Two reviews on populations the health evidence base barely covers

Two systematic reviews published in The Lancet Public Health on the same day, 11 June, examine two very different populations. What they have in common is the shape of their conclusions: a real health burden, and an evidence base too thin and too biased to describe it properly.

Young carers

The first review synthesises longitudinal evidence on the physical and mental health of young carers — people under 25 who provide care — and on the psychosocial factors that shape it [s1].

Databases were searched to 31 December 2024 for longitudinal, health-focused studies that included a non-carer comparison group [s1]. After screening 4,362 records, 18 studies met the inclusion criteria and 17 were retained for narrative synthesis [s1].

Every one of those studies was conducted in a high-income country [s1]. That single sentence defines the limits of what can be said: caregiving by children and young people is not a high-income phenomenon, and the literature is.

The 17 studies contributed 26 associations between caregiving and health — 20 for mental health, six for physical health [s1]. Three studies identified links between psychosocial factors and mental health [s1].

The findings suggest a small-to-moderate risk of poorer mental health among young carers, with stronger effects in subgroups such as high-intensity carers, though the authors flag notable inconsistencies across studies [s1]. Evidence on physical health was inconclusive [s1]. Social connections appear to play a key part in young carers' mental health [s1].

"Small-to-moderate" and "inconsistent" are the honest summary of six associations for physical health and twenty for mental health across seventeen studies. The review calls for further large-scale longitudinal studies to clarify the inconsistencies [s1].

Women in prison

The second review is larger in volume and, if anything, weaker in quality of underlying evidence [s2].

The authors searched peer-reviewed and grey literature for quantitative studies published between 1 January 2003 and 29 January 2025, covering people detained in carceral spaces designated for women as part of the criminal-legal system worldwide, and reporting prevalence of health conditions classified by the Global Burden of Disease Study or ICD-11 [s2].

From 18,008 unique records, 247 studies were included, covering more than 452,261 women [s2].

Then the finding that governs everything else: nearly all studies had a high risk of bias in at least one domain, assessed with the JBI Critical Appraisal Checklist for prevalence data [s2].

Communicable diseases and mental health conditions were the most frequently described topics [s2]. Prevalence of many conditions varied widely between studies and across geographical regions [s2]. The authors identify gaps particularly around non-communicable conditions and in the geographical representation of the data [s2].

The review also names a specific direction of bias. Some studies excluded people with complex health-care needs, and many relied on measures such as self-report that depend on previous access to health care [s2]. Both would push measured prevalence down. The authors conclude that the true burden of health conditions among incarcerated women is likely to be greater than what the literature records [s2].

What "evidence gap" actually means here

It is tempting to read both reviews as saying "we need more research", which is what every systematic review says. The more specific reading is worth extracting.

In the young carers literature, the gap is geographic and structural: no low- or middle-income country studies met the inclusion criteria at all, so nothing in this synthesis describes young carers outside high-income settings [s1].

In the women-in-prison literature, the gap is one of measurement validity: 247 studies exist, and their prevalence estimates are not comparable to each other because they were produced with different instruments in different systems, most with identified risk of bias [s2]. Volume did not produce precision.

Neither review can be used to state a prevalence figure for anything, and neither states one. That is a limitation the authors report rather than one we are imposing on them.

Where they converge

Both populations are defined by a circumstance rather than a diagnosis — providing care, or being detained — and in both cases the circumstance is what makes them hard to study. Young carers are frequently unidentified as carers; incarcerated women are, by definition, in institutions that mediate access to researchers and to health care alike.

The Lancet Public Health review on incarcerated women closes with a call to the correctional, health and research communities to act on health inequities in this population [s2]. The young carers review calls for larger longitudinal studies and highlights social connection as the psychosocial factor with the clearest signal [s1].

What to watch

Whether either literature acquires studies from outside high-income settings, and whether prevalence work in prisons adopts standardised instruments so that the next review can pool rather than narrate [s1][s2].

Sources

• [s1] Health outcomes and psychosocial determinants in young carers: a systematic review. The Lancet Public Health, 11 June 2025. https://doi.org/10.1016/S2468-2667(25)00099-4
• [s2] Health conditions among women in prisons: a systematic review. The Lancet Public Health, 11 June 2025. https://doi.org/10.1016/S2468-2667(25)00092-1