Mental health and benefits insecurity


Many users of mental health services rely on state benefits. Working in a Home Treatment Team, I noticed that service users worry about their benefits and upcoming further assessments. While this experience appears to point towards a relationship between the process of accessing welfare benefits and worsening mental health, without further research, it is difficult to be more specific about this association.

In Britain, the Work Capability Assessment (WCA) was gradually introduced from 2010 in an effort to re-assess eligibility and ensure costs were kept as low as possible. This proved controversial as qualitative research suggested that the assessments for benefits were very stressful and not only does the risk of losing benefits cause distress, but also the bureaucracy that comes with it (Shefer et al., 2016). Indeed, recent data published by the Department of Work and Pensions (2020) itself showed 70% of Fit for Work decisions have been overturned.

The ensuing backlash of the introduction of the WCA led Barr and colleagues to investigate the health effects of the assessment (Barr et al., 2016). They exploited the different rates at which different local authorities introduced the WCA and estimated that for each additional 10,000 people reassessed there were on average an additional 6 suicides, 2,700 cases of self-reported mental health problems, and an additional 7,020 prescriptions for antidepressants. They were unable to assess specific symptoms in benefit claimants, because they only had access to data at population level. Furthermore, the self-reported mental health data were based on the Quarterly Labour Force Survey, containing the questions: “Do you have depression, bad nerves or anxiety?”  and “Do you have mental illness or suffer from phobias, panics or other nervous disorders?” – questions that are somewhat vague and combine conditions.

A similar study has now been conducted by Stuart and colleagues, but they have gone further (Stuart et al., 2020). They wanted to see if the assessments and benefits problems in general led to poorer mental health and, whether the subsequent costs of using health services negated the reduction of financial expenditure on benefits, thereby defeating the purpose of reducing overall costs.

The WCA has proven controversial with studies showing the detrimental effect its introduction has had on mental health.

The Work Capability Assessment has proven controversial with studies showing the detrimental effect its introduction has had on mental health.


The authors used a prospective cohort design to compare an exposed and a control group with follow-ups at 3, 6 and 12 months (Stuart et al., 2020). The exposed group (n=42) contacted the benefits advice centre for help with current assessment and review of their benefits. The control group (n=45) consisted of people who had asked the same benefits centre for advice in the last 18 months and were receiving the correct amount of benefits, according to benefit advisors.

Inclusion criteria:

  • Self-reported mental health problem
  • Aged 16–64 years
  • In receipt of benefits

Exclusion criteria:

  • Inability to provide informed consent

The authors used the following questionnaires:

  • Warwick–Edinburgh Mental Wellbeing Scale (WEMWBS)
  • Zung Self-Rating Anxiety Scale
  • Centre for Epidemiologic Studies Depression Scale (CES-D)
  • Empowerment Scale to measure self-esteem/self-efficacy
  • Client Service Receipt Inventory (CSRI) to measure service use and medical records checked to measure the use of inpatient care

The researchers also collected data on demographics, current benefits and income. The questionnaires were filled in by post, via the telephone or at attendance days with the help of trained volunteers, researchers or benefits advisors.


The exposed and control groups were not completely equal, for example 64.3% of the exposed group indicated that they had never been admitted to psychiatric hospital compared to 46.7% of the control group. Also, 52.4% of the exposed group reported that they were disabled due to both mental and physical health conditions, and 28.6% considered themselves disabled due to a mental health problem only. In contrast, 51.1% of the control group indicated they were disabled due to mental illness only and 35.6% reported to be disabled due to both mental and physical health conditions.

There was also considerable drop-out. At baseline, there were 87 participants, but by 12 months there were only 28 participants remaining (12 in the exposed group and 16 in the control group). Moreover, many participants in the control group were also reassessed for benefits eligibility, even though at recruitment benefits advisors believed they were on the correct benefits.

At baseline, the control group scored higher on the wellbeing scale (WEMWBS) (exposed group = 30.4, control group = 36.2, p<0.05). There were also significant differences in anxiety (exposed group = 3.2, control group = 2.6, p<0.05) as well as the level of depression (exposed group = 43.3, control group = 35.8, p<0.05).

There were no differences between the groups in anxiety and depression at 3, 6 and 12 months, but there was a difference on the WEMBWS wellbeing scale at 12 months (exposed group = 28.8, control group = 38.7, p<0.05), but no difference at 3 or 6 months. There were no differences on the empowerment scale at any time point.

