Study population and sampling
The present study investigated the association between alcohol consumption and OA prevalence and OA-related pain in Koreans using data from the Korean National Health and Nutrition Examination Survey (KNHANES). KNHANES is a national sample survey conducted by the Korea Centers for Disease Control and Prevention (KCDC) under the Korean Ministry of Health and Welfare in South Korea. KNHANES is conducted to evaluate the health and nutritional status for the Korean people. The Korean population residing in Korea with the exception of foreigners, military service personnel, and nursing home and correctional facility residents participated in KNHANES. KNHANES data can be accessed and downloaded from the KNHANES website (https://knhanes.cdc.go.kr/knhanes/index.do). Data were obtained from the fifth KNHANES 2010–2012 (KNHANES V).
Twenty households are selected out of 192 regions each year, and 10,000 individuals aged ≥1 year are the target population for KNHANES. The KNHANES uses a complex, multi-stage probability sample design. The sample collectively represents the total non-institutionalized civilian population of Korea. The survey components are divided into 3 parts; Health Examination, Health Interview, and Nutrition Survey, and the survey and examination items and methods used are determined by KCDC and related academic societies to better monitor trends in risk factors for health and prevalence of major chronic diseases, and provide data for development and evaluation of South Korean health policy and programs. The health interview and examination are conducted by trained staff, including physicians, technicians and interviewers, by means of a mobile examination center, and follow-ups are conducted through dieticians’ visits to the homes of participants [15].
Among the 25,534 participants surveyed in KNHANES V, 7165 individuals aged 50 years or over who responded to drinking-related items were analyzed. Therefore, subjects aged under 50 years (n = 15,382), and subjects with missing Alcohol Use Disorders Identification Test (AUDIT) scores (n = 2987) were excluded. Ultimately, the analyses were performed using data of 7165 subjects, of which the subject inclusion and exclusion process is described in more detail in Fig. 1.
OA
OA screening was conducted by radiologists using X-rays in accordance with the “Professional Surveyor Education and Quality control for OA Examination.” OA examination and readings were performed by two radiologists using the Kellgren-Lawrence (KL) grading system. If the grades are discrepant by more than 2 grades, those digital data were read by another radiologist.
The KL grading classification was originally described with AP knee radiographs. Each radiograph was assigned a grade or 0–4, which correlated to increasing severity of OA (Grade 0 signifying no presence of OA, and Grade 4 severe OA). Additionally, the KL grading system provides detailed radiographic descriptions of OA as follows [16]:
-
0: No joint space narrowing or reactive changes
-
1: Doubtful joint space narrowing, possible osteophytic lipping
-
2: Definite osteophytes, possible joint space narrowing
-
3: Moderate osteophytes, definite joint space narrowing, some sclerosis, possible bone-end deformity
-
4: Large osteophytes, marked joint space narrowing, severe sclerosis, definite bone ends deformity
OA of the knee joint
KL grades of ≥2 as assessed by digital X-ray images of the knee joint were considered radiological OA of the knee joint [16].
OA of the hip joint
KL grades of ≥2 as assessed by digital X-ray images of the hip joint were regarded radiological OA of the hip joint [16].
OA of the lumbar spine
OA of the lumbar spine was evaluated through digital X-ray images of the lumbar spine in accordance with KL grades as follows: 0 = normal, 1 = definite osteophyte, and 2 = intervertebral disk space narrowing, bone sclerosis, and large osteophytes. KL grades were assessed by digital X-ray images of the lumbar spine in consideration of radiological OA of the lumbar spine [16].
Knee pain
Self-reported knee pain was assessed as a survey item included in the Health Interview, and pain severity was measured using the numeric rating scale (NRS). Participants who answered “Yes” to the question “Have you experienced knee pain for 30 days or more over the past three months?” were asked “What is your average knee pain (irrespective of whether you take medication or not)?” Participants indicated their answers using NRS values from 0 to 10 (0 indicative of no pain, and 10, worst pain imaginable).
