Study population
We assessed cross-sectional data collected from 2014–2016 in the China Hainan Centenarian Cohort Study (CHCCS) [20], a longitudinal observational study that targeted the population aged 100 years or older in Hainan, China. The CHCCS is based on a complete sample of the Hainan community-dwelling and institutionalized centenarian population conducted by an interdisciplinary research team. The study protocol was approved by the institutional review board of the Chinese People’s Liberation Army General Hospital (Beijing, China). All study participants provided written informed consent prior to enrollment.
According to the household register provided by the Civil Affairs Bureau in 2014, there were 1811 living centenarians in total. After a rigorous age validation process, we established contact with 1473 eligible individuals. We also excluded individuals who were not conscious and could not perform ADLs or participate in the questionnaire interview, physical health examination, and blood tests. Finally, 1002 centenarians aged 100–116 years were enrolled during recruitment and baseline data were collected between July 2014 and December 2016.
Activities of daily living
We conducted face-to-face interviews to assess the ability to perform ADL and inspections through relatives or caregivers to ensure effective and reliable investigations. ADLs consist of two components: basic ADLs (BADLs) and instrumental ADLs (IADLs). The Barthel index (BI) and Lawton scale were used to evaluate BADLs and IADLs, respectively.
The BI was developed in 1965 and later modified as a scoring technique that measures the patient’s performance in 10 activities of daily life (eating, grooming, bathing, dressing, bowel and bladder care, toilet use, ambulation, transfers, and stair climbing) [21, 22]. Each item is scored 0–5 points, 0–10 points, or 0–15 points according to different evaluation results and summed to a total of 0–100 points. The total BI score is used to classify the individuals’ level of dependence as total (0–20), severe (21–60), moderate (61–90), or slight (91–99). A total score of 100 indicates that the patient is independent of assistance from others [23]. In this study, we defined BADL disability as BI ≤60 points [24].
The Lawton IADL scale is an appropriate instrument used to assess independent living skills that are considered more complex than the BADLs (using the phone, shopping, preparing food, housekeeping, doing laundry, using transportation, handling medications, and handling finances) [25, 26]. Each skill measured by the scale requires some degree of both cognitive and physical function [27]. It takes 10–15 minutes to complete with a summary score ranging from 0–8. The total score of the scale sorts IADL dependence into 4 levels: severe (≤2), moderate (3–5), slight (6–7), and none (8). In this study, we defined IADL disability as a Lawton IADL score ≤2.
Serum albumin
Experienced nurses collected 8 ml of blood from each participant in a fasting state using four vacutainer tubes (2 ml each). All blood samples were transported to the clinical laboratory in a chilled transport container (4°C) and assayed within 4 hours. Comprehensive metabolic panel examinations were performed using a fully automatic biochemical autoanalyzer (Cobas 8000; Roche Products Ltd, Basel, Switzerland). Alb level was categorized into two groups according to the cutoff point of hypoalbuminemia (35 g/L) [28].
Covariates
For the interview and examination, participants were visited at their residences or nearby clinics by a well-trained interdisciplinary medical group. Using a face-to-face interview in the appropriate regional dialect, the interviewers recorded detailed information on a standardized structured questionnaire. Socio-demographic characteristics assessed included age, gender, education, ethnicity, marital status, educational level, and type of residence. We also considered lifestyle characteristics such as smoking status, alcohol drinking status, and weekly exercise. Standing height and weight were measured using a manufactured instrument, and the body mass index (BMI) was calculated by dividing the body weight by the height (in kg/m2). Participants with a BMI <18.5 kg/m2 were categorized as underweight [29]. Respondents were asked to report whether they had been diagnosed and treated by a doctor for any specific medical conditions. The presence of heart disease or stroke was self-reported. The presence of hypertension was defined by a self-report of high blood pressure, and/or sitting systolic blood pressure >140 mmHg and/or diastolic blood pressure >90 mmHg [30]. Similarly, diabetes was defined by self-report and/or a fasting blood glucose concentration of ≥7.0 mmol/L [31]. Impaired renal function was defined by self-report and/or a glomerular filtration rate <60 ml·min-1/1.73 m2 [32].
Statistical analysis
The results are presented as medians and interquartile ranges or means ± standard deviations for continuous variables, and as n (%) for categorical variables. Differences among groups were evaluated using the Mann–Whitney U test or the chi-squared test, as appropriate. Multivariable logistic regression models were constructed to evaluate the associations between Alb and BADL disability and IADL disability. ADL disability (BI ≤60) and IADL disability (Lawton IADL score ≤2) were used as outcomes. Alb level considered dichotomously (35 g/L as the cutoff point) was used as the independent variable. Also, stratified analyses were performed to assess the association between Alb and ADL disability among different sexes. In the multivariable analyses, we considered socio-demographic characteristics, lifestyles, morbidity, and other influential factors including BMI classification, hemoglobin (Hb), total cholesterol (TC) level, and serum 25-hydroxyvitamin D (25OHD) as covariates. All statistical analyses were performed using SPSS Statistics version 24.0 (IBM Corporation, Armonk, NY, USA). All tests were two-sided and performed at a 5% significance level.