The data captured workers at a single point in time, rather than tracking them before and after treatment. Using thousands of matched pairs from the same workplace, it compared wellbeing measures from workers who participated in wellness programs with those of their colleagues who did not.

It is possible that there was selection bias, since workers who enroll in, say, a resilience training program may have lower wellbeing to begin with, Fleming said. To address that, he separately analysed responses from workers with high preexisting levels of work stress, comparing those who did and did not participate. But among this group, too, the survey answers suggested that the programs had no clear benefit.

The findings call into question practices that have become commonplace across job sectors. But researchers said they came as no surprise.

“Employers want to be seen as doing something, but they don’t want to look closely and change the way work is organised,” said Tony D. LaMontagne, a professor of work, health and wellbeing at Deakin University, who was not involved in the study.

Workplace mental health interventions may send the message that “if you do these programs, and you’re still feeling stressed, it must be you,” LaMontagne said. “People who don’t have a critical view might internalise that failure: ‘So I really am a loser.’”

The corporate wellness services industry has ballooned in recent years, with thousands of vendors competing for billions of dollars in revenue. Companies invest in the interventions in hopes of saving money overall by improving worker health and productivity.

Some research supports this expectation. A 2022 study tracking 1,132 workers in the US who used Spring Health, a platform that connects employees with mental health services like therapy and medication management, found that 69.3 per cent of participants showed improvement in their depression. Participants also missed fewer days of work and reported higher productivity.

Adam Chekroud, a co-founder of Spring Health and an assistant professor of psychiatry at Yale, said Fleming’s study examined interventions that were “not highly credible” and measured wellbeing many months later. A blanket dismissal of workplace interventions, he said, risks “throwing the baby out with the bathwater”.

“There is recent and highly credible data that things like mental health programs do improve all those metrics that he mentions,” Chekroud said. “That’s the baby you shouldn’t be throwing out.”

There is also solid evidence that practices like mindfulness can have a positive effect. Controlled studies have consistently demonstrated lower stress and decreased anxiety and depression after mindfulness training.

Loading

The lacklustre benefits that Fleming found may reflect variations in offerings, said Larissa Bartlett, a researcher at the University of Tasmania who has designed and taught mindfulness programs. “Light-touch” interventions like apps, she added, are generally less effective than one-on-one or group trainings.

Fleming’s study, she said, “misses most of these details, condensing intervention types into broad labels, engagement into yes/no, and dismissing the reports from intervention participants that they felt they benefited from the programs they did”.

A key omission, she added, is longitudinal data showing whether participants experience improvement over time. The result is a “bird’s-eye view” of the wellbeing of participants that “skates over changes that may occur at the individual level”, she said.

Fleming said he was aware of the body of research supporting the treatments’ effectiveness, but that he had “never been as convinced by the very positive findings” since the data comes from controlled trials in which the treatment is implemented very well, something that may not be the case in employer-provided programs.

Dr David Crepaz-Keay, head of research and applied learning at the Mental Health Foundation in the UK, has advised the World Health Organisation and Public Health England on mental health initiatives. He said Fleming’s data and analysis as “certainly more robust” than “most of the research that has created the consensus that employee assistance works”.

This article originally appeared in The New York Times.