Social gradients in oral health status in Korea population

Social gradients in oral health status in Korea population

Archives of Oral Biology 95 (2018) 89–94 Contents lists available at ScienceDirect Archives of Oral Biology journal homepage: www.elsevier.com/locat...

235KB Sizes 0 Downloads 7 Views

Archives of Oral Biology 95 (2018) 89–94

Contents lists available at ScienceDirect

Archives of Oral Biology journal homepage: www.elsevier.com/locate/archoralbio

Social gradients in oral health status in Korea population

T

Hye-Sun Shin Department of Dental Hygiene, Eulji University College of Health Science, 553 Sanseong-daero, Sujeong-gu, Seongnam, 13135, Republic of Korea

A R T I C LE I N FO

A B S T R A C T

Keywords: Social gradient Oral health Inequality Tooth loss Chewing ability

Objectives: The aim of the present study was to investigate whether clinical (severe tooth loss) and subjective (chewing difficulties) indicators of oral health outcomes are associated with socioeconomic position and to explore the age-sex differences in the magnitude of the social gradient in Korea using data from the representative national data. Methods: Data were acquired from 10,939 men and women, ≥30 years of age who participated in the Korea National Health and Nutrition Examination Surveys conducted from 2012 to 2014. Education and income were used as socioeconomic position. Self-rated chewing difficulties and severe tooth loss were assessed by dentists and trained interviewers. Confounding variables were demographic factors, general health behaviors, and systemic health status. Results: Significant differences in oral health outcomes were observed according to the quartiles of income and education. In particular, the quartiles of education were significantly associated with oral health outcomes in the fully adjusted model with a dose-response trend. In participants aged 40–49 (OR = 2.30, 95% CI = 1.37–3.88) and 50–59 years (OR = 2.16, 95% CI = 1.49–3.14), the associations between the lowest quartiles of income and chewing difficulties were stronger than in the total population. Conclusions: Our findings demonstrate a clear and distinct social gradient in clinical and subjective oral health indicators based on socioeconomic position.

1. Introduction Research has documented socioeconomic disparities in morality (Giesinger et al., 2014; Pillay-van Wyk & Bradshaw, 2017) and general health outcomes (Sabbah, Tsakos, Chandola, Sheiham, & Watt, 2007). As with other chronic diseases, oral disease disproportionally affects socially disadvantaged members of society and produces a substantial social gradient in many studies (International Centre for Oral Health Inequalities Research & Policy, 2015). In other words, studies have shown that individuals of lower socioeconomic position are more likely to have poorer oral health status. Most epidemiological studies demonstrated an association between socioeconomic position and various oral health outcomes, including self-reported oral health status(Tsakos, Demakakos, Breeze, & Watt, 2011), chewing difficulties(Di Bernardi, Tsakos, Sheiham, Peres, & Peres, 2016), periodontal disease (Lee & Han, 2016; Sabbah et al., 2007), tooth loss (Han & Khang, 2016; Ravaghi, Quinonez, & Allison, 2013), and edentulism (Cunha-Cruz, Hujoel, & Nadanovsky, 2007; Elani, Harper, Allison, Bedos, & Kaufman, 2012; Tsakos et al., 2011). In the Korean National Health and Nutrition Examination Survey (KNHANES) 2014, 61.5% of respondents stated that they had ≥ 20

natural teeth, and the prevalence of chewing difficulty was 43.8%, respectively (Korea Center for Disease Control & Prevention, 2014). These results have changed positively since 2000 and improvement and maintenance of oral health status is now one of the major oral health policies of the Korean government among Korean elders (Korea Center for Disease Control & Prevention, 2016). Some studies have reported age and sex differences in the pattern of oral health inequalities in terms of the number of natural teeth (Ravaghi et al., 2013; Shen, Wildman, & Steele, 2013; Steele et al., 2015) and self-reported oral health status (Shen et al., 2013). Until now, no study has examined the social gradient of oral health outcomes according to age and sex of Korean adults. Additional evidence is needed to determine whether the social gradient and magnitude of oral health in Korea differs according to age group and sex in Korean adults. In addition, we included severe tooth loss and chewing difficulties, which are meaningful indicators of oral health in all Koreans(Locker, 2002). The aim of the present study was to investigate whether clinical (severe tooth loss) and subjective (chewing difficulties) indicators of oral health outcomes are associated with socioeconomic position and to explore the age-sex differences in the magnitude of the social gradient

E-mail address: [email protected] https://doi.org/10.1016/j.archoralbio.2018.07.021 Received 14 December 2017; Received in revised form 27 June 2018; Accepted 30 July 2018 0003-9969/ © 2018 Elsevier Ltd. All rights reserved.