In terms of service use, a higher percentage of the control group used services throughout the study. There were 30 admissions to hospital but only two admissions for mental health, both in the control group. If inpatient costs were excluded, the mean total service cost was lower for the exposed group, but this could be chance variation. Mean income was higher for the control group throughout the study, but there were measurement difficulties, as it was unclear whether everybody counted housing benefit as income.

The results indicate no difference between the groups in levels of anxiety or depression but how confident can we be with significant drop-out?

The results indicate no difference between the groups in levels of anxiety or depression, but how confident can we be with significant drop-out?


The study attempted to answer a very important question, namely whether assessment for benefits eligibility was associated with a deterioration in mental health.

The results should be interpreted with caution. There were some differences between the exposed and the control group at the start of the study which is probably unavoidable in practice with this type of study. There were other difficulties too and the authors stated that their findings only partly supported the hypothesis that the benefits assessment itself was associated with mental health difficulties.

People in the control group reported difficulties with benefits that were not expected, which led the authors to question whether people with mental health conditions are being assessed more frequently than others.

Another problem was the loss to follow-up. The authors suggest that perhaps participants found filling in the same questionnaire repetitive, perhaps they were in hospital, or perhaps they were not able to maintain a telephone line. It is a possibility that the more unwell people dropped out. The authors stated that for future research, a longitudinal follow-up study would be a better method than comparing two groups.

Strengths and limitations


  • Mental health symptoms were measured via standardised questionnaires
  • Although this was not one of the aims of the study, the benefit problems experienced in the control group are an important finding. At the moment, at least in the UK, there is probably no such thing as benefits security; in all likelihood, there will be assessment after assessment


  • Exposed and control groups were different, e.g. more people in the exposed group reported being disabled due to both mental and physical health problems
  • The significant loss to follow-up presents the possibility that greater deterioration in mental health could be one reason for dropping out. Regardless, it limits the strength of the findings
  • Although the authors suggested a longitudinal study to address the limitations, but loss to follow-up will also be a problem with a longitudinal study. Hence, it will be important to discuss at recruitment with potential participants the importance of completing questionnaires repeatedly and to budget for the possibility of making home visits

Of note is that although the control group had on average higher scores on the WEMWBS (general well-being scale), the scores of both groups were lower than at the original validation study. When investigating the validity of the scale, the median score was 50 in a student sample and 51 in a population sample (Tennant et al., 2007). 

A noble effort in answering a very important question but the limitations of the study leave many questions unanswered.

This study makes a noble effort at answering a very important question, but the limitations of the research leave many questions unanswered.

Implications for practice

  • Financial difficulties are common in people who use mental health services and can lead to housing and food insecurity
  • Clinicians should actively ask about financial matters during initial assessment
  • Patients may need to go to a citizen’s advice bureau, the welfare benefits team in the local trust and/or the local food-bank
  • This study adds that it is important for clinicians to realise that there is probably no such thing as benefits security and benefits assessments and reviews can be stressful both because of the bureaucracy involved and the potential loss of income
  • It is important to offer help with this. See for example, Supporting Claimants: a practical guide by Jay Watts in Asylum Magazine from 2018.
Despite the study’s lack of strong conclusions, clinicians should be vigilant during assessments of the added stress of financial difficulties.

Despite the study’s lack of strong conclusions, clinicians should be vigilant during assessments of the added stress of financial difficulties.

Statement of interests

Dieneke Hubbeling does not have any conflicts of interest, other than working in a crisis and home treatment team.


Primary paper

Stuart R, Campbell S, Osumili B, et al. (2020) Do welfare benefit reassessments of people with mental health conditions lead to worse mental health? A prospective cohort study. International Journal of Social Psychiatry 66(2): 136–149.

Other references

Barr B, Taylor-Robinson D, Stuckler D, et al. (2016). ‘First, do no harm’: Are disability assessments associated with adverse trends in mental health? A longitudinal ecological study. Journal of Epidemiology and Community Health 70(4): 339–345.

Department for Work & Pensions. (2020, June). ESA: Work Capability Assessments, Mandatory Reconsiderations and Appeals: June 2020. Retrieved from:

Shefer G, Henderson C, Frost-Gaskin M, et al. (2016). Only Making Things Worse: A Qualitative Study of the Impact of Wrongly Removing Disability Benefits from People with Mental Illness. Community Mental Health Journal 52(7). Springer US: 834–841.

Tennant R, Hiller L, Fishwick R, et al. (2007). The Warwick-Edinburgh mental well-being scale (WEMWBS): Development and UK validation. Health and Quality of Life Outcomes 5: 1–13.

Photo credits

Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+