Disability Index (DI)

1. Abstract

The Disability Index (DI), formulated by Zhang et al. (2023), is a comprehensive multidimensional geriatric assessment instrument engineered to quantify functional disability status and evaluate institutional or familial care requirements among community-dwelling and institutionalized older adults. Grounded in the methodological paradigms established by the Frailty Index (FI; Searle et al., 2008) and the Successful Aging Index (SAI; Cosco et al., 2015), the DI operationalizes disability as a multi-tiered deficit accumulation paradigm across four cardinal functional domains: Activities of Daily Living (ADL), Instrumental Activities of Daily Living (IADL), Physical Performance, and Cognitive Function. The instrument comprises 21 items, integrating behavioral self-care tasks, complex instrumental community-living skills, objective biomechanical physical performance indices, and global cognitive status, the latter quantified via the standardized Chinese version of the Mini-Mental State Examination (MMSE).

Standardized on a robust, nationally representative epidemiological cohort comprising individuals aged 65 to 105 years across 22 provinces in China, the DI yields a continuous composite metric spanning 0 to 100, wherein higher scores denote escalated functional impairment and care dependency. Psychometric evaluations demonstrate high internal consistency for the composite DI (α = 0.94), with individual subscale values ranging from acceptable to excellent (α = 0.72 to 0.95). Convergent and divergent validity are empirically corroborated via strong correlations with external criterion measures across four age subgroups (65–79, 80–89, 90–99, and 100–105 years). Known-groups validity demonstrates discriminatory capacity across age strata, while concurrent validity regarding care dependency yields Area Under the Receiver Operating Characteristic Curve (AUC) values predominantly exceeding 0.80 for the overall cohort. The DI represents an operational tool for clinical gerontology, social care planning, health services research, and population-level health surveillance.

2. Keywords

Disability Index, Activities of Daily Living, Instrumental Activities of Daily Living, Cognitive Function, Physical Performance, Geriatric Assessment, Functional Status, Known-Group Validity, Deficit Accumulation, Care Dependency, Older Adults, Psychometrics

3. Authors

The Disability Index was developed through an interdisciplinary collaboration involving scholars in public health, demographic economics, and gerontology:

• Xuxi Zhang, Ph.D. — Center for Healthy Aging and Development Studies, National School of Development, Peking University, Beijing, China.
• Lizi Lin, Ph.D. — Department of Occupational and Environmental Health, School of Public Health, Sun Yat-sen University, Guangzhou, China.
• Xinying Sun, Ph.D. — School of Public Health, Health Science Center, Peking University, Beijing, China.
• Xiaoyan Lei, Ph.D. — Center for Healthy Aging and Development Studies, National School of Development, Peking University, Beijing, China.
• Gordon G. Liu, Ph.D. — National School of Development, Peking University, Beijing, China.
• Hein Raat, M.D., Ph.D. — Department of Public Health, Erasmus University Medical Center, Rotterdam, The Netherlands.
• Yi Zeng, Ph.D. (Corresponding Author) — Center for Healthy Aging and Development Studies, National School of Development, Peking University, Beijing, China; and Center for the Study of Aging and Human Development, Duke University Medical School, Durham, North Carolina, United States. (ORCID: 0000-0001-8354-2945; Email: [email protected]).

4. Purpose

The global acceleration of population aging presents challenges for healthcare infrastructure, long-term care systems, and informal caregiving networks. Traditional geriatric assessments frequently examine functional limitations, cognitive deterioration, and physical impairment through fragmented, unintegrated inventories. This fragmentation can hinder systematic resource allocation, long-term care insurance eligibility assessments, and clinical prognostication. The Disability Index (DI) was conceived to address these methodological limitations by providing a unified, mathematically continuous, and clinically grounded assessment tool that evaluates the full spectrum of disablement among older adults (Zhang et al., 2023).

