Abstract
Background: A frailty index (FI) can identify individuals with frailty in a population of interest. Previous literature suggests a need for frailty assessment methods for older adults in China and that urbanization may impact frailty status. We used a FI to examine the association between frailty and urbanization as living in a less urbanized area may put older adults at a higher risk frailty and poor healthcare outcomes. Methods: We included adults aged 55 years and older (n = 7695) from the China Health and Nutrition Survey (2018). The FI was based on health outcomes correlating with a deficit score divided by number of health items: robust (<0.08), pre-frail (0.08–0.24), and frail (≥0.25). We used multinomial logistic regression models to examine associations between urbanization tertile (low, medium, and high) and frailty, using our novel FI. We also conducted sub-analyses examining how urbanization level modifies the relationship between frailty status and region of residence, and education and income levels. Results: Living in an area of low versus high urbanization was associated with higher odds of frail versus robust (1.5; 1.2–2.0), and pre-frail versus robust (1.6; 1.4–2.0) status in the fully adjusted model. Generally, higher odds of worse frailty status (e.g., pre-frail or frail) was associated with lower tertiles of urbanization for region, income, and education when compared to the highest urbanization tertile. Conclusions: A FI can help identify specific characteristics that may benefit from individualized interventions to counteract frailty. Living in less urbanized areas was associated with higher odds of pre-frailty and frailty. Inclusion of urbanization level, geographic residence, and social determinants of health in FI development can help identify older adults most at risk of frailty and contribute to individual and policy-level frailty prevention interventions.
Keywords: frailty, geriatrics, health disparities, rural, frailty index
What this paper adds
• To our knowledge, there are few previous works exploring the relationship between frailty and urbanization using a frailty index for Chinese older adults.
• This work enhances the literature by providing a better understanding of the heterogeneity of aging with the use of a frailty index.
Applications of study findings
• In China, a higher odds of pre-frailty and frailty was associated with areas with lower urbanization.
• In future designs of frailty indices, it may be important to consider the inclusion of urbanization as this may be a modifier of frailty status.
• Continuous improvement of frailty index performance is important for encouraging the future development of interventions for frailty mitigation in the areas of most need.
Introduction
Frailty is a state of physical vulnerability that can lead to poor health outcomes and impact individuals along the life span (Rockwood et al., 2011). Frailty is most prevalent in older adults across the globe, leading to an increased risk for morbidity and mortality, including nursing home placement (Lohman et al., 2020). The worldwide prevalence of frailty ranges from 7 to 24%, depending on measurement method (O'Caoimh et al., 2021). Approximately 10% of China’s population is impacted by frailty, with women and individuals 70 years and older most affected (O'Caoimh et al., 2021; Zhou et al., 2023). Vulnerabilities such as having multiple chronic diseases (i.e., three or greater diseases), difficulty with activities of daily living, and low urbanization may increase an older adult’s risk of developing frailty (Vetrano et al., 2019; Yu et al., 2012).
There are various ways to define frailty, using a phenotypic approach (e.g., Fried’s frailty phenotype), an accumulated deficit index (e.g., frailty index), and/or standardized and validated questionnaires (e.g., FRAIL) (Fried et al., 2001; Song et al., 2010; Woo et al., 2015). There is no current gold standard approach to assess frailty, yet a frailty index (FI) can better capture the marked heterogeneity of aging by incorporating a range of variables that well-represent the older adult’s health and functional status, initially developed by Rockwood et al. (Qin et al., 2023; Rockwood & Mitnitski, 2007). A FI presents an opportunity to design a screening that captures the unique characteristics or influencing factors on frailty status in a population, which is important for providing targeted solutions for care needs. China has historically had little research on frailty and frailty interventions, highlighting a need for tools that specifically examine the influencing factors on frailty in Chinese older adults (including geography), as most frailty research has been conducted in the United States and Europe (Ma et al., 2018).
Environmental factors, such as urbanization, may be related to frailty development; however, this relationship is not well-understood. Rural areas in China are associated with higher frailty, lower income and education levels, and less access to healthcare (Fried et al., 2001; Yu et al., 2012). For instance, older adults in rural areas were found to have higher FI scores than those in urban areas, yet a FI does not typically assess how geographical residence or urbanization may impact the frailty status of older adults (Yu et al., 2012; Zeng et al., 2023). Additionally, a previous FI did not incorporate cognitive impairment in scoring, which may impact applicability to an aging population (Zeng et al., 2023). Therefore, an FI provides an opportunity to account for ecological factors that may impact frailty status.
Due to the limitations of previous frailty indices to examine the relationship between urbanization and frailty status, there is a need to better understand the relationship between FI categories and urbanization level to best assess the needs of the population of interest. To address these gaps, we created a FI for older Chinese adults using the China Health and Nutrition Survey (CHNS) in adults from 12 provinces and 3 mega-cities across China and examined associations with urbanization. We used the novel FI to explore the relationship between different frailty states of older adults and urbanization level in China.
Methods
Study Participants
The CHNS started in 1989 with a household survey enrolling participants from 9 provinces but has since expanded and the 2018 survey included participants from 12 provinces and 3 mega-cities. A full explanation of the cohort design and survey procedures are more fully described by Popkin et al. and the China Health and Nutrition Survey website, respectively (Acknowledgements, 2004; Howard et al., 2021; Popkin et al., 2010). This study was approved by the institutional review board from the National Institute for Nutrition and Food Safety, the Chinese Center for Disease Control Prevention, and the University of North Carolina at Chapel Hill, and all participants provided a written informed consent.
For the analytical sample, adults aged 55 years and older were enrolled in the 2018 survey (n = 7695 participants). The survey included questionnaires and standardized data on cognitive and physical function related to frailty, which were asked only of adults 55 and over. We excluded 265 participants who were missing more than half (17 of 37 items or more) of the components of the FI. We also excluded 251 participants who were missing one or more covariates. Our final analytical sample consisted of 7179 participants with an average of 98.8% of FI components available.
