Dementia and cognitive decline in incarcerated populations: Overall considerations for health policy

Introduction

Over the past several decades, prisons in the United States (US) have slowly started to resemble what some advocacy groups and researchers have described as a sort of nursing home behind bars.^1,2 In prisons, individuals are often considered ‘older adults’ at 50 or 55 years of age, while in the general population, that cutoff is generally closer to the age of 65.^3,4 This is in part due to the early onset of cognitive decline, neurodegeneration, and other chronic diseases associated with aging that are observed in incarcerated populations at relatively younger ages.^5–7 Throughout this piece, we use terms such as ‘incarcerated individuals’ or ‘incarcerated older adults’ to refer to adults held in state and federal prisons in the US. While we draw on international data where US-specific evidence is limited, mostly from other high-income countries, where research on dementia in prison is well developed, our policy recommendations are largely directed towards the US correctional system, though they may be applicable and have lessons to draw from globally.

Given these health disparities, there is a risk of an earlier onset of Alzheimer's disease and related dementias (ADRD) in populations experiencing incarceration. ⁸ Studies have found that the prevalence of dementia and cognitive impairment in prison populations substantially varies across geographic regions and can be differentiated by time in incarceration, time of diagnosis, and other demographic indicators.^9–11 One large cross-sectional study analyzed cognitive difficulty for older adults (>55) in prisons vs those in community samples and found the difference to be over two-fold (15.2% versus 7.1%). ¹² A South Carolina linkage study of dementia in correctional settings reported that, in 2016, ADRD prevalence among adults aged 55 years and older was ∼14.4% in the incarcerated population, compared with ∼6.9% in the general non-incarcerated population. ¹³ Outside the US, similar patterns are reported; in England and Wales, an estimated 8.1% of people in prison aged 50 and older have mild cognitive impairment or dementia, a rate much higher than some estimates for the same age group in the general population. ¹⁴

Epidemiology of aging and ADRD in prisons

The older adult population in prisons (in the US) is mostly made up of individuals who aged while in prison, not those who were recently arrested, as, in 2021, older adults (55+) made up 8% of total adult arrests. ¹⁵ Between 1999 and 2016, the number of people incarcerated in the U.S. decreased, but the number of incarcerated older adults increased by 280% (43,300 to 164,400).^16,17 This is a part of a trend in the makeup of prisons globally as incarcerated populations continue to age. For example, in the US, older adults in 1991 made up 3% of the total state and federal prison population, while estimates suggest that by 2022, those aged 55 and older constituted well over 15% of individuals in prison.^15,18 This trend is expected to continue rising and presents a serious challenge for correctional systems, which are not equipped, clinically, financially, or structurally, to care for an aging population with rising rates of ADRD.^1,16 The cost implications are substantial: the National Institute of Corrections has estimated that incarcerated individuals aged 55+ with chronic or terminal illnesses cost two to three times more on average than younger individuals, with more recent federal level analyses finding up to five times more on medical care and 14 times more on medication at facilities with the highest concentrations of older adults, adding to the fiscal and ethical pressure. ¹⁷

The reasons for the higher rates of cognitive decline observed in incarcerated populations are likely multifactorial. Many individuals enter prison with elevated baseline rates of established dementia risk factors, including hypertension, diabetes, smoking histories, and substance use disorders, often with limited access to preventive primary care before incarceration that might otherwise have helped manage them. ⁴ Rates of traumatic brain injury are also higher than in the general population, and average educational attainment tends to be lower, both of which may reduce cognitive reserve and contribute to earlier or more pronounced cognitive decline.^19,20 Once incarcerated, additional exposures may further compound this risk, including sleep disruption from facility schedules and noise, restricted physical activity, and limited access to nutritionally balanced meals. The concentration of these factors within correctional settings may help explain the prevalence gap observed across regional studies.

Mental health may also play an important and often underappreciated role. Rates of depression, anxiety, and post-traumatic stress disorder tend to be higher in incarcerated populations, and depression in particular shares a complicated, possibly bidirectional relationship with cognitive decline, in that depressive symptoms can sometimes mimic or mask cognitive impairment, while emerging cognitive decline can in turn worsen mood and motivation. ¹⁹ The prison environment itself may compound this through chronic stress, perceived threat, and the disruption of social support networks that older adults outside of prison often rely on. Solitary confinement and other forms of restrictive housing, which are sometimes imposed on individuals whose dementia symptoms are misread as misconduct, have also been linked to worsening psychiatric symptoms and may further accelerate this trajectory. ²¹ Personalized prevention and treatment programs for cognitive decline in this population may therefore benefit from integrating mental health care alongside cognitive care, rather than treating the two as separate concerns.