The primary clinical and epidemiological objective of the DI is to measure the severity of functional disability across community-dwelling, assisted-living, and long-term institutionalized populations, explicitly calibrating this quantification to determine formal and informal care needs. Functional decline in late and very late life is rarely restricted to isolated biological systems; rather, it reflects a complex interplay between musculoskeletal frailty, executive cognitive impairment, sensory degradation, and environmental demands. By synthesizing basic self-care, complex independent living tasks, biomechanical performance capacities, and standardized cognitive profiling into a unified metric, the DI provides health practitioners and social policy planners with a granular assessment of functional vulnerability.

In clinical practice, the DI serves as an evaluative and prognostic screening instrument. It enables multidisciplinary geriatric teams to identify emerging dependencies prior to complete functional breakdown, track trajectories of physical or cognitive recovery post-hospitalization, and tailor targeted rehabilitation protocols. In epidemiological and social policy contexts, the DI standardizes disability surveillance across vast demographic cohorts, accommodating oldest-old (aged 85 and older) and centenarian (aged 100 and older) demographics who are frequently subject to floor and ceiling effects on standard instruments. Furthermore, the scale facilitates equitable qualification algorithms for national long-term care insurance (LTCI) benefits by offering an objective, criterion-referenced index that corresponds to caregiver burden and hours of requisite personal support.

5. Psychological Construct

The construct captured by the Disability Index is multidimensional functional disability, operationalized as an individual’s restricted capacity to execute self-care, domestic maintenance, biomechanical mobility tasks, and cognitive processing required for autonomous living. Rooted in the disablement continuum, the DI conceptualizes functional decrement not merely as the presence of somatic pathology, but as the behavioral and biological manifestation of physiological dysregulation impacting daily life. The DI encompasses four distinct yet intercorrelated operational subdomains:

1. Activities of Daily Living (ADL)

The ADL dimension captures basic, biologically oriented self-maintenance behaviors that are fundamental to survival and physical independence. Grounded in the foundational paradigms of Katz et al. (1963), this subdomain evaluates severe functional impairment through universal bodily requirements, including bathing, dressing, indoor ambulation, personal hygiene/toileting, functional transfers (e.g., moving into and out of a bed or chair), and independent feeding. Deficits within this domain represent end-stage functional dependence, typically necessitating intensive, daily direct physical assistance from formal or informal caregivers.

2. Instrumental Activities of Daily Living (IADL)

The IADL dimension measures complex, higher-order adaptational behaviors necessary for sustaining autonomous community living, reflecting the conceptual framework originated by Lawton and Brody (1969). Unlike basic ADLs, IADL tasks demand the integration of executive cognitive faculties, organizational foresight, socio-environmental adaptation, and fine/gross motor competence. This domain encapsulates activities such as independent meal preparation, handling personal finances and monetary transactions, domestic housekeeping, medication management, telecommunication usage, shopping for necessities, and utilizing public transit or navigating outside the immediate household. Functional loss in this domain typically manifests earlier in the disablement cascade than ADL decline, serving as a clinical indicator of cognitive-motor decline.

3. Physical Performance

The Physical Performance subdomain provides an objective, performance-oriented operationalization of biomechanical capability, neuromotor integrity, musculoskeletal strength, and joint range of motion. Distinct from self-reported functional perceptions, this dimension captures observable somatic capabilities, including lower-extremity functional balance, dynamic postural stability, gait control, chair-rise capability, upper-extremity reaching, manual dexterity, and bilateral hand motor integrity. Impairments in physical performance reflect physiological decrements such as sarcopenia, osteoarticular degeneration, and vestibular-proprioceptive decline, serving as an intermediate biological stage between underlying pathology and overt behavioral disability.

4. Cognitive Function

The Cognitive Function domain recognizes the primary role that neurocognitive decline plays in precipitating functional dependence in older adulthood. Operationalized via the Chinese version of the Mini-Mental State Examination (MMSE; Folstein et al., 1975; Zhang et al., 2020), this construct captures multiple neurocognitive faculties, including spatiotemporal orientation, immediate registration, working memory/attention, verbal recall, expressive language, and visuospatial construction. Cognitive deficits are parameterized into four ordinal severity tiers (severe, moderate, mild, none) and rescaled into continuous deficit equivalents. This reflects the reality that progressive neurodegenerative conditions (such as Alzheimer’s disease and vascular dementia) yield functional care requirements that equal or exceed purely musculoskeletal limitations.

