Risk factors for progression of radiographic knee osteoarthritis in elderly community residents in Korea

Table 1 Baseline characteristics of the entire cohort, participants with complete follow-up, and participants lost to follow-up

^aCharacteristics	Entire cohort (n = 504)	Complete follow-up (n = 322)	Lost to follow-up (n = 182)
Age, median (IQR) years	71 (66.0–75.0)	71.0 (66.0–75.0)	72 (65.0–76.0)
Women	54.4	53.4	56.0
BMI, median (IQR) kg/m²	24.7 (22.4–26.7)	24.6 (22.4–26.5)	25.2 (22.4–27.0)
Lower level of education	78.0	75.5	82.4
Low income	24.4	22.0	28.6
Exercise (≥ 3 times/week)	26.0	28.3	22.0
Previous or current smoker	40.5	40.7	40.1
Previous or current alcohol consumption	41.5	39.4	45.1
Manual occupation	19.8	21.4	17
K/L grade in worst knee
Grade 3	9.7	9.6	9.9
Grade 4	9.3	8.4	11.0
Diabetes mellitus	10.1	9.0	12.1
Osteoporosis	19.2	19.3	19.2

IQR interquartile range, BMI body mass index, K-L Kellgren-Lawrence, TKR total knee replacement
^aExcept where indicated otherwise, values are written as percentages. Levels of education were classified as < 10 years or ≥ 10 years. Income was divided into 11 categories and low income was defined as < 500,000 Korean won per month. Exercise status was self-reported and responses were classified as < 3 times/week or ≥ 3 times/week. Smoking was defined as more than 20 packs of cigarettes having ever been smoked during the participants’ lifetime. Alcohol consumption was defined as the drinking of any alcoholic beverage more than once per month. Manual work was defined as work demanding physical or heavy physical exertion. Co-morbidity health information was also self-reported, and was recorded using 29 pre-defined diagnostic categories. Diabetes mellitus was defined as either a fasting glucose level ≥ 126 mg/dL or a 2-h glucose level of ≥200 mg/dL after 75-g oral glucose loading, or treatment for previously diagnosed diabetes mellitus

ISSN: 1471-2474