Alcohol consumption
AUDIT developed by the World Health Organization (WHO) was employed to assess alcohol consumption. AUDIT is a concise measurement tool that evaluates excessive drinking, and was developed to examine the potential contribution of alcohol to development of various disorders [17]. The test is comprised of 10 items which measure the level of alcohol consumption and dependence, and the total sum of scores adds up to 40, with scores of ≥8 indicative of hazardous and harmful alcohol use. Of the 10 items, questions 1–3 measure hazardous alcohol use, and investigate the quantity and frequency of drinking and heavy drinking. Questions 4–6 pertain to alcohol dependency, and inquire into impaired control over drinking, increased salience of drinking, and morning drinking. The remaining 4 items (questions 7–10) ask about resulting recent and lifetime problems from harmful alcohol use, such as guilt after drinking, blackouts, alcohol-related injuries, and whether others are concerned about the respondent’s drinking.
Drinking habits as assessed by AUDIT were classified into 4 zones by level of intervention and risk. Zone I indicates non-drinking, abstinence or low risk drinking with an AUDIT score of ≤7 and intervention level of alcohol education. An AUDIT score of ≥8 is indicative of high risk drinking [18]: Zone II indicates alcoholic use beyond low-risk guidelines with an AUDIT score of 8–15 [19]; Zone III indicates an AUDIT score of 16–19 and intervention level covers brief counseling and continued monitoring in addition to simple advice; and Zone IV indicates an AUDIT score of ≥20 which suggests that the respondent is in need of professional treatment for alcohol dependency. Brief counseling may vary from simple 5-min advice to the aim of lowering dangerous drinking, to multiple counseling sessions in more serious cases. Brief counseling for Zone III respondents is characterized by low intensity and short duration [20].
Covariates
The sociodemographic characteristics of participants investigated in the survey included age, sex, education, income, occupation, marital status, and area of residence. Health-related characteristics included such factors as smoking, body mass index (BMI), and regular exercise.
Education level was categorized into 4 levels: elementary school graduate or lower, middle school graduate, high school graduate, and college graduate or higher. Income level was classified into quartiles by individual income. Occupation was classified into the following 7 categories with the exclusion of military service: (1) managing, administrative and professional positions, (2) office workers, (3) service and sales industry workers, (4) agriculture and fishery workers, (5) engineering and equipment and machinery operation and assembly workers, (6) simple laborers, and (7) unemployment (which included housewives and students). BMI (kg/m2) was further classified into underweight, normal and overweight according to the most commonly used definitions established by the WHO as < 18.5, 18.5 ≤ BMI < 25, and ≥ 25, respectively. Smoking status was trichotomized into (1) former smoker if the person had smoked ≥5 packs of cigarettes in the past and did not smoke at present, (2) current smoker if the person had smoked ≥5 packs in the past and continued to smoke, and (3) non-smoker if they indicated otherwise. Regular exercise was defined as engaging in (1) ≥20 min of rigorous regular exercise more strenuous than usual (e.g., running, climbing, cycling) for ≥3 days during the past week, (2) ≥30 min of regular exercise a little more challenging than usual (e.g., swimming, doubles tennis, volleyball) for ≥5 days during the past week, or (3) ≥30 min of walking for ≥5 days during the past week.
Statistical methods
KNHANES is a nationwide sample survey that employs stratified cluster sampling and weighted values. Stratification, clustering, and weighting were accordingly included as complex sample design elements in this complex sample data analysis. Statistical analyses were performed with SAS version 9.4 package (SAS Institute Inc., Cary, NC, USA), with p < 0.05 regarded to be statistically significant. Continuous variables are presented as mean ± standard deviation (SD), and nominal variables as frequency and percentiles (%). Difference in participant characteristics by drinking and OA was assessed using Rao-Scott Chi-Square test, or ANOVA in analysis of ≥3 groups. Logistic regression analysis for complex sample design was conducted to evaluate the association between AUDIT scores and OA after adjusting for covariates. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated with age and BMI as continuous variables, and other variables as nominal variables in adjustment for covariates.