Urbanization
Urbanization was measured at the “community level using a multicomponent and validated continuous index, derived from 12 components to capture the complexity of urbanization (Jones-Smith & Popkin, 2010). Components included population density, economic activity, traditional markets, modern markets, transportation infrastructure, sanitation, communications, housing, education, diversity, health infrastructure, and social services.” For ease of interpretation, this index was categorized as low, moderate, or high urbanization, based on tertiles of the urbanization index.
Frailty Index Development
We calculated a 37-item FI based on the concept that an accumulation of deficits leads to a higher degree of frailty (Rockwood & Mitnitski, 2007). The FI acts as a marker of health status while acknowledging the heterogeneity of aging across the life span, for various health categories. Similar to the initial FI by Rockwood et al., our FI is composed of various different domains of health, physical function, and quality of life measures (Rockwood et al., 2005).
Our FI included components that are typically considered in routine geriatric assessment, across cultures, and can be referenced in Table 3 (Kim et al., 2018; Rockwood & Mitnitski, 2007; Yu et al., 2012). We included measures of cognitive function, quality of life, body anthropometrics, smoking status, and physical function measures. We matched previously collected health, physical function, and quality of life measures in the CHNS database to deficits used in the initial FI by Rockwood et al., which requires at least 30 items that capture specific domains that can contribute to frailty (e.g., physical function, medical history, and medication use) (Rockwood et al., 2005). Deficits were scored based on standardized health outcomes (Kim et al., 2018; Rockwood & Mitnitski, 2007). The overall score was determined based on the sum of the accumulated deficits, divided by the total possible deficits. Based on previously published frailty cut points, participants were considered to be robust (<0.08), pre-frail (0.08–0.24), or frail (≥0.25) (Rockwood & Mitnitski, 2007). Specific components of the FI are more explicitly defined as follows.
Table 3.
Descriptive Statistics for Selected Variables Used to Calculate the Frailty Index.
| Category | Variable | Scoring deficit points (good = 0, fair = 0.5, and poor = 1) | Lowest tertile urbanization index, N (%) | Moderate tertile urbanization index, N (%) | High tertile urbanization index, N (%) |
|---|---|---|---|---|---|
| Self-reported comorbidities | Cancer | Yes = 1 | 33 (0.5) | 51 (0.7) | 53 (0.7) |
| High blood pressure | Yes = 1 | 837 (11.7) | 969 (13.5) | 1069 (14.9) | |
| Diabetes | Yes = 1 | 147 (2.1) | 233 (3.3) | 291 (4.1) | |
| Myocardial infarction | Yes = 1 | 36 (0.5) | 41 (0.6) | 39 (0.5) | |
| Apoplexy | Yes = 1 | 63 (0.9) | 81 (1.1) | 82 (1.1) | |
| Bone fracture | Yes = 1 | 128 (1.9) | 144 (2.1) | 166 (2.4) | |
| Gastrointestinal reflux | Yes = 1 | 90 (1.4) | 170 (2.6) | 126 (2.0) | |
| Asthma | Yes = 1 | 44 (0.6) | 37 (0.5) | 40 (0.6) | |
| Function | Do you have difficulty bathing? | Yes = 1 | 74 (1.0) | 67 (0.9) | 79 (1.1) |
| Do you have difficulty eating alone? | Yes = 1 | 30 (0.4) | 25 (0.4) | 21 (0.3) | |
| Do you have difficulty putting on clothes? | Yes = 1 | 55 (0.8) | 53 (0.7) | 58 (0.8) | |
| Do you have difficulty combing hair? | Yes = 1 | 45 (0.6) | 36 (0.5) | 47 (0.7) | |
| Do you have difficulty using toilet? | Yes = 1 | 50 (0.7) | 43 (0.6) | 50 (0.7) | |
| Do you have difficulty shopping? | Yes = 1 | 119 (1.7) | 109 (1.5) | 115 (1.6) | |
| Do you have difficulty cooking? | Yes = 1 | 133 (1.9) | 122 (1.7) | 117 (1.6) | |
| Do you have difficulty using public transportation? | Yes = 1 | 255 (3.6) | 205 (1.8) | 184 (1.7) | |
| Do you have difficulty managing money? | Yes = 1 | 144 (2.0) | 126 (5.3) | 121 (5.1) | |
| Do you have difficulty using the telephone? | Yes = 1 | 190 (2.7) | 145 (2.0) | 93 (1.3) | |
| Do you have difficulty walking 200 m? | Yes = 1 | 110 (1.5) | 93 (1.3) | 87 (1.2) | |
| Do you have difficulty walking across the room? | Yes = 1 | 84 (1.2) | 58 (0.8) | 60 (0.8) | |
| Do you have difficulty sitting for 2 hours? | Yes = 1 | 123 (1.7) | 116 (1.6) | 112 (1.6) | |
| Do you have difficulty standing up after sitting? | Yes = 1 | 157 (2.2) | 144 (2.0) | 136 (1.9) | |
| Do you have difficulty climbing 1 flight of stairs? | Yes = 1 | 149 (2.1) | 120 (1.7) | 125 (1.8) | |
| Do you have difficulty climbing a few steps without a pause? | Yes = 1 | 234 (3.3) | 209 (2.9) | 212 (3.0) | |
| Do you have difficulty lifting a 5-kilogram bag? | Yes = 1 | 335 (4.7) | 437 (6.2) | 427 (6.0) | |
| Do you have difficulty squatting? | Yes = 1 | 250 (3.5) | 269 (3.8) | 269 (3.8) | |
| Quality of life | Number of days you have been dissatisfied with you sleep? | ≤1 days = 0 | 820 (11.4) | 666 (9.3) | 643 (9.0) |
| >1 day = 1 | |||||
| Perceived stress score/anxiety (based on Generalized Anxiety Disorder-2 screening) | No anxiety (0) | 59 (0.9) | 71 (1.0) | 102 (1.5) | |
| Mild anxiety (0.5) | 1420 (20.3) | 1382 (19.7) | 1379 (19.7) | ||
| Anxiety (1) | 846 (12.1) | 879 (12.5) | 870 (12.4) | ||
