected, we found that a low level of passive, emotion focused coping, such as wishful thinking, was associated with stable high life satisfaction. Other studies support the harmful effect of emotion focused coping among medical students [12,38]. Studies of another cohort of Norwegian physicians showed that this form of coping during medical school predicts forthcoming mental health problems [19]. Although some of these studies found active coping to promote well-being [11] and to protect against depression [12], these factors reached predictive power in the unadjusted analyses of our study, but not in the adjusted model. Coping strategies can be modified by educational and therapeutic interventions. Training in how to use healthy ways of coping should therefore be provided in medical school, for instance, through the implementation of stress management courses.

We found that personality traits did not predict stable high life satisfaction when we controlled for other coping variables. A low level of vulnerability predicted stable high levels of life satisfaction in the univariate analysis, but this effect seemed to be channelled through wishful thinking in the adjusted model. The role of personality and dispositional variables has been emphasised in several other life satisfaction studies [10,17]. However, another study from our research group found only a rather modest effect of personality on life satisfaction among medical postgraduates and physicians [22]. It seems that for people in this educational and occupational group, stress and coping may be more important for life satisfaction than personality.

While other studies have found that female medical students report more stress [4] and less satisfaction with life than male medical students [39], we found no gender differences in life satisfaction in our study. This is in accordance with earlier reported findings on mental health among Norwegian physicians [19,40] and may reflect the relatively equal position of the genders in Norway.

The major strength of this study is that it is a longitudinal study, following a nationwide cohort of medical students throughout their entire study period. Medical education in Norway consists of six years of education at the university, followed by one and a half years of compulsory residency. Although this is different from, for example the United States, our findings may be generalised beyond Norway, at least within the European system of training. High levels of stress are reported in several countries [1,3,4,12,36,37]. The response rates in the cross-sectional samples of our study were quite high, and although the response rate was somewhat lower in the longitudinal sample, it still consisted of more than half of all the students who entered medical school in Norway in 1993. Considering that those students who dropped out of the longitudinal sample after having responded the first year reported a lower level of life satisfaction when entering medical school than those who participated in all three assessments, the reduction in life satisfaction might have been even larger if all students had been part of the longitudinal sample (type II error). The data on the control groups were collected midway through the study. The comparisons made between the students in their first and last years are hence done with data of other students collected three years later and three years earlier, respectively. We consider any possible effect of this difference in time to be minimal. Another limitation of the study is the employment of a single item as an outcome measure, which may reduce the reliability of the responses, but we consider its correlation with a validated subjective well-being scale [29] to be satisfactory.

What can be done to increase resilience in medical students and ensure that future students are enabled to maintain stable, high levels of life satisfaction throughout medical school? The results of this study stress the importance of making time for personal and social life while in medical school, and the advantage of avoiding the use of passive, emotion focused coping, such as wishful thinking.

Findings from another study, which revealed that many medical students felt guilty for spending time on social activities and personal well-being even though they recognised the importance of doing so [13], may indicate that a change of social norms in medical school is necessary. Medical school educators could play an important role in such a change, by being good role models. In addition, reducing peer pressure is probably also an efficient means for reaching this goal. Evaluation of such interventions may be an important area for future research.