Mental and Physical Health Among People in Prisons

Lancet Public Health | 2024, Vol. 9, No. 4, p. 250 - 260

Mental and Physical Health Morbidity Among People in Prisons: An Umbrella Review

Louis Favril; Institute for International Research on Criminal Policy, Faculty of Law and Criminology, Ghent University, Ghent, Belgium

Josiah D Rich; Warren Alpert Medical School, Brown University, Providence, RI, USA

Jake Hard; HM Prison Cardiff, Cardiff, UK

Seena Fazel; Department of Psychiatry, University of Oxford, Oxford, UK, and Oxford Health NHS Foundation Trust, Oxford, UK

Background: People who experience incarceration are characterized by poor health profiles. Clarification of the disease burden in the prison population can inform service and policy development. We aimed to synthesize and assess the evidence regarding the epidemiology of mental and physical health conditions among people in prisons worldwide. Methods: In this umbrella review, five bibliographic databases (Web of Science, PubMed, PsycINFO, Embase, and Global Health) were systematically searched from inception to identify meta-analyses published up to Oct 31, 2023, which examined the prevalence or incidence of mental and physical health conditions in general prison populations. We excluded meta-analyses that examined health conditions in selected or clinical prison populations. Prevalence data were extracted from published reports, and study authors were contacted for additional information. Estimates were synthesized and stratified by sex, age, and country income level. The robustness of the findings was assessed in terms of heterogeneity, excess significance bias, small-study effects, and review quality. The study protocol was preregistered with PROSPERO, CRD42023404827. Findings: Our search of the literature yielded 1909 records eligible for screening. One thousand seven hundred thirty-six articles were excluded, and 173 full-text reports were examined for eligibility. One hundred forty-four articles were excluded due to not meeting inclusion criteria, resulting in 29 meta-analyses eligible for inclusion. 12 of these were further excluded because they examined the same health condition. We included data from 17 meta-analyses published between 2002 and 2023. In adult men and women combined, the 6-month prevalence was 11·4% (95% CI 9·9–12·8) for major depression, 9·8% (6·8–13·2) for post-traumatic stress disorder, and 3·7% (3·2–4·1) for psychotic illness. On arrival to prison, 23·8% (95% CI 21·0–26·7) of people met diagnostic criteria for alcohol use disorder and 38·9% (31·5–46·2) for drug use disorder. Half of those with major depression or psychotic illness had a comorbid substance use disorder. Infectious diseases were also common; 17·7% (95% CI 15·0–20·7) of people were antibody-positive for hepatitis C virus, with lower estimates (ranging between 2·6% and 5·2%) found for hepatitis B virus, HIV, and tuberculosis. Meta-regression analyses indicated significant differences in prevalence by sex and country income level, albeit not consistent across health conditions. The burden of non-communicable chronic diseases was only examined in adults aged 50 years and older. Overall, the quality of the evidence was limited by high heterogeneity and small study effects.

Interpretation: People in prisons have a specific pattern of morbidity that represents an opportunity for public health to address. In particular, integrating prison health within the national public health system, adequately resourcing primary care and mental health services, and improving linkage with post-release health services could affect public health and safety. Population-based longitudinal studies are needed to clarify the extent to which incarceration affects health.

“Worldwide, more than 11 million individuals are incarcerated on any given day. The life trajectories of people who experience incarceration are typically characterized by poor educational attainment, unemployment, unstable housing, poverty, and trauma—social determinants that negatively affect health. Compared with the general population, people living in prisons, jails, and juvenile detention facilities (collectively referred to as prisons hereafter) disproportionately experience mental health problems, substance misuse, infectious diseases, and chronic conditions” (p. 250). 

This article “conducted an umbrella review to systematically collect and review published meta-analyses examining the prevalence and incidence of mental and physical health conditions in prison populations… Our systematic search of the literature yielded 1909 records eligible for screening. Following the exclusion of 1736 articles based on title and abstract, 173 full-text reports were examined for eligibility. 144 articles were subsequently excluded on the basis of design,  population, and outcome assessment, resulting in 29 meta-analyses that met our inclusion criteria. 12 of these were further excluded because they examined the same health condition, resulting in 17 meta-analyses being included in our review” (p. 251-253).

