The ASPRA is a prospective cohort study of older adults who are registered in the NHS and reside in 2 rural towns (Haanmi-ri and Gaesu-ri) in Daehwa-myeon, Pyeongchang-gun, Gangwon-do, Korea. Pyeongchang-gun (County) is located 180 kilometers east of Seoul, with a total population size of 43,660 and older adults comprising 22% (Fig. 1). Due to the geographical distance to cities, the residents receive medical care almost exclusively from the regional CHPs.

Between October and December 2014, screening and recruitment took place in Haanmi-ri and Gaesu-ri. From the NHS member roster, potentially eligible residents were identified. They received a letter or a phone call from nurses to visit the CHP for their upcoming annual health examination or vaccination. If they were not able to attend, nurses made a home visit. A resident was considered eligible if he/she was: 1) aged 65 years or older; 2) registered in the NHS; 3) ambulatory with or without an assistive device; 4) living at home at the time of the assessment; and 5) able to provide informed consent. Excluded were those who were: 1) living in a nursing home; 2) hospitalized; or 3) bed-ridden and receiving a nursing-home level care at home. Because the study presents no more than minimal risk to the participants, a verbal informed consent was obtained.

To assess whether the ASPRA cohort is representative of older Koreans in rural communities, we analyzed data from the 6th Korea National Health and Nutrition Examination Survey (KNHANES) (14), conducted in 2013 and including 8,018 adults aged 65 years or older.

An experienced nurse conducted an interview, comprehensive geriatric assessment, and measurement of body composition at the regional CHP. Based on this information, we defined the key geriatric conditions as summarized in Table 1.

The nurse obtained information on demographic characteristics, age, sex, marital status (married, unmarried, widowed, divorced, or separated), living status (living alone, living with spouse or family, or living with others), occupation (agriculture, livestock farming, forestry, managing a store, working at a factory, or others), income (receiving Medical aid due to income < 500 USD per month), education level (no formal education, elementary school [≤ 6 years], middle or high school [7-12 years], and college or higher [> 12 years]), current drinking (yes/no), alcohol intake per week, current smoking (yes/no), cumulative lifetime smoking history (pack years), chronic conditions (angina, arthritis, asthma, cancer excluding minor skin cancer, chronic lung disease, congestive heart failure, diabetes, heart attack, hypertension, kidney disease, and stroke), current use of prescription and over-the-counter medications, and history of fall in the past year.

The nurse measured functional status, physical activity, disability, mood, nutritional status, and frailty status using validated instruments as described follows: 1) We administered the Korean version of the Mini-Mental State Examination (K-MMSE) (range: 0-30). Cognitive dysfunction was defined if K-MMSE score < 24 (15). 2) We used the Korean version of the Physical Activity Scale for the Elderly (K-PASE) (16). 3) As a physical function test, we measured handgrip strength (kg) using a dynamometer (T.K.K. 5401 Grip-D; Takei, Tokyo, Japan) and averaged two measurements from the dominant hand. Usual gait speed (m/second) was calculated from the time taken to walk four meters following one meter of unmeasured acceleration space. Slow gait was defined as < 0.8 m/second (17). Dismobility, a very slow gait, was defined as < 0.6 m/second (18). 4) To evaluate disability, using the validated scales for Koreans, we assessed dependence (i.e. requiring assistance from another person) in the following seven activities of daily living (ADL): bathing, continence, dressing, eating, toileting, transfer, and washing face and hands; and in the following ten instrumental activities of daily living (IADL): food preparation, household chores, going out short distance, grooming, handling finances, laundry, taking personal medication, shopping, using public transportation, and using the telephone (19). 5) We assessed depressive symptoms using the Korean version of the Center for Epidemiological Studies depression (CES-D) scale (range: 0-60) (20). Depression was defined if CES-D score ≥ 21 (21). 6) We defined malnutrition if the Mini-Nutritional Assessment Short Form (MNA-SF) score was < 8 points, and at risk for malnutrition if the MNA-SF score was 8 to 11 points (22). 7) We used two different validated frailty scales: Cardiovascular Health Study (CHS) frailty criteria (exhaustion, low activity, slowness, weakness, and weight loss) (4) and the Korean version of the FRAIL scale (K-FRAIL, an acronym for fatigue, resistance, ambulation, illnesses, and loss of weight) (2324). Each scale assigns scores from 0 to 5 and classifies participants as robust if a score of 0, prefrail if a score of 1-2, and frail if a score ≥ 3. Although both scales are validated and widely used in clinical practice and research, the K-FRAIL scale that relies on self-reported response is relatively easier to obtain compared to the CHS criteria which requires a lengthy physical activity questionnaire and a measurement of physical performance.

To assess body composition including total mass, lean mass, and fat mass, we used bioimpedance analysis (Inbody 620, InBody, Seoul, Korea) with measuring frequencies of 5, 50, and 500 kHz. The device measured the impedance of four limbs while participants were standing after overnight fasting to minimize the differences due to inter-individual fluid balance. Cut points for decreased muscle mass were a height adjusted appendicular skeletal muscle mass (ASM/height²) of < 7.0 kg/m² for men and < 5.7 kg/m² for women (17). This method has been previously validated in the Korean elderly population (25), with an R² value of 0.890 compared to dual energy X-ray absorptiometry. Sarcopenia was defined according to the consensus report of the Asian Working Group for Sarcopenia (17): ASM/height² of < 7.0 kg/m² for men and < 5.7 kg/m² for women and decreased physical performance (by either handgrip strength < 26 kg for men or < 18 kg for women, or gait speed of < 0.8 m/second).