6. Theoretical Framework

The structural and operational architecture of the Disability Index is grounded in two prominent theoretical models in gerontology and psychometrics: the Disablement Process Model proposed by Verbrugge and Jette (1994), and the Cumulative Deficit Model of Frailty articulated by Rockwood and Mitnitski (2007) and operationalized by Searle et al. (2008).

The Disablement Process and the ICF Model

According to Verbrugge and Jette’s Disablement Process Model, the pathway from chronic medical conditions to fundamental life restriction progresses along a distinct sociomedical pathway: Pathology (biochemical or structural somatic abnormalities) → Impairments (dysfunctions and structural abnormalities in specific body systems) → Functional Limitations (restrictions in basic physical and mental actions, such as walking, reaching, or remembering) → Disability (inability or limitation in performing socially defined roles and activities within a sociocultural environment). This aligns directly with the World Health Organization’s International Classification of Functioning, Disability and Health (ICF).

The DI systematically captures this theoretical sequence. The physical performance tasks operationalize direct physiological impairments and immediate functional limitations; the cognitive evaluation measures neurological impairments; and the ADL and IADL domains document manifest disabilities in personal care and social role participation. By integrating these interconnected stages, the DI avoids treating disability as an isolated, all-or-nothing threshold, instead assessing it as an evolving bio-psycho-social spectrum.

The Cumulative Deficit Paradigm

Rather than relying purely on classical psychometric item-response paradigms that treat all scale elements as interchangeable reflective manifestations of a single latent trait, the DI incorporates the mathematical framework of the Frailty Index (FI; Searle et al., 2008) and the Successful Aging Index (SAI; Cosco et al., 2015). Under the cumulative deficit model, health and functional status are viewed through the lens of complex systems biology. As individuals age, molecular, physiological, and behavioral deficits accumulate across distributed biological subsystems.

Within this framework, each individual deficit acts as an indicator within a formative or cumulative index. The DI adopts this methodology by converting ADL, IADL, physical performance, and cognitive impairments into standard deficit proportions, averaging them into a normalized metric from 0 to 100. This synthesis bridges the gap between functional performance status and frailty indices, creating a continuous quantitative assessment calibrated directly to societal and personal care requirements.

7. Validity

The psychometric validation of the Disability Index was conducted using data from the Chinese Longitudinal Healthy Longevity Study (CLHLS), encompassing a geographically and demographically expansive cohort of older adults aged 65 to 105 residing in 22 Chinese provinces (Zhang et al., 2023). Empirical findings substantiate multiple facets of construct, criterion, and discriminative validity:

Convergent and Divergent Validity

To evaluate convergent and divergent validity, the four DI subdomains were correlated against validated alternative external instruments measuring parallel and distinct physiological and psychosocial constructs. Across the total population and within pre-specified age strata (65–79, 80–89, 90–99, and 100–105 years), each DI subdomain exhibited robust, statistically significant correlations with its corresponding construct-specific reference criterion:

• The ADL subdomain demonstrated strong convergent correlations with alternative physical disability scales and caregiver hour estimates.
• The IADL subdomain correlated highly with external measures of independent community functioning and social network engagement.
• The Physical Performance subdomain correlated directly with validated somatic frailty criteria and self-rated physical mobility indices.
• The Cognitive Function subdomain exhibited convergent agreement with external neurocognitive inventories, clinical dementia staging metrics, and proxy-reported cognitive status.

Simultaneously, divergent validity was established: correlations between distinct subdomains (e.g., purely physical performance metrics versus abstract neurocognitive scores) were moderate to low, confirming that the four subdomains capture non-redundant, complementary facets of functional impairment.