Archives of Oral Biology 95 (2018) 89–94

H.-S. Shin

to intraoral problems, including teeth, gums, or dentures? (If you are using dentures, please select the condition you have while wearing the dentures.)”. A trained interviewer interviewed the participants using a structured questionnaire. The survey staff members were required to complete an intensive training course seven times per year and to conduct supervised practice before working on the survey. The response options were assessed by a 5-point rating scale: very much, quite a lot, some, very little, or not at all. The response options were dichotomized into no (some, very little, and not at all) and yes (very much and quite a lot). The criteria for this categorization were based on the annual KNHANES report (Korea Center for Disease Control & Prevention, 2014). The prevalence of chewing difficulties is an indicator that has been monitored in the elderly. A trained dentist conducted a full-mouth oral examination and recorded dental status to obtain information on the number of existing permanent teeth. Calibration for dental status was conducted annually and the mean kappa values for interexaminer reliabilities were as follows: 0.892 in 2012, 0.931 in 2013, and 0.920 in 2014. The number of existing permanent teeth was obtained after excluding implants, missing teeth, impacted teeth, and wisdom teeth. According to the World Health Organization report (World Health Organization, 1992), having ≥ 20 natural teeth is defined as functional dentition. Retention of a minimum of 20 natural teeth is also a key indicator of the oral health plan in Korea (Korea Center for Disease Control & Prevention, 2014). The number of existing permanent teeth was dichotomized into having ≥ 20 teeth and having ≤ 19 teeth. As with previous evidence (Han, Khang, & Lee, 2015; Shin, 2017), severe tooth loss was defined as having ≤ 19 teeth.

in Korea using data from the KNHANES 2012–2014. 2. Methods 2.1. Study design The present analysis is based on data from the 2012–2014 KNHANES, an ongoing surveillance system in Korea. KNHANES is a nationwide cross-sectional survey conducted every year since 1998 by the Korea Centers for Disease Control and Prevention (KCDC)(Kweon et al., 2014). The target population consists of nationally representative non-institutionalized Korean citizens. The sampling framework of the KNHANES uses the most recent National Census data. The sampling plan follows a complex, stratified, multistage, and probability-cluster design. The sampling method is a two-stage stratified sampling method with survey area and household as the first and second sampling units. The sampling frame was stratified based on province and city, town and village, housing type, and the ratio of residential area and education of the head of household was used as internal stratification criteria (Kweon et al., 2014). A total of 192 primary sampling units (PSUs) were chosen from geographical PSUs in Korea over the 3 years. A total of 567 PSUs were extracted for the 3 years, and the final 20 sampled households were selected using the systematic sampling method. All household members aged ≥ 1 year were selected as the target population. Sample weights were constructed for the sample participants to represent the Korean population by accounting for the complex survey design, survey nonresponse, and post-stratification. The weights based on the inverse of the selection probability and response rate were modified by adjusting them to the age- and sex-specific Korean population. A more detailed explanation of the sampling and selection methods in KNHANES is available (Kweon et al., 2014; Shin, 2017). KNHANES provides information about participant sociodemographic factors, health and oral health status, and nutritional status from a health interview, an oral and health examination, and a nutrition survey. The national survey was approved by the Institutional Review Board of the KCDC (2012-01EXP-01-2C, 2013-07CON-03-4C, and 2013-12EXP-03-5C). The individual health questionnaire data were collected in face-to-face interviews. The physical examinations and blood sampling were performed at a mobile examination center, and clinical measurements were carried out by trained medical personnel. The participation rates were 80% in 2012, 79.3% in 2013, and 77.8% in 2014.

2.4. Assessment of the socioeconomic position Education and income were used to define socioeconomic position. Trained interviewers rated socioeconomic position in face-to-face interviews. Information on education level was collected by asking the participants to indicate the highest level of education they had achieved (It was classified as previous education if completed, dropped out, enrolled, or absent from school). We categorized the level of education into four groups: below primary school (lowest), middle school, high school, and college or higher (highest). Income was collected by asking participants for their monthly household income. Monthly household income was adjusted for the number of household members and divided into quartiles as follows: first quartile (lowest), second quartile, third quartile, and fourth quartile (highest) of total equivalized income.