| Do you have as much pep as last year? | No different, agree, strongly agree = 0 | 1514 (21.4) | 1293 (18.3) | 1187 (16.8) | |
| Disagree, strongly disagree = 1 | |||||
| Are you older and things are better? | No different, agree, strongly agree = 0 | 1353 (19.2) | 1141 (16.2) | 1095 (15.6) | |
| Disagree, strongly disagree = 1 | |||||
| Are you as happy as when you were younger? | No different, agree, strongly agree = 0 | 1530 (21.6) | 1283 (18.1) | 1179 (16.7) | |
| Disagree, strongly disagree = 1 | |||||
| How do you rate your current life? | Excellent, good = 0 | 1260 (17.6) | 1523 (21.3) | 1567 (21.9) | |
| Fair = 0.5 | 988 (13.8) | 732 (10.2) | 733 (10.3) | ||
| Poor = 1 | 142 (2.0) | 123 (1.7) | 79 (1.1) | ||
| How do you rate your current health status? | Excellent, good = 0 | 949 (13.3) | 1181 (16.6) | 1237 (17.4) | |
| Fair = 0.5 | 1100 (15.4) | 941 (13.2) | 937 (13.2) | ||
| Poor = 1 | 331 (4.6) | 255 (3.6) | 195 (2.7) | ||
| Cognition | Cognitive status based on the Telephone Interview for Cognitive Status (TICS) | Cognitive impairment (TICS <6 points) = 1 | 329 (4.6) | 257 (3.6) | 195 (2.7) |
| Mild cognitive impairment (>6 to <11 points) = 0.5 | 551 (7.7) | 475 (6.6) | 453 (6.3) | ||
| Normal cognition (>11 points) = 0 | 1524 (21.2) | 1653 (23.0) | 1742 (24.3) | ||
| Objective measurements | Body mass index (kg/m2) | Underweight (<18.5 kg/m2) = 1 | 121 (2.0) | 69 (1.1) | 59 (0.9) |
| Normal weight (18.5–23.9 kg/m2) = 0 | 982 (15.2) | 880 (13.6) | 866 (13.4) | ||
| Overweight (24–27.5 kg/m2) or obesity (≥27.5 kg/m2) = 1 | 1087 (16.8) | 1223 (18.9) | 1172 (18.2) | ||
| Hospitalization within last 4 weeks | Yes = 1, No = 0 | 38 (0.5) | 37 (0.5) | 27 (0.4) | |
| Smoking | Classified as never, former, or current smoker | Never = 0 | 1375 (19.2) | 1406 (19.6) | 1541 (21.5) |
| former = 0.5 | 453 (6.3) | 476 (6.6) | 434 (6.1) | ||
| Current smoker = 1 | 575 (8.0) | 503 (7.0) | 415 (5.8) |
Cognitive Function
We assessed cognitive function using the Telephone Interview for Cognitive Status (TICS-27) (Abdulrahman et al., 2022; Fong et al., 2009; Hlávka et al., 2022). Based on literature and cross walks with Mini-Mental Status Exam, we determined cut-offs for dementia and mild cognitive impairment: dementia ≤6 points; mild cognitive impairment >6 to ≤11 points; and normal cognition >11 points (Crimmins et al., 2011; Hlávka et al., 2022).
Quality of Life
Quality of life was assessed using the questions, “Do you have as much pep as last year?”, “Are you older and things are better?”, and “Are you as happy as when you were younger?” Participants were also asked to assess their current health status by rating their health as excellent, good, fair, or poor.
Anxiety
Anxiety was assessed using two questions similar to the questions validated for Generalized Anxiety Disorder-2 screening. A score of ≥3 was consistent with having anxiety (Kroenke et al., 2007).
Smoking
Smoking status was defined as “never a smoker,” “former smoker,” or “current smoker.”
Body Anthropometrics
BMI categories were based on the World Health Organization’s recommendations for Asian adults: underweight (<18.5 kg/m2), normal weight (18.5–23 kg/m2), overweight (23–27.5 kg/m2), or obesity (≥27.5 kg/m2) (Lynch et al., 2023).
Physical Function
Physical function variables (e.g., climbing one set of stairs and lifting 5-kilogram bag), activities of daily living, and instrumental activities of daily living were self-reported by each individual. A participant’s illnesses/comorbidities were defined by self-report, as received by a doctor.
Covariates
Covariates included age, sex, education, region, and household income. Sex and education were all self-reported in survey questionnaires. Age was categorized in five-year age groups, including 55–60, 60–65, 65–70, 70–75, and ≥75 years and education level was categorized as not having completed primary school (ages 6 through 15 years), completed primary school, or completed post-primary school education. Household income was based on individual and household questionnaires and assets and inflated to 2018 yuan currency. Region was categorized as Northern, Central, Southern, and mega-cities of China. Given the extremely skewed nature of household income and to account for potential non-linearity in the association with the FI, income was categorized in tertiles.
Statistical Methods
Descriptive statistics were calculated overall and by urbanization tertile. Chi-squared test and unadjusted analysis of variance test were used to test for differences in percentages and means, respectively, between urbanization tertiles for both covariates and FI components. Multinomial logistic regression models were used to explore how the FI was associated with urbanization. A minimally adjusted multinomial logistic regression model controlled for age and sex, and a fully adjusted model additionally adjusted for education, income, and region. To illustrate these associations, the estimates of adjusted odds ratios and 95% confidence intervals were provided. Odds ratios were generated for low and moderate urbanization tertiles compared to high urbanization. Sub-group analyses were conducted across education level, income level, and region of residence, stratifying by urbanization tertile. All analysis was conducted using SAS Version 9.4.
