Global trends in high impact psychiatry research

The utility of psychiatry research, when indexed as the number of times it is cited by other researchers, is most centrally predicted by the impact factor of the publishing journal1. In keeping, scientists' track record in high impact journals is directly related to academic promotions and merit increases in academic psychiatry institutions2, and the acquisition of research funding3. Thus, much pressure exists for psychiatric researchers to publish their research in high impact psychiatry journals.

Trends in high impact psychiatry journals are therefore important to examine. This is especially true when considering that psychiatry is thought to be particularly vulnerable to publication bias. While studies have noted possible biases in the preponderance of underpowered studies stemming from pharmacological company sponsorship4, and the underrepresentation of research stemming from low‐ and middle‐income countries5, less evidence exists in detailing possible bias relating to the content of psychiatric research. We therefore assessed global publication trends in high impact general psychiatry journals according to specific illness typology and methodology of research. In total, we extracted 12,218 manuscripts from six high impact psychiatry journals, over a 16‐year period (2000‐2016).

In terms of psychopathology, stable trends emerged relating to comorbid psychiatric presentations, mood disorders and psychotic disorders, which remained the most studied psychopathologies throughout this period. Equally robust trends emerged around the least commonly studied psychopathologies – personality, eating and obsessive‐compulsive disorders. The coverage of two of them (eating and personality disorders) consistently decreased over time.

In terms of manuscript typology, less stable trends emerged and, although epidemiological and behavioral studies were the most prevalent type across the entire period, they appeared to be steadily decreasing in prevalence over time. By contrast, editorial manuscripts almost tripled in prevalence throughout this 16‐year period and, by 2016, these manuscripts made up the most common manuscript type in high impact psychiatry journals.

In interpreting these findings, several possibilities are worth considering. First, while data relating to the ratio of submitted versus accepted manuscripts are unavailable, one possibility is that more manuscripts relating to comorbid presentations, mood disorders and psychotic disorders are submitted to high impact psychiatry journals, relative to other psychiatric disorders, which may account for their greater overall prevalence.

Relatedly, evidence from the US National Institute of Mental Health suggests that research pertaining to mood and psychotic disorders consistently receives more funding than other psychopathologies6. Greater financial support may allow for a greater number of large‐scale high‐quality studies in the domain of mood and psychotic disorders, relative to other disorders, which may account for the greater overall volume of these studies in high impact psychiatry journals.

Similarly, the decreasing prevalence of epidemiological and behaviorally‐focused studies in high impact psychiatry journals may reflect a declining submission rate. This would be consistent with the ongoing paradigm shift which posits that innovative research ought to uncover the precise pathophysiology of psychiatric disorders, rather than documenting behavioral sequalae or epidemiological trends7. However, it is important to note that the presence of genetic, cellular or neuroimaging studies in high impact psychiatry journals is not increasing.

A second possibility in accounting for these trends, in the event of comparable submission rates, is a potential greater acceptance rate of manuscripts relating to specific psychopathologies and methodologies. In this instance, it would be important to consider what drives a higher acceptance rate, and therefore a greater emphasis on scientific discovery, among some specific psychopathologies relative to others. For example, in considering nationally representative data relating to the prevalence of psychiatric disorders in the US, evidence suggests that anxiety disorders are the most common psychiatric disorders, yet they represent only 4‐7% of manuscripts in high impact psychiatry journals, and their coverage is steadily declining over time. In contrast, psychotic disorders feature prominently in high impact psychiatry journals, representing 20‐27% of all manuscripts, yet demonstrate a lifetime prevalence of less than 1%.

Burden of illness data suggest that mood disorders are the leading cause of disability‐adjusted life years (DALYs) stemming from psychiatric illness8, which is consistent with the prevalence of manuscripts relating to these disorders in high impact psychiatry journals. However, other major contributors to DALYs stemming from psychiatric illness include anxiety disorders and substance use disorders, which represent a notably more marginal focus in high impact psychiatry journals.

The rapid increase in editorial manuscripts over this period is particularly noteworthy, and it is unclear whether this reflects a sharp increase in the overall volume of editorial submissions, relative to other types of methodologies, or a greater acceptance rate among editorial manuscripts. In fact, editorial manuscripts are unique in that they are typically solicited from editors, although it is unclear as to whether this alters the course of editorial and peer review. This sharp increase is important, however, in that editorials do not offer novel data, and more frequently offer novel perspectives or syntheses of emerging findings.

This may involve the advantage of assisting in the interpretation and contextualization of complex empirical findings, while simultaneously increasing the citation count of articles included in the journal (although not the impact factor, which is based on citations received by articles in the following two years), since editorials typically relate to manuscripts in the same journal edition.

In summary, these findings suggest a demonstrable skew in high impact psychiatry journals towards comorbid, mood and psychotic disorders, alongside editorial type manuscripts. These trends are important as they relate to the dissemination and uptake of psychiatric research, since potential bias may stymie advances in research relating to less commonly featured psychopathologies.

Our results suggest that trends in high impact psychiatry journals may be comparable to trends in psychiatric research funding, which is consistent with the notion that extramural research funding appears to be linked to researchers' track records in high impact journals3. The potentially recursive influence between lower extramural research funding and the lower likelihood of high impact publications may present an additional barrier for research innovation among psychopathologies which attract less extramural funding, which ultimately may result in slower advances for patients.