2.2. Eligibility and exclusion criteria 2.5. Assessment of potential confounders A total of 23,626 respondents participated in the 2012–2014 KNHANES. Of these, adults aged > 30 years were included in the analysis. Because the prevalence of severe tooth loss and chewing difficulties was very low for participants in their 20 s, participants aged < 30 years were excluded from analyses (n = 7196). Exclusion criteria for the 16,430 participants aged > 30 years were as follows: 1) those who did not receive an oral examination or health interview (including socioeconomic status variables) and 2) those with missing information for the confounders. The final sample comprised 10,939 participants (4516 males and 6423 females, age range: 30–93 years; mean age: 52.7 years), and 5491 participants were excluded.

Confounders in this study included demographic factors, healthrelated behavioral factors, and health status variables. Interviews using structured questionnaires were administered to assess potential confounders. Age, sex, and area of residence were selected as the demographic factors. Age was categorized into four groups for statistical analyses: 30–39 years, 40–49 years, 50–50 years, ≥ 60 years. The area of residence was classified into urban (called Dong) and rural (called Eup and Myeon) according to administrative district. Health-related behaviors included smoking, dental visits, and toothbrushing frequency. The participants were asked whether they were currently smoking, and the answers were categorized into: no (never smoker and past smoker) and yes (current smoker). Regular dental checkups were identified by asking “Have you visited a dentist to determine your oral health condition without any oral problems during the last year?” The answers were categorized into: no and yes. Daily tooth-brushing frequency was assessed by asking the following question “How many times did you brush your teeth yesterday?” The answers were grouped into less than two times and two or more times. In order to obtain information regarding the general health status,

2.3. Assessment of the oral health outcomes Presence or absence of chewing difficulties (subjective oral health outcome) and severe tooth loss (clinical oral health outcome) were used as outcome variables. One of the most critical consequences of oral disorders is reduced chewing ability (Locker, 2002). Self-rated chewing difficulty was assessed through the question “Do you have difficulties chewing food due 90

Archives of Oral Biology 95 (2018) 89–94

H.-S. Shin

education and chewing difficulties remained significant and showed a clear social gradient. In crude analyses, the OR of chewing difficulties in the lowest education quartile was 4.77 compared to participants in the highest income quartile. In particular, the strength of the association was considerably weakened after controlling for all potential confounders, but stronger than the quartiles of income (first quartile: OR = 2.32, 95% CI = 1.88–2.85, second quartile: OR = 1.81, 95% CI = 1.47–2.24, and third quartile: OR = 1.36, 95% CI = 1.16–1.60). A dose-effect trend was clearly observed for the association with quartiles of education and chewing difficulties.

the blood test and anthropometric measurements were conducted by trained medical professionals and interviewers. The definition of hypertension was systolic blood pressure ≥ 140 mm Hg or diastolic blood pressure ≥ 90 mm Hg or being medicated for hypertension. Body mass index (BMI) was used to define obesity. BMI was calculated by dividing weight in kilograms by the square of height in meters. Obesity was defined as a BMI ≥ 25.0 kg/m2. Diabetes mellitus was defined as having a fasting glucose level > 126 mg/dL or taking medication for diabetes. Medical decision-making often requires two groups (normal/ abnormal, cancerous/benign, and so on). Dividing all participants into two groups can simplify statistical analysis and provide easy interpretation and presentation of results of the study (Royston, Altman, & Sauerbrei, 2006).

3.3. Association between socioeconomic positions and severe tooth loss The same pattern of association was also observed for severe tooth loss. Significant social gradients existed for the first and second quartile, but not the third quartile of income (1 st quartile: OR = 2.14, 95% CI = 1.69–2.72, 2nd quartile: OR = 1.66, 95% CI = 1.31–2.11, and 3rd quartile: OR = 1.26, 95% CI = 0.96–1.64). In the unadjusted model, although the association with the quartiles of education and severe tooth loss and education was considerably strong (OR = 12.40, 95% CI = 9.55–16.10), the association was considerably weakened after including systemic health status. The fully adjusted ORs of the lowest income quartile for severe tooth loss were: 2.23 for the first quartile; 1.80 for the second quartile; and 1.52 for the third quartile) (Table 2).