Structural and institutional barriers to diagnosis and care

One of the most evident shortcomings of health responses in prison systems in addressing ADRD among the resident populations is the potential underdiagnosis of their impairments. ²² The inadequacy of systems set up for the routine health screening and preventative health measures among populations experiencing incarceration has been well demonstrated previously. One study based in England and Wales reported that only 30% of prisons routinely utilized certain screening tools to identify potential cognitive impairment.^14,23 A cross-sectional study in the same region that assessed 869 older adults (>50) in prison found that only 3% of those who screened positively for the Montreal Cognitive Assessment (MoCA) had a confirmed diagnosis of dementia written in their healthcare notes. ²² In the Texas prison system, the largest of its kind in the US, one study found that only 15.4% of patients who met a dementia threshold via MoCA had a diagnosis for it documented in their medical records. ²⁴

One barrier in identifying and caring for the individuals experiencing cognitive impairment while in prisons is the lack of resources devoted to addressing it, coupled with federal or state-level policies that may sideline or ignore it in health system planning. ¹⁴ Given that estimates of its prevalence vary and there are routine shortages in attention to primary care and healthy longevity for people in incarceration, underdiagnosis remains a concern. This, in turn, can delay pathways to care and potential interventions for those in need. Many individuals who may have dementia are also not being readily assessed by clinical teams until there is substantial impairment, in part due to the collapse of close-knit social and support networks in prisons that are otherwise useful for identifying lapses in memory and cognition (and initiating treatment). ²⁵

In turn, this can lead to individuals facing difficulty in understanding, comprehending, and following prison rules or instructions by correctional officers. Some actions may be cast as disruptive, which could result in disciplinary infractions, relocation to more restrictive conditions and housing arrangements, verbal and/or physical abuse from peers, all of which may in turn serve to further deteriorate health and potentially accelerate cognitive decline. ¹⁶

There remain limited options for older adults to exit such a system, with some jurisdictions having a form of compassionate release and commuted sentences at times for severe declines in health.^26–29 Nevertheless, the process, time, and resources needed to navigate such an avenue are often burdensome and difficult to access, and often yield uncertain results. The absence of routine patient- and person-centered care that might enable individuals who are incarcerated to have access to the sort of resources necessary to track decline in health and cognitive status can further compound this. ³⁰ Some prisons have special units designated for those with dementia. ¹⁶ Entry into such units typically requires that some form of screening, referral, or clinical observation has already taken place, since these are often reserved for individuals with established diagnoses or with clear functional impairment noted by correctional or medical staff. Even so, by bringing geriatric and cognitive expertise into one place, these units may also help with case-finding more broadly, in that staff who are trained to recognize signs of decline may be better positioned to observe subtle changes in nearby older adults and to refer them for further assessment. This could be particularly useful for incarcerated individuals suspected of having ADRD who otherwise might find it difficult to advocate for themselves or secure remediations in their sentences. ³¹ But the presence of such units is rare across the world's prisons, and even compassionate release, when formulated into state statutes or policies, may base its analyses on physical impairments rather than cognitive decline. Some individuals may very well pass away before there are substantial improvements to their living conditions or a final decision is rendered on their cognitive decline claims.

Another limitation to addressing the needs of the aging population in prisons who face health declines and poor health outcomes stems from the prioritization of other goals embedded into prison systems and their associated structures, often designed to dehumanize individuals. For example, many private prisons, which, as of 2022, hold around 7.4% of incarcerated individuals in the US, require that the state keep a certain percentage of their beds filled or pay penalties.^32,33 Despite an executive order in 2022 aiming to end the practice of privately-operated prisons, it was applied only for federally-contracted facilities, with individual states continuing to operate them. ³² Private prison corporations have been reported to lobby for harsher policies that create an incentive to maintain high incarceration rates rather than pursue releases such as those granted through the compassionate release process. ³⁴ For example, it was found in a study that incarcerated individuals in private prisons have a delayed release of ∼90 more days compared to similar individuals in public prisons, in part, potentially due to a higher use of conduct violations. ³⁵ A state's attention to meeting bed guarantee thresholds may very well eclipse focus on the mental and physical health of individuals within such prisons.