Known-Groups Validity

Known-groups validity was tested across demographic and clinical cohorts known to exhibit distinct functional capacities. The full DI and its four subdomains showed clear discriminatory capacity across the age span. The mean DI score increased systematically from the youngest cohort (65–79 years) through octogenarians (80–89 years) and nonagenarians (90–99 years), reaching peak deficit levels among centenarians (100–105 years). The scale also discriminated between older adults with divergent chronic disease burdens, those residing in community settings versus nursing facilities, and individuals with and without documented hospitalizations within the preceding 12 months (all comparisons p < 0.001).

Concurrent and Criterion-Related Validity

The concurrent validity of the full DI was assessed using Area Under the Receiver Operating Characteristic Curve (ROC-AUC) analyses against critical external care indicators, including caregiver burden, formal healthcare service utilization, and heavy informal care dependency. In the total sample, the DI demonstrated concurrent validity, achieving AUC values ≥ 0.80 for predicting comprehensive care dependency.

Within individual age subgroups, discriminatory performance remained generally acceptable, with most subgroup AUCs ≥ 0.70. However, when evaluating the specific criterion of “Heavy Informal Care Burden,” predictive performance in the youngest (65–79 years) and oldest (100–105 years) cohorts yielded AUC values ranging from 0.64 to 0.69, falling slightly below the conventional 0.70 threshold. This divergence likely reflects age-dependent caregiving contexts: among young-old adults, heavy informal care is often triggered by acute catastrophic events (e.g., stroke) not yet captured by chronic functional scales, whereas among centenarians, intensive informal assistance is routinely provided as a normative cultural default regardless of incremental differences in physical test scores.

8. Reliability

The internal consistency reliability of the Disability Index was examined across the complete epidemiological cohort and within distinct age groups (Zhang et al., 2023):

Internal Consistency

Reliability was primarily estimated using Cronbach’s alpha coefficient. Across the full normative sample (N = thousands across 22 provinces), internal consistency metrics were as follows:

• Full Disability Index (21 items): α = 0.94
• Instrumental Activities of Daily Living (IADL) Domain: α = 0.95
• Activities of Daily Living (ADL) Domain: α = 0.87
• Physical Performance Domain: α = 0.72

The composite DI and the IADL and ADL subdomains demonstrate high internal consistency, well exceeding the ≥ 0.80 benchmark recommended for basic research and the ≥ 0.90 threshold for individual-level clinical decision-making. The high alpha for the IADL domain (α = 0.95) indicates high internal coherence among complex instrumental living tasks.

Within specific age subgroups (65–79, 80–89, 90–99, and 100–105 years), internal consistency for the Full DI, the ADL domain, and the IADL domain remained stable, with all subgroup alpha values consistently ≥ 0.70. For the Physical Performance domain, subgroup alphas were moderately lower, varying between 0.66 and 0.68 across age brackets. This moderate reliability is common in brief physical performance batteries that assess diverse motor subsystems (e.g., lower-extremity balance versus fine manual dexterity), where heteroscedasticity across specific physical challenges is expected.

Test-Retest and Longitudinal Reliability Considerations

Test-retest reliability was not formally assessed during the original validation study due to the longitudinal design and sampling intervals of the underlying national survey. As noted by Zhang et al. (2023), establishing short-interval test-retest reliability and intra-class correlation coefficients (ICC) across stable clinical intervals remains an objective for future psychometric research, particularly within sub-acute and post-rehabilitation clinical environments.

9. Factor Analysis

No formal exploratory factor analysis (EFA) or confirmatory factor analysis (CFA) was reported in the primary validation article by Zhang et al. (2023). This absence is rooted in the psychometric and theoretical conceptualization of the instrument. The DI was constructed primarily using a causal/formative indicator framework and the cumulative deficit modeling methodology, rather than a classical reflective latent variable framework.