Results
The cohort is approximately half men with an average age of 66.7 ± 8 years across the three tertiles of urbanization (Table 1). Most participants were classified as pre-frail across urbanization tertiles (low 74.0%, moderate 69.9%, and high 66.1%).
Table 1.
Descriptive Statistics by Urbanization Index Tertile.
| All sample | Low urbanization tertile | Moderate urbanization tertile | High urbanization tertile | p-value a | |
|---|---|---|---|---|---|
| N = 7179 | N = 2404 | N = 2385 | N = 2390 | ||
| Age, years (mean (standard deviation)) | 66.7 (8.0) | 66.0 (7.6) | 66.6 (8.0) | 67.4 (8.4) | <0.0001 |
| Participant (%) in each age group | |||||
| 55–59 | 21.8 | 23.8 | 21.8 | 19.8 | <0.0001 |
| 60–64 | 25.6 | 26.2 | 26.0 | 24.7 | |
| 65–69 | 22.6 | 23.2 | 22.9 | 21.8 | |
| 70–74 | 14.3 | 14.1 | 14.0 | 15.0 | |
| 75+ | 15.6 | 12.7 | 15.4 | 18.7 | |
| Men (%) | 47.2 | 47.4 | 47.7 | 46.5 | 0.6929 |
| Highest education level completed (%) | |||||
| Less than primary school | 11.0 | 18.7 | 9.4 | 4.9 | <0.0001 |
| Primary school | 60.7 | 70.7 | 63.8 | 47.6 | |
| High school and beyond | 28.3 | 10.6 | 26.8 | 47.6 | |
| Married b (%) | 85.0 | 85.3 | 85.1 | 84.5 | 0.7505 |
| Currently working c (%) | 41.6 | 67.6 | 38.7 | 18.2 | <0.0001 |
| Region of residence (%) | |||||
| North | 12.3 | 17.1 | 6.4 | 13.4 | <0.0001 |
| Central | 34.7 | 43.1 | 34.5 | 26.5 | |
| South | 30.2 | 29.2 | 28.4 | 32.9 | |
| Mega-cities | 22.8 | 10.6 | 30.7 | 27.2 | |
| Total gross household income (K) per capita inflated to 2018 (mean (standard deviation)) | 28.9 (58.1) | 22.9 (83.6) | 29.1 (37.0) | 34.6 (41.1) | <0.0001 |
| Urbanization index d (mean (standard deviation)) | 72.6 (17.6) | 51.6 (6.7) | 74.3 (6.1) | 91.9 (5.1) | <0.0001 |
| Frailty (%) | |||||
| Robust (<0.08) | 17.6 | 12.4 | 17.6 | 22.8 | <0.0001 |
| Pre-frail (0.08-<0.25) | 70.0 | 74.0 | 69.9 | 66.1 | |
| Frail (≥0.25) | 12.4 | 13.6 | 12.5 | 11.2 | |
aFor the comparisons across urbanization tertiles: analysis of variance was used for means of variables and the chi-squared test was used for percentages. p-value <.05 was considered statistically significant.
bMarried versus not-married.
cWorking versus not-working.
dUrbanization index score 0–120, higher score indicating higher level of urbanization.
In the unadjusted model, low and moderate urbanization tertiles were associated with higher odds of pre-frailty and frailty than the highest tertile of urbanization (Figure 1). Similar trends were present in the fully adjusted model, but only reached statistical significance in the pre-frail versus robust groups in the low and moderate urbanization tertiles and only in the frail versus robust group in the low urbanization tertile (Figure 1). For both the unadjusted and adjusted models, odds of frailty increased with each age group, with highest odds of frailty in the 75+ years age group (Table 2).
Figure 1.
Minimally and fully adjusted multinomial logistic regression models testing the association between frailty status (robust, pre-frail, and frail) and tertiles. Minimally adjusted models controlling for age, sex, and urbanization and fully adjusted models controlling for age, sex, urbanization, education, income, and region.
Table 2.
Minimally and Fully Adjusted Multinomial Logistic Regression Models Testing the Association Between Frailty Status (Robust, Pre-frail, and Frail) and Urbanization Tertiles. a
| Odds ratio | ||||||
|---|---|---|---|---|---|---|
| Minimally adjusted model a | Fully adjusted model a | |||||
| Frail versus pre-frail | Frail versus robust | Pre-frail versus robust | Frail versus pre-frail | Frail versus robust | Pre-frail versus robust | |
| OR (95% CL) | OR (95% CL) | OR (95% CL) | OR (95% CL) | OR (95% CL) | OR (95% CL) | |
| Low versus high urbanization | 1.4 (1.1, 1.6) | 3.0 (2.4, 3.8) | 2.2 (1.9, 2.6) | 0.9 (0.8, 1.2) | 1.5 (1.2, 2.0) | 1.6 (1.4, 2.0) |
| Moderate versus high urbanization | 1.2 (1.0, 1.4) | 1.7 (1.4, 2.1) | 1.4 (1.2, 1.6) | 1.0 (0.8, 1.2) | 1.2 (1.0, 1.5) | 1.2 (1.0, 1.4) |
| Age 60–64 versus 55–59 | 1.3 (1.0, 1.8) | 1.8 (1.3, 2.5) | 1.4 (1.2, 1.6) | 1.2 (0.9, 1.6) | 1.5 (1.1, 2.1) | 1.3 (1.1, 1.5) |
| Age 65–69 versus 55–59 | 1.9 (1.4, 2.5) | 2.9 (2.1, 4.0) | 1.6 (1.3, 1.9) | 1.6 (1.2, 2.2) | 2.2 (1.6, 3.1) | 1.4 (1.2, 1.7) |
| Age 70–74 versus 55–59 | 3.5 (2.6, 4.8) | 6.2 (4.4, 8.7) | 1.7 (1.4, 2.2) | 3.1 (2.3, 4.1) | 4.8 (3.4, 6.7) | 1.6 (1.3, 1.9) |
| Age 75+ versus 55–59 | 9.4 (7.1, 12.3) | 32.1 (22.5, 45.7) | 3.4 (2.6, 4.5) | 7.5 (5.7, 9.9) | 22.5 (15.7, 32.4) | 3.0 (2.3, 3.9) |
| Male versus female | 0.9 (0.7, 1.0) | 1.1 (0.9, 1.3) | 1.3 (1.1, 1.5) | 1.0 (0.9, 1.2) | 1.5 (1.2, 1.8) | 1.4 (1.2, 1.6) |
| Less than primary school versus high school and beyond | 2.4 (1.8, 3.1) | 4.8 (3.2, 7.0) | 2.0 (1.5, 2.7) | |||
| Primary versus high school and beyond | 1.2 (1.0, 1.5) | 1.7 (1.4, 2.2) | 1.5 (1.3, 1.7) | |||
| Low versus high income | 1.8 (1.4, 2.2) | 2.8 (2.1, 3.6) | 1.6 (1.3, 1.8) | |||
| Moderate versus high income | 1.4 (1.2, 1.8) | 1.8 (1.4, 2.3) | 1.2 (1.1, 1.4) | |||
| Central versus Southern | 0.7 (0.6, 0.9) | 0.8 (0.6, 0.9) | 0.7 (0.6, 0.9) | |||
| Mega cities versus Southern | 1.0 (0.8, 1.3) | 0.9 (0.7, 1.0) | 1.0 (0.8, 1.3) | |||
| Northern versus Southern | 0.7 (0.5, 1.0) | 0.7 (0.5, 0.8) | 0.7 (0.5, 1.0) | |||
aMinimally adjusted models controlling for age, sex, and urbanization and fully adjusted models controlling for age, sex, education, income, urbanization, and region.
