Study Finds High Malnutrition Burden Among Institutionalized Adults Over Eighty

A cross-sectional study evaluating adults aged 80 years and older admitted to integrated medical and long-term care facilities in Chengdu, Sichuan, China, found that more than half of newly admitted residents exhibited screening scores indicating active malnutrition. Conducted across three institutions combining inpatient medical services with long-term residential nursing, the investigation documented that 52.7 percent of participants met the threshold for malnutrition on the Mini Nutritional Assessment–Short Form, with a 95 percent confidence interval spanning 46.1 to 59.4 percent. An additional 44.1 percent of residents were identified as being at risk of malnutrition, establishing that 96.8 percent of the enrolled cohort presented with some degree of nutritional vulnerability at entry. Only 3.2 percent of the evaluated individuals demonstrated normal nutritional status, establishing a baseline of profound clinical vulnerability among this very old institutionalized population.

The demographic backdrop of the study centers on the rapid demographic aging observed across China, where census estimates confirm expanding proportions of older citizens, particularly the oldest-old cohort aged 80 years and older. In response to increasing multimorbidity, cognitive impairment, and physical frailty, health authorities have increasingly emphasized integrated models connecting clinical treatment directly to daily residential nursing. However, nutritional compromise frequently compounds these existing chronic illnesses without receiving dedicated early screening. The authors noted that malnutrition in advanced age is closely intertwined with chronic inflammation, anorexia, sarcopenia, and physical disability, each accelerating institutionalization and healthcare utilization. While the prognostic relevance of nutritional status has been widely documented in community and acute hospital settings, systematic data detailing admission profiles specifically within integrated long-term residential facilities in non-core urban areas have historically remained limited.

To investigate admission vulnerability, researchers enrolled 220 newly admitted residents aged 80 years and older between July 2024 and June 2025 across three designated care institutions in Chengdu. Eligible facilities were required to house more than two hundred beds and operate dual-capacity services spanning medical rehabilitation and long-term assistance. Sampling utilized a stratified convenience cluster technique across administrative districts, enrolling residents who completed structured baseline screenings within their first week of admission. Residents receiving palliative end-of-life care, individuals under age 80, and those who were completely bedridden and incapable of completing assessments were systematically excluded, as were residents who declined biochemical sampling. The final sample size of 220 participants exceeded the minimum calculation of 139 subjects required for prevalence estimation, providing more than ten outcome events per candidate predictor variable in subsequent multivariable models.

Data collection followed a rigorous standardized protocol administered by trained investigators at initial admission. The primary screening instrument was the Mini Nutritional Assessment–Short Form, an established six-item evaluation that reviews food intake decline, recent involuntary weight loss, mobility restrictions, acute illness or psychological distress, neuropsychological problems, and body mass index. Scoring on the instrument ranges from 0 to 14 points, stratifying respondents into ranges denoting malnutrition from 0 to 7, malnutrition risk from 8 to 11, and normal nutritional standing from 12 to 14. Anthropometric assessments involved calibrated standing, chair, or bed scales, alongside height determinations calculated through direct supine length or Bassey demi-span equations using sternal-notch-to-finger measurements. Activities of daily living were systematically quantified using the ten-component Barthel Index, which grades independence across self-care and mobility tasks from 0 to 100.

Baseline profile evaluations revealed a mean participant age of 87.46 years, with women representing 65.5 percent of the total analytic group. Educational attainment was predominantly at or below the middle-school level for 83.2 percent of participants, while 75.9 percent were widowed or living without a partner. Functional assessments through the Barthel Index indicated pronounced dependency, with 79.5 percent of participants registering scores between 0 and 40, reflecting severe limitations in fundamental self-care tasks such as feeding, transfers, and mobility. Comorbidity audits established that 72.3 percent presented with three or fewer chronic clinical conditions, whereas 27.7 percent carried four or more diagnosed conditions. The most prevalent chronic illnesses recorded at admission included hypertension, stroke, chronic obstructive pulmonary disease, and constipation, the last of which was present in 29.5 percent of the resident cohort.