Formative Indices versus Reflective Latent Models

In standard reflective psychometric scales (such as inventories measuring depression or personality traits), items are assumed to be manifestations of an underlying, unobserved latent construct. In such models, the latent variable causes variations in the observed item responses; consequently, high inter-item correlations and a clear common factor structure (identifiable via CFA with indices such as RMSEA, CFI, and TLI) are mandatory prerequisites for construct validity.

Conversely, the Disability Index, like the Frailty Index (Searle et al., 2008), operates under a formative/composite logic. Within a formative measurement model, the individual indicators accumulate to define and compose the construct of disability. A deficit in one subsystem (e.g., severe cognitive impairment) does not necessarily require the simultaneous presence of a deficit in an unrelated subsystem (e.g., bilateral manual dexterity). While cognitive and physical declines frequently co-occur in late life, an individual can experience severe cognitive deficits while maintaining intact ambulation, or conversely, possess intact cognitive faculties alongside quadriplegia.

Applying standard reflective factor analysis to formative indices can yield misleading interpretations, as low correlations between certain disparate somatic tasks do not indicate structural invalidity. Instead, the construct validity of the DI relies on its multi-tiered deficit coverage, high internal consistency of its subdomains, convergent criterion associations, and its predictive ability regarding functional dependence.

10. Instrument / Measurement Tool

The practical administration, structural parameters, and computational architecture of the Disability Index are summarized below:

• Instrument Name: Disability Index (DI)
• Primary Reference Publication: Zhang, X., Lin, L., Sun, X., Lei, X., Liu, G. G., Raat, H., & Zeng, Y. (2023). Development and validation of the Disability Index among older adults. The Journals of Gerontology: Series A: Biological Sciences and Medical Sciences, 78(1), 111–119.
• Assessment Type: Multidimensional Geriatric Health Index (Composite Formative Index combining self/proxy report, performance-based observational tasks, and standardized cognitive testing).
• Target Population: Community-dwelling, assisted living, and institutionalized older adults across the full geriatric life span (adults aged 65 to 105+ years).
• Total Item Count: 21 composite items spanning four distinct functional subdomains.
• Subdomain Composition:
  • Activities of Daily Living (ADL) — Focuses on basic self-maintenance (e.g., bathing, dressing, indoor mobility, toileting, functional transfers, eating).
  • Instrumental Activities of Daily Living (IADL) — Focuses on independent community living (e.g., meal preparation, financial management, housekeeping, medication adherence, phone usage, shopping, navigation).
  • Physical Performance — Focuses on objective motor/neuromuscular tasks (e.g., balance, chair stands, functional reaching, bilateral manual performance).
  • Cognitive Function — Derived from the Chinese version of the Mini-Mental State Examination (MMSE; 30-item neurocognitive test).
• Response Formats and Scaling:
  • Standard ADL, IADL, and Physical Performance items: Typically scored on a 3-point functional scale: 0 = “Able to perform the activity without difficulty” (no impairment), 0.5 = “Able to perform with some difficulty” (partial impairment), and 1.0 = “Unable to perform the activity independently / Requires complete assistance” (complete impairment).
  • Physical Performance manual tasks (Items 4–6): Scored on a biological performance scale: 0 = “Both hands can perform the activity”, 0.5 = “Only one hand can perform the activity”, 1.0 = “Neither hand can perform the activity”.
  • Cognitive Function domain (MMSE categorization): The total MMSE score (0–30) is transformed into an ordinal deficit score across four clinical severity bands:
    • No cognitive impairment (MMSE 26–30): Deficit score = 0
    • Mild cognitive impairment (MMSE 22–25): Deficit score = 33
    • Moderate cognitive impairment (MMSE 18–21): Deficit score = 67
    • Severe cognitive impairment (MMSE 0–17): Deficit score = 100
• Composite Scoring Algorithm:
  1. Each of the ADL, IADL, and physical performance items is converted to a fractional index value (ranging from 0.0 to 1.0) and multiplied by 100 (yielding item values of 0, 50, or 100).
  2. The cognitive domain score (0, 33, 67, or 100) enters the equation as an individual item index.
  3. The total Disability Index (DI) score is calculated by summing all item scores and dividing by the total number of evaluated items (21 items), yielding a continuous metric ranging from 0 to 100:

  DI Score = (Σ Item Scores / 21)

  Higher scores correspond to elevated functional impairment, heightened frailty, and greater care dependence.

