Scientific research, the quest for the unknown, should perhaps be considered the most exciting job there is, but data indicate that scientists are generally unhappy with their lives. For PhD students, this is particularly concerning.

PhD depression

According to recent studies, PhD students are two times more likely to suffer from mental health problems, and three-quarters of them are found to be under more than average stress. Further studies evidence a worrying prevalence of factors associated with depression in doctoral students. Simply put, PhD students are under severe stress and are disturbingly likely to suffer mental health problems.

These findings are not surprising for anyone in academia. In fact, it is widely acknowledged in science that PhD researchers are generally unhappy, often anxious, and depressed. 

Confirming the extent of the problem, worldwide rates of depression and anxiety are six times higher in PhD students than those in the general public. These results were published in the journal Nature Biotechnology, a publication from the prestigious scientific journal Nature. The authors conclude —  as reported in a short commentary by Nature itself on the 26th of March 2018 — that academic institutions should “provide students with training to help them manage their time and cope with stress”.

According to the authors of the study published in Nature Biotechnology and to the editors of the journal Nature, the reason for this worrying depression trend in PhD researchers has to be attributed to the lack of psychological training for students in academic settings. In other words, rather than looking at the cause of students’ mental health problems, Nature invites institutions to apply mitigation measures, to allow students to better cope with any sort of pressure they are subjected to during their PhD.

To consolidate and add to their point, on the 15^th of September 2019 Nature published an editorial article entitled “Being a PhD student shouldn’t be bad for your health”, where they suggested that “solutions are at hand”. Without any particular scientific evidence, they claim that “supervisors need comprehensive, compulsory training to identify, assist and understand researchers facing mental health problems. Students could have more than one supervisor, so that they can find support without worrying about damaging their career.” And also that “universities need to make sure that the mental-health services they admirably make available to undergraduates also reach graduate students and postdocs. And academia must learn to respect the work-life balance that many researchers struggle to find”.

That’s not nearly enough.

The mitigation measures proposed by Nature in their editorial article seem, to me and many other colleagues, clearly insufficient to tackle the concerning mental health crisis in academia. After all, the root of the problem isn’t addressed — and the root may be the journals themselves.

Publish or perish

In order to prevent scientists from condemning their hurried conclusions, Nature designed a survey to understand the causes of the unhappiness of PhD researchers. But before diving into that, let’s shed some light on the bizarre model of scientific publishing.

You spend months or years researching a particular topic (typically, there’s a whole team working). You write the manuscript, where it is reviewed by experts in your field, who typically ask for some changes or edits to be made. If everything goes according to plan, you revise the manuscript and it’s published in the journal. But here’s the thing: the authors don’t get paid, the reviewers don’t (usually) get paid, instead it’s only the journal that gets paid.

Notably, in academia, there is a growing movement of scientists trying to oppose journals, in particular high impact journals such as Nature, which are basically private businesses that have monopolized scientific communication. Although many recognize journals as one of the major issues in modern science, they feel helpless to oppose their power. But that hasn’t dissuaded some researchers from taking the journals head-on.

The battle against journals’ interests is being fought by some dedicated scientists who are not afraid to raise their voices but are often left alone on the front line. For instance, Nobel Prize Laureate Randy Schekman called Nature a “hegemonic” journal in a recent interview for Culturico. One of the main issues with branded journals such as Nature is that, in science, success is measured by the record of publications. You publish in journals, you get a ladder to help in your academic ambitions. You don’t? There’s no room for you in academia.

For this reason, Nature’s survey already constitutes a conflict of interest, especially as Nature itself — together with other prestigious journals and their policies — could indeed be one of the causes of the mental health crisis in academia.

At this point, it should be clear that it’s not Nature itself who should be conducting studies on the young researchers’ mental health struggles. Rather, a third party should be in charge of identifying the root causes for them. However, no scientific intergovernmental institution currently exists to regulate science, and the publication of scientific studies is substantially monopolized by journals. In other words, scientists have no other way to communicate their research findings effectively if not via journals themselves.

Nature gathered data from its survey in an article entitled “PhDs: the tortuous truth” and published it on the 13^th of November 2019. Nature asked students whether they are satisfied with their decision to pursue a PhD, whether their satisfaction rate improved or worsened during the course of their studies, but also whether they are stressed because of the number of working hours, because of the lack of support from their institution or because they are subjected to harassment from their supervisors or colleagues. I will return to these specific points very soon.

Unsurprisingly, all of these were identified as stressors that contribute to the mental health crisis in academia. It turns out, almost everything about the PhD environment is a stressor. However, this is where the conflict of interest steps in: there was no question and no mention of the role played by journals in this crisis. After all, who would challenge themselves so much to go against their own interests? The best way to preserve the status quo and avoid being questioned is to become the questioner. Nature attempted to remove itself from the dispute by becoming superpartes, the seemingly disinterested arbiter of the dispute.

