This study utilized data of CHARLS from 2011 to 2020 to explore the associations between BMI, WC, ABSI, WWI, WHtR, and ADL impairment in middle-aged and older adults aged 45 years and above. The study participants were representative samples from 28 provinces across China, selected using a stratified cluster sampling method (15). Baseline data covered information such as age, gender, height, weight, WC, education level, marital status, residence place, smoking, drinking, hypertension, diabetes mellitus (DM) and dyslipidemia. To ensure data completeness and accuracy, samples with missing data, outliers, or logical errors were excluded. Participants were excluded if they met the following criteria: (1) age below 45; (2) missing ADL data; (3) missing BMI, WC, ABSI, WWI, WHtR data; (4) missing demographic data; (5) WC less than 40 cm. Ultimately, 8,700 subjects from 2011 were included in the cross-section analysis. The prospective study used baseline data from 2011 and follow-up data from 2013, 2015, 2018, and 2020, with a total of 5,945 participants included. This study was approved by the Peking University Medical Ethics Committee (IRB00001052-11015), and all participants signed informed consent to participate in the follow-up surveys.

ADL impairment was mainly assessed using the ADL scale. Each item on the scale had four response options: “no difficulty, ““some difficulty but can still complete, ““difficulty and needs help, “and “unable to complete” (16). The scale included six tasks essential for independent living: eating, toileting, bathing, dressing, indoor activities, and continence control, as well as tasks related to community living, including housekeeping, cooking, shopping, medication use, telephone use, and managing finances. ADL impairment was defined as the inability to perform any of the above tasks independently (17).

In the CHARLS study, measurements of WC, height, and weight were conducted following strict protocols. WC was measured using a calibrated tape measure at the midpoint between the lower edge of the rib cage and the iliac crest, with the subject standing in a natural breathing position and the reading taken at the end of expiration. Height was measured with the participant standing barefoot, with eyes looking straight ahead. For weight measurement, participants were required to remove heavy outer clothing and any accessories that might affect the results, and the value was recorded once stabilized. BMI was calculated using the formula: B M I = W eight Heigh t 2 kg/m², which is used to assess weight status (18). ABSI was calculated using the formula: ABSI = W C B M I 2 / 3 × Heigh t 1 / 2 (19). WHtR was calculated by dividing WC by height, incorporating body shape into the assessment of central obesity. Finally, the WWI was derived by dividing WC by the square root of weight, reflecting the presence of excessive body fat buildup and the loss of muscular mass (14).

To assess the impact of potential risk factors, demographic and health status variables were collected through a structured questionnaire. These included age (as a continuous variable), gender (male or female), marital status (married or non married), place of residence (urban or rural), education level (illiterate, primary/middle school, high school or above), smoking, and drinking (categorized as no or yes based on current behavior).

Hypertension diagnosis was based on blood pressure measurements and physician diagnosis. Dyslipidemia was diagnosed if any of the following criteria were met: total cholesterol ≥6.2 mmol/L, low-density lipoprotein cholesterol ≥4.1 mmol/L, triglycerides ≥2.3 mmol/L, High-density lipoprotein cholesterol <1 mmol/L, or physician diagnosis (20). The diagnosis criteria for DM included any of the following: fasting blood glucose ≥7 mmol/L, random blood glucose ≥11.1 mmol/L, glycated hemoglobin ≥6%, or physician diagnosis (21).

Participants were divided into two groups based on the presence or absence of ADL impairment. Means±standard deviations were used to represent continuous variables, while percentages were used for categorical variables. T-test was used for continuous variables, and a chi-square test was used to assess differences between groups for categorical variables. Binary logistic regression models were employed in the cross-sectional study to calculate odds ratio (OR) with corresponding 95% confidence interval (95% CI). In the prospective study, Cox proportional hazards regression models were utilized to estimate hazard ratio (HR) with 95% CI, complemented by Kaplan–Meier (KM) survival curve visualization. To control for potential confounding factors, adjustments were made: Model 1 was adjusted for age and gender; Model 2 was further adjusted for smoking, alcohol consumption, residence, marital status, and education level; Model 3 was additionally adjusted for hypertension, dyslipidemia, and DM. Subsequently, restricted cubic spline (RCS) curve was used to identify trends, with subgroup analyses performed according to gender, marital status, residence, smoking, drinking, hypertension, dyslipidemia, DM, BMI and age. For enhanced clinical interpretability of OR and HR, ABSI values were standardized by multiplying by 100. Similarly, WHtR values were standardized by multiplying by 10. Diagnostic performance was quantified through receiver operating characteristic (ROC) curve analysis, with area under the curve (AUC) values computed. The statistical significance set at p < 0.05. All statistical analyses were performed using R 4.4.1.