• Administration Time: Approximately 15 to 25 minutes (dependent on respondent cognitive processing speed and mobility status; proxy administration is permissible for ADL/IADL in cases of advanced cognitive impairment).

11. Permissions & Fee and Test Year

The Disability Index was formally published in 2023 by Zhang and colleagues in The Journals of Gerontology: Series A: Biological Sciences and Medical Sciences (Oxford University Press). The tool was developed utilizing public academic research infrastructure supported by Peking University, Duke University Medical Center, and the Chinese Longitudinal Healthy Longevity Study (CLHLS).

Licensing and Commercial Use:

• Commercial Status: Non-commercial.
• Fee: Free of charge for academic, non-commercial research, and public health surveillance purposes.
• Permissions and Access: While the methodological framework, subscales, and scoring algorithms are described in the peer-reviewed literature, the complete, integrated survey instrument and instructions for implementation should be requested from the corresponding author, Prof. Yi Zeng (Center for Healthy Aging and Development Studies, Peking University / Duke University Medical School, Email: [email protected]), or accessed via official CLHLS research protocols. Commercial deployment, automated electronic health record software integration, or proprietary clinical trials require written authorization.

12. References

Cosco, T. D., Stephan, B. C., & Brayne, C. (2015). Operationalising active ageing in the European Social Survey: A successful ageing index. International Psychogeriatrics, 27(6), 907–914. https://doi.org/10.1017/S104161021400276X

Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). “Mini-mental state”: A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12(3), 189–198. https://doi.org/10.1016/0022-3956(75)90026-6

Katz, S., Ford, A. B., Moskowitz, R. W., Jackson, B. A., & Jaffe, M. W. (1963). Studies of illness in the aged: The index of ADL: A standardized measure of biological and psychosocial function. JAMA, 185(12), 914–919. https://doi.org/10.1001/jama.1963.03060120024016

Lawton, M. P., & Brody, E. M. (1969). Assessment of older people: Self-maintaining and instrumental activities of daily living. The Gerontologist, 9(3_Part_1), 179–186. https://doi.org/10.1093/geront/9.3_Part_1.179

Rockwood, K., & Mitnitski, A. (2007). Frailty in relation to the accumulation of deficits. The Journals of Gerontology: Series A: Biological Sciences and Medical Sciences, 62(7), 722–727. https://doi.org/10.1093/gerona/62.7.722

Searle, S. D., Mitnitski, A., Gahbauer, E. A., Gill, T. M., & Rockwood, K. (2008). A standard procedure for creating a frailty index. BMC Geriatrics, 8(1), Article 24. https://doi.org/10.1186/1471-2318-8-24

Verbrugge, L. M., & Jette, A. M. (1994). The disablement process. Social Science & Medicine, 38(1), 1–14. https://doi.org/10.1016/0277-9536(94)90186-0

World Health Organization. (2001). International Classification of Functioning, Disability and Health: ICF. World Health Organization. https://apps.who.int/iris/handle/10665/42407

Zhang, X., Lin, L., Sun, X., Lei, X., Liu, G. G., Raat, H., & Zeng, Y. (2023). Development and validation of the Disability Index among older adults. The Journals of Gerontology: Series A: Biological Sciences and Medical Sciences, 78(1), 111–119. https://doi.org/10.1093/gerona/glac059

Zhang, Z., Gu, D., & Hayward, M. D. (2020). Early life conditions and cognitive impairment among oldest-old Chinese. The Journals of Gerontology: Series B: Psychological Sciences and Social Sciences, 75(4), 834–845. https://doi.org/10.1093/geronb/gby073

13. Items of the Scale