Potential strategies and recommendations

There are a number of strategies that can be explored to address these limitations in caring for the aging incarcerated population. Regular screening for ADRD in older incarcerated individuals would allow for early detection and treatment that could delay the onset of advanced ADRD symptoms. Nevertheless, there is a lack of clinicians trained to do these sorts of screenings; many detention facilities may rely (formally or informally) upon correctional officers and other prison staff who may lack advanced education in the identification, assessment, or management of ADRD. ²⁵ Potential strategies to address this gap may include contracting external parties or clinicians to conduct regular assessments, creating specialized facilities for individuals with suspected cognitive decline, bringing in staff interested in working with such populations, and perhaps training the workforce within correctional facilities to administer cognitive screenings on incarcerated individuals. ³⁶ For example, in Fishkill (New York), a specific unit has been created to provide accommodations for individuals with dementia and is linked to the prison's medical facility. ²⁴ This makes it so that dementia symptoms that would otherwise be left unaddressed due to workforce limitations can be identified by social workers or correctional officers who would already be spending more time observing incarcerated individuals.

While the MoCA is able to be rapidly administered and can be useful for early screening, some have noted that the environment in correctional facilities is unique and quite different from other parts of the world and may thus require a context-specific assessment and diagnostic tools. ³⁶ This may need to be coupled with an architecture that is able to surveil and screen for dementia early on. For example, an intake and annual screening for everyone above a certain age threshold (e.g., 50) using a brief validated tool plus structured functional/PADL (prison activities of daily living) signals, perhaps followed by confirmatory assessment for positives with more detailed cognition testing, collateral history, and reversible-cause workup. Beyond age-based screening alone, there may also be value in being more specific about what should prompt a screening in the first place, rather than relying on ad hoc judgment. Triggers could include staff or peer reports of memory lapses, disorientation, or wandering, as well as repeated rule infractions that seem inconsistent with an individual's prior behavior, declining performance on activities of daily living such as hygiene or medication self-administration, self-reported cognitive concerns, and referrals following disciplinary incidents that may reflect cognitive impairment rather than willful misconduct. Embedding these sorts of triggers into existing intake, sick-call, and disciplinary review processes may help create a more streamlined identification pathway and reduce reliance on individual staff judgment. For patients who are identified to be at high risk or able to be clinically diagnosed, clinicians can then create individualized, dementia-specific care plans.

Another important aspect of treatment is to consider redesigns to the physical and cultural landscape within prisons and such facilities. Older adults in prisons may have physical limitations, impaired hearing or vision, and potentially a history of traumatic brain injury, all of which may serve to amplify potential risk factors for dementia, miscommunication, or misclassification of actions into grounds for disciplinary consequences. ¹² Sensory-friendly navigation clues, safer designs of housing for mobility (e.g., lower-bunk access), scheduled reorientations, fall prevention, and clinician-led medication reconciliation with attention to growing geriatric needs may be needed. ³⁷ Beyond providing safety assurances, similar redesigns can also function to provide an overall improvement in daily function. For example, the Special Needs Program for Inmates with Dementia (SNPID) provides alterations to the physical space for incarcerated individuals, such as having pictorial instructions of the steps to hand washing, which can be very helpful for individuals with ADRD. ³ Beyond physical adaptations, there may also be value in considering structured cognitive and cultural engagement, such as reading groups, educational programming, art and music activities, religious or community gatherings, and supervised group exercise, all of which may provide some degree of cognitive stimulation. Such activities are often cited in community-based dementia prevention frameworks but tend to remain underdeveloped in correctional settings, where access to programming is sometimes restricted as a disciplinary lever rather than considered as part of a broader health intervention. ¹⁹ There may also be benefits to considering programs for individuals to be paired in a sort of support network system, where carers may be able to receive recognition also for such services to aid in future employment prospects or reconsideration of their sentence. ²⁶