In the fully adjusted model (Table 2), central and northern cities compared to southern region had decreased odds of pre-frailty development; however, when stratifying region by the urbanization level in our sub-analyses, the northern region of China had the highest odds of worse frailty status (Supplemental Table 1). Lower education level (less than primary school) and low income were also associated with higher odds of pre-frailty and frailty in the fully adjusted model. Yet completing primary school and beyond in lower levels of urbanization was associated with lower odds of worse frailty status in our stratified sub-analysis (Supplemental Table 3). Generally, there were higher odds of worse frailty status (e.g., pre-frail or frail) in the lower tertiles of urbanization for region, income, and education when compared to the highest urbanization tertile (Supplemental Tables 1-3).
We found differences in some components of the FI across urbanization tertiles (Table 3). Higher proportions of participants in low urbanization than those in areas with moderate to high urbanization reported more difficulty with transportation (3.6% (low) versus 1.8% (moderate) and 1.7% (high)) and using the telephone (2.7% vs. 2.0% and 1.3%). Participants in areas of lower urbanization also reported worse self-reported health status (4.6% vs. 3.6% and 2.7%) and cognitive impairment (dementia: 4.6% vs. 3.6% and 2.7%; mild cognitive impairment: 7.7% vs. 6.6% and 6.3%). Proportions across urbanization levels were similar for mild (range 19.7–20.3%) and full anxiety (range 12.1–12.5%). There was a higher proportion of participants in less urbanized areas (vs. moderate and high urbanization) that reported “more pep” (21.4% vs. 18.3% and 16.8%), improved life in older age (19.2% vs. 16.2% and 15.6%), and similar happiness level as in younger age (21.6% vs. 18.1% and 16.7%).
Discussion
Our study describes the development of a FI to identify older adults at risk of frailty development and the association between frailty status and urbanization levels in China. Our study adds to the limited literature on the development of a FI for older adults in China and provides insight into the relationship between urbanization level and frailty prevalence. Using a FI, we found that areas of lower urbanization are associated with increased odds of having pre-frailty and frailty and is observed across varying places of residence, education, and income levels. Our study is unique as it describes the odds of a worse frailty status for specific social determinants of health (e.g., income, education, urbanization, and residence), which may suggest that FI development should include individual and ecological factors. Identifying these environmental influences on frailty status can lend to the development of individual and system-level interventions to optimize “aging in place” for older adults in China.
Our findings of higher odds of frailty associated with low urbanization are similar to that from Fan et al. (2020) and Zeng et al. (2023). in China, finding higher rates of pre-frailty and frailty in rural versus urban China (Fan et al., 2020; Zeng et al., 2023). Like Zeng et al., we also observed higher odds of pre-frailty and frailty in Northern China, as this is typically an area of lower urbanization and a more rural landscape than Southern China (Guo & Li, 2024; Zeng et al., 2023). Interestingly, Fan et al. observed higher FI scores for younger than older adults using a 28-item FI with a sample of 30–79 years old from the China Kadoorie Biobank. This difference may be due to multiple reasons when compared to our study: (1) Fan et al.’s sample included predominantly robust participants, where our study participants exhibited predominantly pre-frailty status; (2) Fan et al.’s FI included 9 less items than ours; and (3) of Fan et al.’s 28 items, 14 were comorbidities versus our 8 comorbidities out of 37 items. These differences may suggest that different FI components and/or weighting are needed for different age groups to best reflect the heterogeneity of aging. Additionally, older adults with frailty in less urbanized areas of China reported higher rates of working, which may be due to jobs in less urbanized areas lacking pensions and required retirement ages (National Research Council Panel on Policy & Data Needs to Meet the Challenge of Aging in, 2012).
Our findings lend to speculation that differences in FI scores by urbanization could be secondary to healthcare access and other social determinants of health. For example, we observed less access to transportation in less urbanized areas in China and generally higher odds of frailty across all levels of education and income in our sub-analyses for the lowest urbanization tertile (Supplemental Tables 2 and 3). Lower access to healthcare in rural areas in China is associated with worse health outcomes and increased frailty prevalence (Gu et al., 2019; Zhou et al., 2023). This is further supported by older adults with frailty in rural China having a higher hazard ratio for all-cause mortality than those in urban areas (Fan et al., 2020). Therefore, a FI may be most helpful in capturing healthcare disparities, contrary to other frailty screenings, due to its flexibility to include social determinants of health and residence in FI scoring.