2.6. Data analysis The outcome variables were oral health status (chewing difficulty and severe tooth loss) and the explanatory variable was socioeconomic position (education and income). Differences in the general characteristics of the participants were compared according to the oral health status using the chi-square test. All data are presented as weighted numbers and percentages. A multivariable logistic regression analysis was performed to examine the associations between socioeconomic position and subjective and objective oral health status. A multivariable logistic regression was used to compute unadjusted and adjusted odds ratios (ORs) and confidence intervals (CIs). Model 1 was unadjusted. Model 2 was adjusted for demographic variables, including age, gender and area of residence. Model 3 was adjusted for demographic factors, health-related variables, including smoking, regular dental checkup, and tooth-brushing frequency. Model 4 was adjusted for all confounding variables, including hypertension, obesity, and diabetes mellitus. We also performed a subgroup analysis by age and gender. All statistical analyses were performed using the SPSS 19.0 (SPSS Inc., Chicago, IL, USA) statistical program.

3.4. Age- and gender- stratified association between socioeconomic positions and oral health outcomes In the subgroup analyses, the stratified association between socioeconomic position and oral health outcome was moderate to strong across subgroups (Table 3). In participants aged 40–49 (OR = 2.30, 95% CI = 1.37–3.88) and 50–59 years (OR = 2.16, 95% CI = 1.49–3.14), the associations between the lowest quartiles of income and chewing difficulties were stronger than in the total population. When considering the results of the separate subgroup analyses according to sex, the association was strong in men (OR = 1.83, 95% CI = 1.37–2.46) for income, and strong in women (OR = 2.78, 95% CI = 2.07–3.73) for education. The associations between the indicators of socioeconomic position and severe tooth loss with income and quartiles of education were strong in participants aged 50–59 years (OR = 2.69, 95% CI = 1.51–4.79 for income; OR = 3.09, 95% CI = 1.52–6.28 for education). According to the results of a sex-stratified analysis, the indicators of socioeconomic position were strongly and significantly associated with severe tooth loss in females (OR = 2.47, 95% CI = 1.80–3.38 for income; OR = 4.77, 95% CI = 2.59–8.77 for education). The results of severe tooth loss should be interpreted carefully because only 12 participants aged 30–39 had < 20 teeth and 32 aged 40–49 years had < 20 teeth.

3. Results 3.1. Characteristics of the participants Table 1 presents the distribution of participants according to oral health outcomes (chewing difficulties and severe tooth loss). Of the 10,939 study participants, 25% (n = 2730) of the participants had selfrated chewing difficulties and 14.9% (n = 1625) had severe tooth loss. These two indicators of oral health outcome tended to increase with age, and the majority of participants (77.2%) with severe tooth loss were ≥ 60 years. Participants with poor oral health outcomes were more likely to live in a rural area than those without. The percentage of participants with a poor oral health outcome was significantly higher in those with hypertension, obesity, and diabetes mellitus. No differences were identified based on sex. Participants with poor oral health outcomes were significantly more likely to have a lower socioeconomic position than those with a normal oral health status (p < 0.001).

4. Discussion Our findings demonstrate a clear and distinct social gradient in clinical and subjective oral health indicators based on socioeconomic position after adjusting for age, sex, area of residence, smoking, regular dental check-ups, tooth-brushing frequency, hypertension, obesity, and diabetes mellitus. These results support previous studies that found an association (Di Bernardi et al., 2016; Han & Khang, 2016; Ravaghi et al., 2013). The magnitude of the social gradient was stronger in middle-aged adults than elders, and females than males in a nationally representative sample of Korean adults. Although there is no optimal indicator of socioeconomic position, the indicators used in our research were income and education. The education indicator is a strong determinant of future employment and income (Davey Smith et al., 1998). It captures the long-term effects of

3.2. Association between socioeconomic positions and chewing difficulty The quartiles of income were consistently associated with chewing difficulties and all potential confounders in the logistic models throughout the adjustment process, except for the third quartile (Table 2). In crude analyses, the OR of chewing difficulties in the lowest quartile of income was 1.91 compared to participants in the highest quartile of income. After adjusting for demographic variables, healthrelated behavior factor, and health status variables in that order, the association between quartiles of income and chewing difficulties was slightly weakened (first quartile: OR = 1.74, 95% CI = 1.44–2.11, second quartile: OR = 1.41, 95% CI = 1.20–1.66, and third quartile: OR = 1.13, 95% CI = 0.97–1.32). The association between quartiles of 91

Archives of Oral Biology 95 (2018) 89–94

H.-S. Shin

Table 1 Distribution of participants according to the oral health outcomes (N = 10,939).  