Table 1 presents the baseline characteristics of the 8,700 participants. Among all participants, 3,603 (41.41%) were male, and 5,097 (58.59%) were female. The results indicated that 1,776 participants suffered from ADL impairment. Compared to those without ADL impairment, participants with ADL impairment were generally older, female, unmarried, living in rural area, no drinking, less educated, and more likely to have hypertension, and DM, lower weight and BMI, while higher ABSI WHtR and WWI. However, no significant difference in WC was observed between the two groups.

Table 2 shows the associations between BMI, ABSI, WWI, WHtR, and ADL impairment using the binary logistic regression model. After adjusting for age, gender, smoking, drinking, residence place, marital status, education, hypertension, dyslipidemia, and DM, BMI (OR = 0.98, 95% CI: 0.97–1.00) showed negative association with ADL impairment while ABSI (OR = 1.25, 95% CI: 1.14–1.37), WWI (OR = 1.16, 95% CI: 1.09–1.24) were positively associated with ADL impairment.

The baseline population in 2011 consisted of 5,945 participants. 3,096 participants had experienced ADL impairment during the follow-up period. Among them, 2,392 participants (40.24%) were male, and 3,553 participants (59.76%) were female. Among all participants, 3,202 experienced a complete 9-year follow-up period. The results indicated that participants with ADL impairment tended to be older, female, unmarried, less educated, living in rural areas, non-smoker, non-drinker, and were more likely to have hypertension and DM. Additionally, these participants had lower BMI and higher ABSI, WHtR, WWI compared to those without ADL impairment. However, there was no significant difference of WC between the two groups (Table 3).

The prospective associations between BMI, ABSI, WWI, WHtR and ADL impairment are presented in Table 4. After adjusting for confounding factors, both ABSI (HR = 1.07, 95% CI: 1.01–1.14) and WWI (HR = 1.05, 95% CI: 1.01–1.10) retained their significant associations with ADL impairment. The results of quartile analysis indicated that, in comparison with the lowest quartile, the highest quartile of WWI (HR = 1.14, 95% CI: 1.02–1.28) was significantly associated with an elevated risk of ADL impairment.

Figure 1 illustrates the KM survival curves for ADL impairment across quartile groups of ABSI and WWI. Notably, compared with the lowest quartile group, the risk of ADL impairment in the highest quartile group showed a more significant upward trend. Moreover, the median occurrence time for both ABSI and WWI in the highest quartile group was 7 years, and there was no significant difference between the two indicators.

Figure 2 presents the results of the RCS analysis among ABSI, WWI, and ADL impairment in the prospective analysis. After adjustment for confounding factors, positive linear correlations were observed among ABSI, WWI, and ADL impairment, suggesting that as ABSI and WWI increase, the risk of ADL impairment occurring rises significantly.

The results of the subgroup analysis indicates that there is no difference in the interaction among subgroup variables after adjusting for confounding factors, whether for ABSI or WWI. This implies that the impacts of ABSI and WWI on ADL impairment were generally stable (Figure 3).

The ROC curve was employed to evaluate the diagnostic value of ABSI and WWI in predicting ADL impairment. As depicted in Figure 4, the areas under the curves (AUC) of the two composite indices reflecting central obesity were both greater than 0.5. This indicates their potential in identifying the risk of ADL impairment. Notably, the predictive ability of WWI was slightly higher than ABSI, with an AUC value of 0.598.