When examining anthropometric metrics, investigators applied two discrete classification systems to evaluate body mass index distributions. Under conventional Chinese adult guidelines, 26.8 percent of participants were classified as underweight with values below 18.5 kilograms per square meter, 59.1 percent fell within the standard range of 18.5 to 23.9, and 14.1 percent measured at or above 24.0. Applying the newer Chinese health industry standard WS/T 868—2025 tailored specifically for the oldest-old, which sets an appropriate threshold at 22.0 to 26.9 kilograms per square meter, 68.2 percent of participants exhibited values below 22.0, while only 5.5 percent exceeded 26.9. Laboratory markers showed that 13.6 percent presented with serum albumin concentrations below 35 grams per liter, while anemia, defined by sex-specific hemoglobin cutoffs of below 120 grams per liter for men and below 110 grams per liter for women, was documented in 35.0 percent of subjects.

Multivariable logistic regression evaluated clinical characteristics linked to screening-identified malnutrition. Because body mass index contributes points directly to the screening score, models including it were designated as exploratory, while primary models excluded body mass index to avoid incorporation bias. In the primary models, receiving a liquid or semi-liquid diet was consistently associated with higher odds of malnutrition, yielding an adjusted odds ratio of 2.092 in the full sample and 2.223 when excluding the seven well-nourished individuals. Conversely, having four or more comorbid diagnoses was associated with lower odds of malnutrition, yielding adjusted odds ratios of 0.469 and 0.505 across the models.

In exploratory multivariable models that incorporated body mass index, higher anthropometric values showed a persistent inverse association with screening-defined malnutrition. Evaluated continuously, each one-kilogram-per-square-meter increase in body mass index corresponded to an adjusted odds ratio of 0.676, reflecting roughly 32.4 percent lower odds of malnutrition. When grouped by conventional classification categories, the ordinal transition across tiers yielded an odds ratio of 0.227, while categorizing by the older-adult industry standard WS/T 868—2025 yielded an odds ratio of 0.129. Furthermore, restricted cubic spline analysis with knots at the 10th, 50th, and 90th percentiles demonstrated an approximately linear dose-response curve without evidence of non-linear departure, confirming that odds of malnutrition declined steadily across the observed range. The researchers emphasized that these mathematical associations capture cross-sectional relationships within an integrated index rather than independent prospective predictive capacity.

Dietary texture modifications retained robust associations with malnutrition status throughout all exploratory model adjustments. Residents prescribed liquid or semi-liquid diets faced more than twofold higher odds of screening-identified malnutrition, with odds ratios ranging from 2.258 to 2.487 across the various model specifications. The researchers underscored that this statistical association must be interpreted cautiously as confounding by indication rather than proof of a direct harmful effect caused by food consistency. In clinical care, texture modifications are frequently ordered in response to existing swallowing impairments, advanced neurological decline, severe dental loss, or gastrointestinal disturbances, conditions that independently depress caloric intake.

Several critical methodological constraints contextualize the study findings. The cross-sectional architecture precludes any temporal or causal inferences regarding whether low body mass index and diet textures directly cause nutritional deterioration or simply reflect underlying multisystem decline. The exclusion of completely bedridden residents and those in palliative care may have led to an underestimation of true malnutrition rates, while the lack of objective swallowing evaluations, body composition analyses, and muscle mass assessments prevented formal diagnosis under Global Leadership Initiative on Malnutrition criteria. Sampling was restricted to three facilities within Chengdu, meaning geographic and institutional variations could limit broader generalizability. The authors concluded that routine admission screening with instruments like the Mini Nutritional Assessment–Short Form, paired with heightened monitoring and caloric enrichment for individuals receiving modified textures, represents a vital practical strategy for institutional geriatric care.

Subject of Research: Malnutrition risk and associated factors among institutionalized oldest-old adults in Chengdu, China

Article Title: High malnutrition risk burden in the oldest-old in integrated medical and long-term care institutions for older adults: a cross-sectional study

Article References: Gao, J., Pu, H., Yan, Y., Zhong, L., Zheng, W., Wu, Y., Zhang, X., & Yu, X. (2026). High malnutrition risk burden in the oldest-old in integrated medical and long-term care institutions for older adults: a cross-sectional study. Frontiers in Public Health, 14, Article 1886677. https://doi.org/10.3389/fpubh.2026.1886677

Image Credits: AI Generated

DOI: 10.3389/fpubh.2026.1886677

Keywords: malnutrition risk, oldest-old, Mini Nutritional Assessment-Short Form, body mass index, integrated medical and long-term care, Chengdu, geriatric health, texture-modified diets, Barthel Index, multivariable logistic regression, restricted cubic splines, clinical nutrition, functional dependence, incorporation bias, WS/T 868-2025, institutional care