At the time of writing this protocol, we plan to follow the ASPRA cohort for 5 years. During the follow-up period, we will conduct repeated interviews, geriatric assessments, and body composition measurements to determine the presence of selected geriatric conditions as listed in Table 1. We will also gather information on vital status, hospitalizations, and nursing home admission every three months by telephone interview that will be confirmed by contacting the hospital or facility.

We characterized our study population using the mean and standard deviation (SD) for continuous variables and proportions for categorical variables. To assess the generalizability of the ASPRA cohort, we compared demographic characteristics of the ASPRA participants with those of the nationally representative KNHANES cohort. We used stratified clustered weighting to build the standard population (4,441,345 people aged 65 years or older in urban area and 1,696,357 people aged 65 years or older in rural area) of Korea in 2013. In KNHANES, standard error (SE) was used to show variations of data; otherwise, standard deviation (SD) was used. To determine the burden of frailty and geriatric conditions in rural population, we assessed the prevalence of frailty and geriatric conditions between men and women in the ASPRA cohort, after adjusting for age using logistic regression. We compared the two frailty scales using the Kappa statistic and Spearman correlation coefficient as well as their associations with each geriatric condition. Using logistic regression, we computed the odds ratio (OR) and 95% confidence interval (CI) of each geriatric condition associated with frailty versus no frailty according to each frailty scale, independently of age, sex, education level, and Medical aid status. A two-sided P value of < 0.05 was considered statistically significant.

The study protocol was reviewed and approved by the institutional review board of the Asan Medical Center, Seoul, Korea (IRB No. 2014-0988). Informed consent was waived by the board.

We screened 244 (98.8%) of 247 residents in Haanmi-ri and 150 (95.5%) of 157 in Gaesu-ri (Fig. 2). Of 404 residents who were screened, we excluded 12 who were institutionalized in a nursing home (n = 9), a hospital (n = 1), or bed ridden and receiving nursing care in their home (n = 2). Ten of 392 eligible residents declined to participate in the study, leaving a total of 382 (167 men and 215 women) participants.

Sociodemographic characteristics of the ASPRA participants were compared with the nationally representative KNHANES cohort (Table 2). The mean age (74.4 years) and proportion of males (43.7%) in the ASPRA cohort were similar to those of the KNHANES population in rural as well as urban area. A unique characteristic of the ASPRA cohort was that more individuals were currently working (53.4% engaged in agriculture) and had no formal education (44.8%) compared to the older population in rural area in KNHANES (25.1% in agriculture and 22.6% with no formal education). However, the proportion of college graduates (8.4%) was similar to that in older population in urban area in KNHANES (8.4%). The proportion of individuals with income below the absolute poverty limit (monthly income < USD 500) who received Medical aid service (5.8%) was similar to the rural KNHANES population (6.2%), but lower than urban KNHANES population (10.3%). More participants were married (63.4%) in the ASPRA cohort than rural (57.6%) and urban (60.5%) KNHANES populations. The proportion of those living alone (22.8%) was smaller than that of the rural KNHANES population (26.4%).

Complete data on CHS frailty criteria were available in 380 participants and K-FRAIL scale data were available in 382 participants. According to the CHS frailty criteria, 66 participants (17.4%) were frail and 200 (52.6%) were prefrail (Table 3). Of the 5 CHS frailty components, weakness (50.9%) and exhaustion (33.2%) were more common than other components (16.0%-19.9%). Among the 66 participants with frailty, 17 (25.8%) were frail without multimorbidity or disability. After adjusting for age, more women were frail or prefrail than men (22.4% vs. 10.8% for frailty; 57.0% vs. 47.0% for prefrail status, mainly due to high prevalence of exhaustion (38.6% vs. 26.3%) and weakness (66.7% vs. 30.7%) in women.

The overall burden of geriatric conditions other than frailty was high in the ASPRA cohort, particularly in women (Fig. 3); 33.5% of men and 58.6% of women had ≥ 2 geriatric conditions. The most prevalent geriatric condition was IADL disability (38.5%), followed by malnutrition risk (37.9%), cognitive dysfunction (33.4%), multimorbidity (31.9%), and sarcopenia (27.8%), whereas dismobility (8.4%), incontinence (8.4%), and polypharmacy (2.6%) were relatively uncommon (Table 3). After adjusting age, women were more likely than men to live alone and have multimorbidity, sarcopenia, cognitive dysfunction, depression, history of fall, and malnutrition risk.

Using the K-FRAIL scale, 105 participants (27.5%) were frail and 165 (43.2%) were prefrail. Of the five K-FRAIL components, resistance (55.0%), fatigue (46.9%), and ambulation (44.5%) were more common than loss of weight (18.6%) and illness (1.0%). The agreement between the two frailty scales was moderate (Kappa statistic: 0.41, P < 0.001; Spearman correlation coefficient: 0.64, P < 0.001).

Compared with the K-FRAIL scale, frailty according to the CHS criteria was more strongly associated with the presence of several geriatric conditions (Table 4): ADL disability (OR comparing the CHS criteria vs. K-FRAIL scale: 7.63 vs. 1.37), IADL disability (7.97 vs. 2.66), dismobility (15.8 vs. 2.04), at risk for malnutrition (4.24 vs. 2.30), polypharmacy (4.17 vs. 1.53), and incontinence (4.27 vs. 1.11). The associations with multimorbidity, sarcopenia, cognitive dysfunction, depression, and history of fall did not substantially differ between the two frailty criteria. Medical aid status (monthly income < USD 500) and living alone were associated with both definitions of frailty.