Our conflicting quality of life findings as part of the FI further highlight the heterogeneity of aging. Older adults in less urbanized areas reported more “pep,” happiness, and “better” life in older age, while yet reporting worse quality of life than their more urban counterparts. In contrast, a systematic review by Lobanov-Rostovsky and a cross-sectional study by You et al. observed positive quality of life measures in more urbanized areas (Lobanov-Rostovsky et al., 2023; You et al., 2019). Lobanov-Rostovsky speculates this is due to better access to healthcare in more urbanized areas. Notably, our quality of life screenings did not include a measure of pain, which was a focus of the three-level EuroQol quality of life screening by You et al. Quality of life may be indirectly impacted by healthcare access, as we observed higher proportions of participants in areas of low urbanization reporting difficulty with transportation, using telephone, and lower self-reported health status, compared to areas with moderate and high urbanization.
Our study is cross-sectional (2018), so we are unable to examine evolution of FI scores and morbidity and mortality risk changes as urbanization changes over time. However, we employed methods similar to Song et al., where earlier cross-sectional FI scores correlated with mortality (Song et al., 2010). We did find higher rates of pre-frailty and frailty for community-dwelling older adults than in a systematic review by Zhou et al., which may limit applicability of our findings as a result of our sample (Zhou et al., 2023). However, Zhou et al. compared various frailty assessments in addition to frailty indices. Most notably, our FI is limited by the accuracy of the individual indices to capture the individual’s functional status due to possible self-report bias. From a biopsychosocial aging standpoint, our FI did not include biomarkers such as C-reactive protein, IL-6 and hemoglobin A1c, TNF-alpha, and other cardiovascular metabolic markers which have been correlated with frailty and worse health outcomes in rural areas (Aizawa, 2019; Purser et al., 2008; Saedi et al., 2019; Zhang & Crimmins, 2019). Including biomarkers may be important for acknowledging healthcare disparities with a FI and applying directed health interventions to address specific disease processes. Lastly, and ideally, we would have controlled for healthcare access as a confounder, but this was not a part of the dataset.
The strengths in our study lie in our use of a FI, a measure that is not often used in China. As much of the literature on rural versus urban health outcomes relies on self-reported measures, the strength of an FI lies in incorporation of multiple subjective and objective measures to track changes over time (Li et al., 2022; Lin et al., 2023; Yuan et al., 2023). Our FI components include 37 variables, which is comparable to and exceeds published frailty indices (Fan et al., 2020; Qin et al., 2023; Song et al., 2010; Wang et al., 2017). While number of variables alone does not improve the performance of an FI, our inclusion of quality of life measures and cognitive performance is unique and imperative to fully capturing influential factors of functional status and frailty development. Because we did not examine the performance of our FI against other Chinese frailty screenings, this is a necessary step for validation.
Conclusion
A FI has the potential to be a useful tool for identifying health vulnerability for older adults in China. Our study shows the ability of a FI to discern differences in frailty states for older adults at different levels of urbanization and the importance of including social determinants of health variables in future FI development to identify individuals most at risk of frailty. Future efforts should include comparisons of current frailty indices in the population of interest and relation to healthcare outcomes to best identify potential targets for interventions that mitigate pre-frailty, frailty, and poor health outcomes.
Supplemental Material
Supplemental Material for The Association Between Urbanization and Frailty Status in China by Hillary B. Spangler, David H. Lynch, Annie Green Howard, Hsiao-Chuan Tien, Shufa Du, Bing Zhang, Huijun Wang, Penny Gordon Larsen, and John A. Batsis in Journal of Applied Gerontology
Author Contributions: All authors listed participated in the study’s design, data analysis, and manuscript preparation.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research uses data from the China Health and Nutrition Survey (CHNS). We appreciate the grant funding from the National Institute for Health (NIH), the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) for R01HD30880, the National Institute on Aging (NIA) for R01AG065357, the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) for R01DK104371, the National Institute of Heart, Lung, and Blood for R01HL108427, the NIH Fogarty grant D43 TW009077 since 1989, and the China-Japan Friendship Hospital, Ministry of Health for support for CHNS 2009, the Chinese National Human Genome Center at Shanghai since 2009, and the Beijing Municipal Center for Disease Prevention and Control since 2011 (Chinal Health and Nutrition Survey, 2024). We are grateful for support from the Carolina Center for Population Aging and Health (P30AG066615), the National Institute for Nutrition and Health, the China Center for Disease Control and Prevention, the Beijing Municipal Center for Disease Control and Prevention, and the Chinese National Human Genome Center at Shanghai. (Chinal Health and Nutrition Survey, 2024) We are also grateful for National Opinion Research Center support (P30DK056350-23).
Sponsor’s Role: Dr Batsis participated in the study’s design, data analysis, and manuscript preparation.
IRB Approval: This study was approved by the institutional review board from the National Institute for Nutrition and Food Safety, the Chinese Center for Disease Control Prevention, and the University of North Carolina at Chapel Hill (20-0025), and all participants provided a written informed consent.
Supplemental Material: Supplemental material for this article is available online.