For patients with diagnosed or suspected dementia, release planning could begin at incarceration and be further mapped out 90 days prerelease: be it medication continuity, summary of cognitive care received, community appointment booking before gate release, caregiver/family briefing, benefits activation, and warm handoff to relevant community primary care, geriatrics, or behavioral health staff. ³⁸ This might be more feasible now as, in the US, the Center for Medicare and Medicaid Services (CMS) has approved multiple state Section 1115 reentry demonstrations or waivers, which permit targeted Medicaid-covered services before release (including a 90-day prerelease window in approved designs). POPS (Project for Older Prisoners), which is operated by law school students, is an example of a possibly effective method of advocating for a fair pathway to the release or the change in setting of incarceration for those who suffer from ADRD. ³ The use of a similar legal infrastructure or advocacy groups run by medical trainees and physicians that exist outside of the prison administration and is dedicated to older incarcerated individuals may help improve the chances of receiving compassionate releases.

Social isolation is itself thought to be a modifiable risk factor for cognitive decline and dementia, and incarceration tends to limit social contact in fairly structural ways, through restrictions on visitation, the use of segregation, and the erosion of social networks that may have been in place before incarceration. ¹⁹ Restrictive housing and solitary confinement, which are sometimes applied to individuals whose dementia symptoms have been misread as misconduct, may further compound this. ²¹ In that context, peer-care programs may serve a dual purpose, in that they provide some structured care for those with dementia while also offering meaningful social engagement and a sense of purpose for the caregivers themselves.

Although the overall evidence remains limited due to minimal adoption, there are a few case studies of their use that suggest that dementia-aware policy planning within prisons may aid in improving safety, continuity, and person-centered care for older adults. Some facilities in the US (e.g., in New York and California) have created specific units for patients with dementia, with some adopting specialized training for their prison staff as well. ²⁶ Others have also experimented with peer-support networks and developed networks of allied health professionals to support patients with dementia (e.g., in Australia). ²⁶ For example, the Gold Coats program in California is one in which incarcerated individuals with records of good conduct are selected and trained to identify signs of dementia and to provide basic care for fellow incarcerated individuals who are showing symptoms. ³ Its success in lessening behavioral problems, as well as its use of incarcerated individuals, makes it a resource-saving, effective, and realistic integration that can bring forth real improvements.

Conclusion

The challenges outlined here can also be understood as one part of a broader and more long-standing inadequacy in correctional healthcare more generally. The Federal Bureau of Prisons, despite operating under the Department of Justice and with some involvement from the US Public Health Service, tends to function with relatively limited external clinical oversight and limited expert ethical guidance. State systems vary widely but face many of the same structural constraints. Despite fairly widespread recognition of these issues, the popular press, news media, and government have generally been slow to engage with healthcare in prisons in any sustained way. If the social determinants of health are to be taken seriously in this setting, then addressing ADRD in incarcerated populations may require more than clinical recommendations or scholarly commentary alone, and may also involve sustained legislative engagement, independent oversight, and a more public reckoning with the conditions that follow individuals into and out of carceral settings. The aging incarcerated population is, in some ways, a proxy for a much larger humanitarian concern with implications across many parts of wider society.

There currently remains a paucity of research regarding ADRD in individuals experiencing incarceration (an area where general medical literature is lacking anyhow). The resulting lack of knowledge of the true prevalence of ADRD in incarcerated populations is also further exacerbated by the aforementioned reasons for underdiagnosis. Another contributor to the gap in understanding of how to address cognitive decline is that the Bureau of Justice Statistics (BJS), which is the principal source of national prison health statistics in the US, does not appear to systematically collect data on dementia or cognitive impairment in its surveys of incarcerated populations, which means that national-level prevalence estimates often have to rely on extrapolation from regional studies that use varying methodologies. ¹³ Greater inclusion of such populations in research designs may assist in creating strategies for future guidelines, recommendations, and policy proposals. For example, making it so that governmental and, more broadly, research institutions collect data targeted towards ADRD in incarcerated populations may help uncover what's currently unknown. This will inform prison systems, clinicians, and advocacy groups to allocate the necessary resources needed to create a more humanizing approach to treating incarcerated populations affected by ADRD. Meaningful change, however, will likely require legislative engagement alongside clinical and research efforts, since the scale of the problem is unlikely to be addressed by scholarly commentary alone.