In line with their previous articles, Nature concludes that “institutions […] have much to learn”, again blaming universities for their inaction. This time, they support their statements with data gathered in the survey, data that could be questioned as inconclusive, as questions were biased to confirm their conclusions published in the previously mentioned editorial article. 

Nature is far from the only journal dodging the blame and pointing at the lack of educational training for supervisors as the root cause of the mental health crisis in academia. Last year, Wang and colleagues published a study investigating the prevalence and associated factors of depression among doctoral students in the journal Psychology Research and Behavior Management. They conclude that “educational experts should pay close attention to the mental health of doctoral students. Active strategies and interventions that promote research self-efficacy and mentoring relationships might be beneficial in preventing or reducing depression and anxiety.”

Of course, there is much to be said about the role of universities and supervisors, but there is an underlying bias in these studies: they narrowly focus on one factor without analyzing its relative importance among other factors, diverting the focus and leading other scientists to misunderstand the causes and seriousness of the issue. For instance, Karen Berry, Emma Warnecke, and Megan Woods proposed in “The Conversation” that the solution to the problem is a psychological approach referred to as “mindfulness”, a method for managing feelings of stress and anxiety. As if playing psychological games could eradicate, or even superficially scratch the real problem.

But this is a bit like a company that overworks and underpays its employees and then offers yoga classes to help with the stress: it does little to address the underlying issue.

Different survey with different results

If you’ve followed me until now, you probably realise that I believe — as do many others in academia — that journals are the biggest contributors to the stress and anxiety that PhD researchers suffer from. It is about time to check whether this is indeed the case and to do it with a scientific and unbiased approach.

For this purpose, I designed a survey to identify the root causes of the mental health crisis in academia and asked PhD researchers to answer a few questions.

The survey, which I developed in collaboration with Culturico, can be found here. It is designed similar to Nature’s survey, and in fact the first few questions are virtually identical. This serves as a control, a check that my survey and Nature’s yield similar results. In order to reach unbiased conclusions, my questions touched upon all the topics discussed in Nature’s survey, including questions about whether a student feels stressed, whether they are satisfied with their decision to pursue a PhD, and how much each contributing factor causes stress (on a scale from 0 to 10) – i.e. the supervisor’s contribution to the stress level, the atmosphere in the research group, discrimination or harassment.

In addition, however, I added two other factors that weren’t taken into consideration in the survey designed by Nature. These were the fear of presenting in front of colleagues and other scientists (PhD students have to deal with public presentations of their research quite often) and the contribution to the desire of publishing in high impact journals on a student’s stress level. We will soon understand why the public communication of science and data sharing between scientists is a key factor in this survey.

The results of the survey

From the 175 respondents 86% confirmed they felt stressed during their PhD, thus confirming how worrying the mental health crisis in academia really is. 

However, the majority of respondents (73%) stated they are either very satisfied or somewhat satisfied with their decision to pursue a PhD, thus highlighting that PhD programs themselves are not the problem.

In other words, although unhappy and stressed, PhD students think they made the right decision in pursuing a PhD.

These results are in agreement with those of Nature, for which 75% of the interviewees responded they are either very satisfied or somewhat satisfied with their decision of pursuing a PhD. But the problem lies elsewhere.

As Nature did, I also asked if the respondent’s satisfaction level increased, worsened, or remained the same since the start of their PhD. In this case, although the results slightly differ from those acquired by Nature, the conclusion is the same. In my survey, 38% of the respondents declared their satisfaction level worsened (Nature: 45%), and 30% of responders answered that it increased instead (Nature: 42%). In my survey I had a larger number of students responding that their satisfaction level has remained the same (32%, Nature: 13%).

The PhD survey performed by Nature asked respondents to either agree, disagree or show neutrality for given statements. One example is: “My supervisor has a good awareness of support services and was able to direct me to them if needed”.

To this statement, 42% of the respondents disagreed, whereas 28% agreed (30% showed neutrality). This and other similar results prompted the most prestigious scientific journal to conclude, as they did in previous articles, that more should be done “to provide more training for supervisors”. Although we can all relate to this qualitative statement, this point may be only one of the factors to be concerned about, and may not be the predominant one. 

To understand whether this is indeed the case, I quantitatively measured the impact supervisors had on a student’s stress levels, as well as other factors considered in Nature’s survey, such as whether discrimination or harassment are important stressors. I compared these factors with others that weren’t taken in consideration by Nature, i.e. how much the fear of presenting in front of other scientists contributes to a student’s stress level, as well as the contribution of the atmosphere in the research group and of the desire of publishing in high impact journals (such as Nature).

As previously mentioned, while Nature only asked PhD students to either agree, disagree or show neutrality for given statements, I asked respondents to grade how much each factor contributes to their general stress level, from 0 (not at all) to 10 (the maximum), in order to quantify each stress factor allowing comparisons to be made between them.