ORCID iDs
Hillary B. Spangler https://orcid.org/0000-0003-4027-4113
John A. Batsis https://orcid.org/0000-0002-0845-4416
References
- Abdulrahman H., Jansen E., Hoevenaar-Blom M., van Dalen J. W., van Wanrooij L. L., van Bussel E., van Gool W. A., Richard E., Moll van Charante E. P. (2022). Diagnostic accuracy of the telephone interview for cognitive status for the detection of dementia in primary care. The Annals of Family Medicine, 20(2), 130–136. 10.1370/afm.2768 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Aizawa T. (2019). Urban developments and health: Evidence from the distributional analysis of biomarkers in China. SSM - Population Health, 8, Article 100397. 10.1016/j.ssmph.2019.100397 [DOI] [PMC free article] [PubMed] [Google Scholar]
- China Health and Nutrition Survey . (2004). Acknowledgements. https://www.cpc.unc.edu/projects/china [Google Scholar]
- Crimmins E. M., Kim J. K., Langa K. M., Weir D. R. (2011). Assessment of cognition using surveys and neuropsychological assessment: The health and retirement study and the aging, demographics, and memory study. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 66(Suppl 1), i162–i171. 10.1093/geronb/gbr048 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fan J., Yu C., Guo Y., Bian Z., Sun Z., Yang L., Chen Y., Du H., Li Z., Lei Y., Sun D., Clarke R., Chen J., Chen Z., Lv J., Li L., China Kadoorie Biobank Collaborative Group . (2020). Frailty index and all-cause and cause-specific mortality in Chinese adults: A prospective cohort study. The Lancet Public Health, 5(12), e650–e660. 10.1016/s2468-2667(20)30113-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fong T. G., Fearing M. A., Jones R. N., Shi P., Marcantonio E. R., Rudolph J. L., Yang F. M., Kiely D. K., Inouye S. K. (2009). Telephone interview for cognitive status: Creating a crosswalk with the mini-mental state examination. Alzheimer’s and Dementia: The Journal of the Alzheimer’s Association, 5(6), 492–497. 10.1016/j.jalz.2009.02.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fried L. P., Tangen C. M., Walston J., Newman A. B., Hirsch C., Gottdiener J., Seeman T., Tracy R., Kop W. J., Burke G., McBurnie M. A., Cardiovascular Health Study Collaborative Research Group . (2001). Frailty in older adults: Evidence for a phenotype. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 56(3), M146–M156. 10.1093/gerona/56.3.m146 [DOI] [PubMed] [Google Scholar]
- Gu J., Chen H., Gu X., Sun X., Pan Z., Zhu S., Young D. (2019). Frailty and associated risk factors in elderly people with health examination in rural areas of China. Iranian Journal of Public Health, 48(9), 1663–1670. [PMC free article] [PubMed] [Google Scholar]
- Guo Y., Li X. (2024). Regional inequality in China’s educational development: An urban-rural comparison. Heliyon, 10(4), Article e26249. 10.1016/j.heliyon.2024.e26249 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hlávka J. P., Yu J. C., Lakdawalla D. N. (2022). Crosswalk between the mini-mental state examination and the telephone interview for cognitive status (TICS-27/30/40). Alzheimer’s and Dementia: The Journal of the Alzheimer’s Association, 18(11), 2036–2041. 10.1002/alz.12569 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Howard A. G., Attard S. M., Herring A. H., Wang H., Du S., Gordon-Larsen P. (2021). Socioeconomic gradients in the Westernization of diet in China over 20 years. SSM - Population Health, 16, Article 100943. 10.1016/j.ssmph.2021.100943 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jones-Smith J. C., Popkin B. M. (2010). Understanding community context and adult health changes in China: Development of an urbanicity scale. Social Science & Medicine, 71(8), 1436–1446. 10.1016/j.socscimed.2010.07.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kim D. H., Schneeweiss S., Glynn R. J., Lipsitz L. A., Rockwood K., Avorn J. (2018). Measuring frailty in medicare data: Development and validation of a claims-based frailty index. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 73(7), 980–987. 10.1093/gerona/glx229 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kroenke K., Spitzer R. L., Williams J. B., Monahan P. O., Löwe B. (2007). Anxiety disorders in primary care: Prevalence, impairment, comorbidity, and detection. Annals of Internal Medicine, 146(5), 317–325. 10.7326/0003-4819-146-5-200703060-00004 [DOI] [PubMed] [Google Scholar]
- Li H., Zeng Y., Gan L., Tuersun Y., Yang J., Liu J., Chen J. (2022). Urban-rural disparities in the healthy ageing trajectory in China: A population-based study. BMC Public Health, 22(1), Article 1406. 10.1186/s12889-022-13757-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lin J., Yang D., Zhao X., Xie L., Xiong K., Hu L., Xu Y., Yu S., Huang W., Gong N., Liang X. (2023). The action logic of the older adults about health-seeking in South Rural China. BMC Public Health, 23(1), Article 2487. 10.1186/s12889-023-17314-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lobanov-Rostovsky S., He Q., Chen Y., Liu Y., Wu Y., Liu Y., Venkatraman T., French E., Curry N., Hemmings N., Bandosz P., Chan W. K., Liao J., Brunner E. J. (2023). Growing old in China in socioeconomic and epidemiological context: Systematic review of social care policy for older people. BMC Public Health, 23(1), Article 1272. 10.1186/s12889-023-15583-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lohman M. C., Sonnega A. J., Resciniti N. V., Leggett A. N. (2020). Frailty phenotype and cause-specific mortality in the United States. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 75(10), 1935–1942. 10.1093/gerona/glaa025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lynch D. H., Howard A. G., Tien H. C., Du S., Zhang B., Wang H., Gordon-Larsen P., Batsis J. A. (2023). Association between weight status and rate of cognitive decline: China health and nutrition survey 1997-2018. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 78(6), 958–965. 10.1093/gerona/glad051 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ma L., Zhang L., Sun F., Li Y., Tang Z. (2018). Frailty in Chinese older adults with hypertension: Prevalence, associated factors, and prediction for long-term mortality. Journal of Clinical Hypertension, 20(11), 1595–1602. 10.1111/jch.13405 [DOI] [PMC free article] [PubMed] [Google Scholar]