“Findings indicate that incarcerated individuals experience poor health across a wide range of mental and physical conditions. However, prevalence estimates should be interpreted in light of high heterogeneity, small-study effects, and risk of bias in the underlying meta-analyses. We report four main results. First, the burden of treatable mental disorders among incarcerated individuals is substantial. One in every ten people was diagnosed with depression (11%) or PTSD (10%), and psychotic illness affected about 4% of the prison population. A quarter (24%) of people who enter prison were found to have an alcohol use disorder and 39% a drug use disorder” (p. 255). 

“Second, our findings indicate a high prevalence of infectious diseases. Around one in six (18%) people in prisons had a current or past hepatitis C virus infection, with relatively lower estimates found for hepatitis B virus (5%), HIV (3%), and tuberculosis (3%). Bacterial sexually transmitted infections such as chlamydia (9%) were also common, with higher rates in women than in men. Among people in prison aged 50 years and older, non-communicable diseases including hypertension (39%), diabetes (14%), and asthma (7%) were generally more prevalent compared with their younger peers in prisons” (p. 255).

“Third, pooled prevalence estimates should be interpreted in the context of small-study effects and high heterogeneity, which were present in many meta-analyses contributing to this umbrella review. Small study effects might indicate publication bias, likely leading to inflated prevalence estimates. The high level of heterogeneity (with I²>90% for most conditions) might be due to the primary studies being conducted in a large variety of prison settings and expected changes in prevalence over time (eg, due to policy reforms). Prison populations are likely to vary substantively between countries owing to differences in national policies regarding management of health conditions within the criminal justice system (eg, alternative sentencing and diversion strategies), which might further contribute to heterogeneity…Furthermore, around two-thirds of meta-analyses included in our review had moderate or high risk of bias, with common limitations including insufficient consideration of heterogeneity and bias in primary studies” (p. 255).

“Finally, this umbrella review highlights several key gaps in the meta-analytic evidence base. First, none of the included meta-analyses examined incidence rates. Second, there was a notable lack of evidence on the prevalence of non-communicable diseases such as cancer, diabetes, and cardiovascular disease. The only identified meta-analysis in this area was restricted to older adults. Third, meta-analyses on epilepsy and personality disorders were published more than years ago and thus require updating, for example, by including prevalence data on borderline personality disorder in men. Fourth, the global burden of bipolar disorder, anxiety disorders, eating disorders, and autism spectrum disorder in prison populations has not yet been meta-analytically reviewed to date. In contrast, intellectual disability and traumatic brain injury require further investigation based on reliable and clinically informative diagnostic criteria. Fifth, we did not identify any meta-analyses examining the comorbidity between mental and physical health conditions” (p. 255 - 256).  

“The disproportionate burden of mental and physical disease in the prison population presents both challenges and opportunities. The prison context poses unique challenges to the delivery of healthcare services, including security requirements. Frequent movements between prisons (often with no transfer of medical records) and short stays make engagement with health care difficult, and structural barriers such as overcrowding and understaffing further impede the optimal delivery of  Additionally, individual-level barriers that prevent people in prisons from accessing available services include distrust of the health-care system, low health literacy and help-seeking behaviour, and fear of stigmatization. Despite these challenges, incarceration provides a unique time window during which the multifaceted health needs of underserved populations can be assessed, diagnosed, and treated—often for the first time. Because almost all people in prisons will be released at some point, improving their health during imprisonment can equally improve the health of the communities to which they will return, hence producing a public health benefit. Treatment of mental illness and substance misuse might additionally contribute to public safety by decreasing rates of reoffending.105 In turn, these effects could lead to economic benefits by reducing the burden on health and criminal justice systems. In conclusion, incarceration provides an important opportunity to address unmet health needs in a vulnerable population, which can positively impact public health, public safety, and society as a whole” (p. 257).

“Strengths of this umbrella review include synthesising a broad range of health conditions and using methodological tests to assess the quality of evidence. However, there are also several limitations. First, our findings only apply to health conditions in general prison populations and might not be generalisable to selected and high-risk groups, in which prevalence is likely to be higher. Second, because we only considered health conditions that have been subject to meta-analysis, other health conditions reported in narrative or systematic reviews (without quantitative synthesis) were not included in our overview. Third, when multiple eligible meta-analyses evaluated the same health condition, we retained the one with the highest quality, provided that individual-level study estimates were available. This latter criterion led to findings on infectious diseases being based on meta-analyses of low quality. Fourth, our global scope might have masked important differences between countries and regions in terms of prevalence, resources, and policies” (p. 256).