In Nature’s survey, 21% of the respondents declared that they experienced discrimination or harassment in their PhD programme, and Nature concluded that universities are not doing enough to tackle this issue. Instead of asking the respondents if they experienced discrimination, I asked them how much this influences their stress level. My analysis revealed that this problem does not cause mental health issues for the vast majority of students. This was identified as the least important factor with an average score of 2.9 out of 10, with 42% of the respondents stating this is not a problem at all (score=0), and only 5% of the respondents declared this was in fact a huge element of stress. Although harassment and discrimination, whether based on gender, age, or disability should be condemned, it is highly likely that academia does not perform any worse than other sectors, and thus it is not the reason for the overwhelming mental health problems PhD researchers are facing.

On the contrary, and here I agree with Nature’s considerations, supervisors should be trained to educate and teach their students. This can strongly influence the general mood and atmosphere in a research group and ultimately contribute to improving the wellbeing of all scientists and the quality of science itself. My survey identified that supervisors contribute 5.4 out of 10 to a student’s stress level, and the research group atmosphere contributes 4.9 out of 10. Here respondents gave more disparate answers. Some feel that supervisors or their colleagues are the main source of stress, and some feel they don’t contribute at all. In other words, these are not necessarily systemic factors, but they are nonetheless important contributors to the stress level of young researchers. It depends strongly on one’s supervisor, and on individual experiences. I agree with Nature that supervisors should be trained, but I also think that Professors and Principal Investigators should be selected for their teaching capabilities rather than their record of publications in high impact journals. After all, this very problem is still connected to the role journals have in shaping working dynamics in science.

More or less in line with these data, the fear of presenting in front of audiences — something scientists are required to do on a regular basis — also appears to be a relevant contributing factor (5.3 out of 10).

As with the other two factors, the contribution of presenting to a student’s overall stress level is highly variable depending on the respondent, with some feeling more comfortable in presenting and others feeling very uncomfortable. This highlights the lack of presentation skills training for students, perhaps combined with the pressure of presenting high quality – publishable – findings, which again reminds us of the role journals play in placing PhDs under publication pressure. Possibly, in a scientific world without journals, PhD students may feel less anxious about presenting their data, without having to prove to their supervisors and colleagues that their results make up to a publishable story, a brilliant discovery, or a well-designed and attractive study.

Last, but definitely not least — in fact, this is the most important point — I asked respondents to score how much the desire of publishing in high impact journals contributes to their stress level.

This particular point wasn’t (perhaps obviously) considered in Nature’s survey. This factor scored 6.7 out of 10, the most relevant of the considered factors by a long way. Further, most respondents considered the desire of publishing in high impact journals as the largest stress contributor (39%). In this special ranking presenting came second (22%), the role played by the supervisor third (19%), followed by the atmosphere in the research group (13%), and last by having experienced discrimination or harassment (7%).

It won’t be difficult for an attentive reader to see that those factors that weren’t considered by Nature are actually the most relevant ones.

The pressure of publishing

Students feel the pressure of publishing in high impact journals, especially when they want to continue in their academic careers. In fact, positions in academic institutions are given based on the track record of publications, and particularly valuable are those publications in high impact journals, such as Nature for scientific disciplines. I explained elsewhere how the scientific publishing system works and its repercussions for science and scientists. 

However, in another editorial article published on the same day of the survey’s results (entitled “The mental health of PhD researchers demands urgent attention”), Nature envisions dangerous criticism, and they appear to try to prevent it by concluding that “it also lies in recognizing that mental ill-health is, at least in part, a consequence of an excessive focus on measuring performance – something that funders, academic institutions, journals and publishers must all take responsibility for”.

By recognizing itself to be playing a (minor) part of the problem, Nature stresses that they are at least doing something to address the problem. However, Nature’s strategy is nothing other than semantic satiation, the psychological phenomenon in which repeated words become meaningless for the listener. Nature’s articles and surveys are just aiming to shift the problem elsewhere, not to take the blame for it. The opinion that journals are to be blamed is floating around in scientists’ minds, but this doesn’t really matter as long as scientific evidence of this is not circulating online, and as long as journals play the role of institutions providing good services for public science.

The publish or perish culture in academia finds no platform, no online space for discussion other than the journals themselves, as I previously mentioned. Nature smartly provides the platform for this discussion, positioning itself as a moderator. Nature acts in practice as a sort of authoritarian government that appears as a democratic one. It publishes articles showing concerns about the excessive need for measuring performance, but only from students – whose dissenting voices cause no big problems – and only selected messages are conveyed. I surely couldn’t publish this piece in Nature, nor in any other high-profile journal. Even criticism of journals goes via journals. Even criticism of Nature, the journal every scientist wishes to publish in, goes via Nature (or its other publications). This should not be the case. 

Publishers and journals should take most of the blame for the current mental health crisis in academia. In fact, scientists should push for the formation of intergovernmental bodies that regulate how the scientific publishing system works, mitigating the effects of the economic interests of journals, who prioritize their profit margins over the needs of scientists. This may ultimately halt the mental health crisis young academics are facing, and help them to enjoy scientific research, and journey into the unknown with smiles on their faces.

Note: the survey was conducted in collaboration with Culturico.