Despite the growing interest in the prevalence and consequences of loneliness, the way it is measured still raises a number of questions. In particular, few studies have directly compared the psychometric properties of very short measures of loneliness to standard measures.

We conducted a large epidemiological study of midwife students (n = 1742) and performed a head-to-head comparison of the psychometric properties of the standard (20 items) and short version (3 items) of the UCLA Loneliness Scales (UCLA-LS). All participants completed the UCLA-LS-20, UCLA-LS-3, as well as other measures of mental health, including anxiety and depression.

First, as predicted, we found that the two loneliness scales were strongly associated with each other. Second, when using the dimensional scores of the scales, we showed that the internal reliability, convergent-, discriminant-, and known-groups validities were high and of similar magnitude between the UCLA-LS-20 and the UCLA-LS-3. Third, when the scales were dichotomized, the results were more mixed. The sensitivity and/or specificity of the UCLA-LS-3 against the UCLA-LS-20 were systematically below acceptable thresholds, regardless of the dichotomizing process used. In addition, the prevalence of loneliness was strikingly variable as a function of the cut-offs used.

Overall, we showed that the UCLA-LS-3 provided an adequate dimensional measure of loneliness that is very similar to the UCLA-LS-20. On the other hand, we were able to highlight more marked differences between the scales when their scores were dichotomized, which has important consequences for studies estimating, for example, the prevalence of loneliness.

Applying the concept of burnout to medical students before residency is relatively recent. Its estimated prevalence varies significantly between studies. Our objective was to estimate the prevalence of burnout in medical students worldwide.

We systematically searched Medline for English-language articles published between January 1, 2010 and December 31, 2017. We selected all the original studies about the prevalence of burnout in medical students before residency, using validated questionnaires for burnout. Statistical analyses were conducted using the OpenMetaAnalyst software.

Prevalence of current burnout was extracted from 24 studies encompassing 17,431 medical students. Among them, 8060 suffered from burnout and we estimated the prevalence to be 44.2% [33.4%–55.0%]. The information about the prevalence of each subset of burnout dimensions was given in nine studies including 7588 students. Current prevalence was estimated to be 40.8% for ‘emotional exhaustion’ [32.8%–48.9%], 35.1% [27.2%–43.0%] for ‘depersonalization’ and 27.4% [20.5%–34.3%] for ‘personal accomplishment’. There is no significant gender difference in burnout. The prevalence of burnout is slightly different across countries with a higher prevalence in Oceania and the Middle East than in other continents.

The results of this meta-analysis suggest that one student out of two is suffering from burnout, even before residency. Again, our findings highlight the high level of distress in the medical population. These results should encourage the development of preventive strategies.