Systematic review finds predictive models for dysphagia risk in older adults show variable performanceCan we predict swallowing trouble in older adults before it happens?

Frontiers in Medicine Published April 3, 2026 Study authors: Shujun Fan, Bingbing Chen, Chuan Yin, Yu Hao, Wanning Liu, Yu Zhang, Qin Liu, Zhe Zhang DOI ↗ Editorial oversight: Dr. Sofia Müller, MD · Lifespan & Whole-Person Care

AI-generated summary of the cited source, checked by automated accuracy review. How we work

Key Takeaway

Note: Predictive models for dysphagia risk show promise but require more validation before clinical use.

This systematic review and meta-analysis evaluated the performance of predictive models for dysphagia risk in older adults. The analysis included 17 studies identified from 7,113 records, though specific study settings, interventions, and comparators were not reported. The primary outcome was model performance measured by Area Under the Curve (AUC).

The pooled AUC across studies was 0.82 (95% CI: 0.77–0.88), with individual study AUCs ranging from 0.682 to 0.926. Common predictors identified across models included advanced age, smoking history, reduced self-care ability, polypharmacy, frailty, malnutrition, cognitive impairment, and poor oral health. No safety or tolerability data were reported.

Key limitations significantly constrain interpretation. All included studies showed a high overall risk of bias, and the authors noted methodological flaws throughout the evidence base. Only two studies performed external validation, indicating insufficient testing of model generalizability. The review concludes that predictive modeling for dysphagia in older adults remains in an early stage of development.

For clinical practice, this evidence suggests current predictive models show variable discrimination but lack proven reliability. The high risk of bias and limited external validation mean these tools should not yet be relied upon for clinical decision-making. Further rigorous development and validation are needed before considering implementation.

Imagine being able to know which older adult might develop trouble swallowing before they ever choke on a meal. A fresh look at the research suggests we're getting closer to building that kind of early warning system. The analysis combined data from 17 studies and found that predictive models, which use clues like a person's age, frailty, medication list, and cognitive health, showed promise in identifying risk. On average, these tools performed moderately well at sorting who might be at risk. But here's the crucial catch: this is very early science. The researchers themselves warn that every study they looked at had serious flaws in how it was designed. Only two studies tested their models on a separate group of people to see if they really worked, which is a critical step. In short, the idea is promising, but the current tools are not reliable enough for real-world use. Doctors and families still need to rely on careful observation and clinical exams.

What this means for you:

Early tools to predict swallowing risk show promise but are too flawed for clinical use.

Study Details

Study typeMeta analysis

EvidenceLevel 1

PublishedApr 2026

View Original Abstract ↓

Dysphagia is a common condition among older adults, closely linked to aging and neurodegenerative diseases. It can lead to malnutrition, frailty, and aspiration pneumonia, thereby impairing quality of life and clinical outcomes. Although various interventions may improve swallowing function, early identification of high-risk individuals remains challenging. Existing predictive models show inconsistent performance and lack systematic evaluation. This study aimed to systematically review and assess predictive models for dysphagia risk in older adults. A comprehensive search was conducted across CNKI, the Chinese Science and Technology Journal Database, the Chinese Biomedical Literature Database, Wanfang Data, PubMed, Web of Science, and the Cochrane Library, covering studies published up to September 15, 2025. Of 7,113 records identified, 17 met inclusion criteria, with only two performing external validation. Reported AUCs ranged from 0.682 to 0.926, and all studies showed a high overall risk of bias. The pooled AUC from the meta-analysis was 0.82 (95% CI: 0.77–0.88). Common predictors included advanced age, smoking history, reduced self-care ability, polypharmacy, frailty, malnutrition, cognitive impairment, and poor oral health. Overall, predictive modeling for dysphagia in older adults remains in an early stage, limited by methodological flaws and insufficient external validation. Future research should follow PROBAST standards and conduct large, multicenter validations to improve model reliability and clinical utility.

Systematic review finds predictive models for dysphagia risk in older adults show variable performanceCan we predict swallowing trouble in older adults before it happens?

Study Details

Group dance interventions improved physical performance and cognition in community-dwelling older adults with sarcopenia

Why Group Dance Is Becoming Medicine for Aging Muscles

Clinical research that matters. Delivered to your inbox.

Systematic review finds predictive models for dysphagia risk in older adults show variable performanceCan we predict swallowing trouble in older adults before it happens?

More on Dysphagia

Study Details

Group dance interventions improved physical performance and cognition in community-dwelling older adults with sarcopenia

Why Group Dance Is Becoming Medicine for Aging Muscles

Clinical research that matters. Delivered to your inbox.

Related in Geriatrics & Aging

From Other Specialties