- National Research Council Panel on Policy, R., & Data Needs to Meet the Challenge of Aging in, A . (2012). The national academies collection: Reports funded by national institutes of health. In Smith J. P., Majmundar M. (Eds.), Aging in Asia: Findings from new and emerging data initiatives (p. 486). National Academies Press (US). [PubMed] [Google Scholar]
- O’Caoimh R., Sezgin D., O'Donovan M. R., Molloy D. W., Clegg A., Rockwood K., Liew A. (2021). Prevalence of frailty in 62 countries across the world: A systematic review and meta-analysis of population-level studies. Age and Ageing, 50(1), 96–104. 10.1093/ageing/afaa219 [DOI] [PubMed] [Google Scholar]
- Popkin B. M., Du S., Zhai F., Zhang B. (2010). Cohort profile: The China health and nutrition survey--monitoring and understanding socio-economic and health change in China, 1989-2011. International Journal of Epidemiology, 39(6), 1435–1440. 10.1093/ije/dyp322 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Purser J. L., Kuchibhatla M. N., Miranda M. L., Blazer D. G., Cohen H. J., Fillenbaum G. G. (2008). Geographical segregation and IL-6: A marker of chronic inflammation in older adults. Biomarkers in Medicine, 2(4), 335–348. 10.2217/17520363.2.4.335 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Qin F., Guo Y., Ruan Y., Huang Z., Sun S., Gao S., Ye J., Wu F. (2023). Frailty and risk of adverse outcomes among community-dwelling older adults in China: A comparison of four different frailty scales. Frontiers in Public Health, 11, Article 1154809. 10.3389/fpubh.2023.1154809 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 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. 10.1093/gerona/62.7.722 [DOI] [PubMed] [Google Scholar]
- Rockwood K., Song X., MacKnight C., Bergman H., Hogan D. B., McDowell I., Mitnitski A. (2005). A global clinical measure of fitness and frailty in elderly people. Canadian Medical Association Journal, 173(5), 489–495. 10.1503/cmaj.050051 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rockwood K., Song X., Mitnitski A. (2011). Changes in relative fitness and frailty across the adult lifespan: Evidence from the Canadian national population health survey. Canadian Medical Association Journal, 183(8), E487–E494. 10.1503/cmaj.101271 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Saedi A. A., Feehan J., Phu S., Duque G. (2019). Current and emerging biomarkers of frailty in the elderly. Clinical Interventions in Aging, 14, 389–398. 10.2147/cia.S168687 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Song X., Mitnitski A., Rockwood K. (2010). Prevalence and 10-year outcomes of frailty in older adults in relation to deficit accumulation. Journal of the American Geriatrics Society, 58(4), 681–687. 10.1111/j.1532-5415.2010.02764.x [DOI] [PubMed] [Google Scholar]
- Vetrano D. L., Palmer K., Marengoni A., Marzetti E., Lattanzio F., Roller-Wirnsberger R., Lopez Samaniego L., Rodríguez-Mañas L., Bernabei R., Onder G., Joint Action ADVANTAGE WP4 Group . (2019). Frailty and multimorbidity: A systematic review and meta-analysis. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 74(5), 659–666. 10.1093/gerona/gly110 [DOI] [PubMed] [Google Scholar]
- Wang C., Ji X., Wu X., Tang Z., Zhang X., Guan S., Liu H., Fang X. (2017). Frailty in relation to the risk of Alzheimer’s disease, dementia, and death in older Chinese adults: A seven-year prospective study. The Journal of Nutrition, Health & Aging, 21(6), 648–654. 10.1007/s12603-016-0798-7 [DOI] [PubMed] [Google Scholar]
- Woo J., Yu R., Wong M., Yeung F., Wong M., Lum C. (2015). Frailty screening in the community using the FRAIL scale. Journal of the American Medical Directors Association, 16(5), 412–419. 10.1016/j.jamda.2015.01.087 [DOI] [PubMed] [Google Scholar]
- You X., Zhang Y., Zeng J., Wang C., Sun H., Ma Q., Ma Y., Xu Y. (2019). Disparity of the Chinese elderly’s health-related quality of life between urban and rural areas: A mediation analysis. BMJ Open, 9(1), Article e024080. 10.1136/bmjopen-2018-024080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yu P., Song X., Shi J., Mitnitski A., Tang Z., Fang X., Rockwood K. (2012). Frailty and survival of older Chinese adults in urban and rural areas: Results from the Beijing longitudinal study of aging. Archives of Gerontology and Geriatrics, 54(1), 3–8. 10.1016/j.archger.2011.04.020 [DOI] [PubMed] [Google Scholar]
- Yuan L., Yu B., Gao L., Du M., Lv Y., Liu X., Sun J. (2023). Decomposition analysis of health inequalities between the urban and rural oldest-old populations in China: Evidence from a national survey. SSM - Population Health, 21, Article 101325. 10.1016/j.ssmph.2022.101325 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zeng X. Z., Meng L. B., Li Y. Y., Jia N., Shi J., Zhang C., Hu X., Hu J. B., Li J. Y., Wu D. S., Li H., Qi X., Wang H., Zhang Q. X., Li J., Liu D. P. (2023). Prevalence and factors associated with frailty and pre-frailty in the older adults in China: A national cross-sectional study. Frontiers in Public Health, 11, Article 1110648. 10.3389/fpubh.2023.1110648 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang Y. S., Crimmins E. M. (2019). Urban-rural differentials in age-related biological risk among middle-aged and older Chinese. International Journal of Public Health, 64(6), 831–839. 10.1007/s00038-018-1189-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhou Q., Li Y., Gao Q., Yuan H., Sun L., Xi H., Wu W. (2023). Prevalence of frailty among Chinese community-dwelling older adults: A systematic review and meta-analysis. International Journal of Public Health, 68, Article 1605964. 10.3389/ijph.2023.1605964 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Material for The Association Between Urbanization and Frailty Status in China by Hillary B. Spangler, David H. Lynch, Annie Green Howard, Hsiao-Chuan Tien, Shufa Du, Bing Zhang, Huijun Wang, Penny Gordon Larsen, and John A. Batsis in Journal of Applied Gerontology