 

Chewing difficulty

Variables

Total n

No (N = 8209) n (%)

Yes (N = 2730) n (%)

2161 2365 2469 3944

1953 2086 1822 2348

(28.6) (29.7) (23.2) (18.4)

208 (10.4) 279 (15.4) 647 (29.4) 1596 (44.7)

4516 6423

3371 (48.4) 4838 (51.6)

8779 2160

Age (years) 30-39 40-49 50-59 ≥60 Sex Male Female Area of residence Urban (Dong) Rural (Eup, Myeon) Smoking status Never&former Current Regular dental checkup No Yes Toothbrushing frequencies <2 ≥2 Hypertension No Yes Obesity No Yes Diabetes mellitus No Yes Income I (< 25%) II (25-49%) III (50-75%) Ⅳ (> 75%) Education I (primary school) II (middle school) Ⅲ (high school) Ⅳ (college)

 

Severe tooth loss*

P-value†

No (N = 9314) n (%)

Yes (N = 1625) n (%)

< 0.001

2149 2333 2275 2557

12 (1.6) 32 (3.3) 194 (17.9) 1387 (77.2)

< 0.001

1145 (47.3) 1585 (52.7)

0.406

3802 (48.2) 5512 (51.8)

714 (47.3) 911 (52.7)

0.537

6750 (82.9) 1459 (17.1)

2029 (75.7) 701 (24.3)

< 0.001

7654 (82.7) 1660 (17.3)

1125 (69.9) 500 (30.1)

< 0.001

8875 2064

6729 (77.2) 1480 (22.8)

2146 (72.3) 584 (27.7)

< 0.001

7550 (75.8) 1764 (24.2)

1325 (78.8) 300 (21.2)

0.046

7848 3091

5807 (70.9) 2402 (29.1)

2041 (72.6) 689 (27.4)

0.168

6489 (70.0) 2825 (30.0)

1359 (82.0) 266 (18.0)

< 0.001

1230 9709

764 (9.5) 7445 (90.5)

466 (15.9) 2264 (84.1)

< 0.001

874 (9.5) 8440 (90.5)

356 (22.6) 1269 (77.4)

< 0.001

7463 3476

5892 (75.2) 2317 (24.8)

1571 (61.4) 1159 (38.6)

< 0.001

6736 (75.3) 2578 (24.7)

727 (46.8) 898 (53.2)

< 0.001

7315 3624

5550 (66.4) 2659 (33.6)

1765 (65.1) 965 (34.9)

0.316

6295 (66.7) 3019 (33.3)

1020 (61.6) 605 (38.4)

0.003

9675 1264

7451 (92.0) 758 (8.0)

2224 (83.2) 506 (16.8)

< 0.001

8429 (91.7) 885 (8.3)

1246 (76.4) 379 (23.6)

< 0.001

1941 2800 3035 3163

1099 2006 2450 2654

842 794 585 509

< 0.001

1253 2329 2764 2968

(11.0) (24.8) (31.4) (32.9)

688 471 271 195

(39.7) (29.3) (18.3) (12.7)

< 0.001

2671 1256 3510 3502

1468 (13.3) 840 (9.2) 2823 (37.2) 3078 (40.3)

< 0.001

1715 1012 3200 3387

(13.9) (9.9) (37.4) (38.8)

956 244 310 115

(55.3) (15.7) (21.1) (7.8)

< 0.001

(10.7) (24.2) (31.6) (33.4)

(26.2) (28.8) (24.1) (20.8)

1203 (36.8) 416 (15.1) 687 (30.1) 424 (18.0)

 

(27.5) (29.4) (25.4) (17.7)

 

P-value†

 

Data are presented as numbers and weighted percentage. Bold denotes statistical significance at P < 0.05. * Severe tooth loss was defined as having ≤ 19 teeth. † Obtained from chi-square test.

In contrast, chewing difficulties are related to whether you feel uncomfortable with your mouth at present, regardless of your experience. More importantly, the social gradients differed across age groups and the sexes. Our results provide evidence for the age and sex difference in Koreans for the first time. Our results show that the magnitude of the gradient was considerably weakened after controlling for age and sex, and it was assumed that the magnitude was affected by age group and sex. The association between education and chewing difficulties and severe tooth loss was highlighted in females rather than the total population. The strength of the association was strongest in women across all age and sex groups. These results are congruent with three previous studies (Ravaghi et al., 2013; Shen et al., 2013; Steele et al., 2015). Ravaghi et al. (Ravaghi et al., 2013) quantified health inequalities and realized a greater magnitude of inequality among women in terms of the number of missing teeth, not oral pain. Another study (Steele et al., 2015) presented income inequalities and the number of natural teeth among older groups, excluding the younger groups. Other studies (Shen et al., 2013) have emphasized the tendency for a difference between the number of teeth and OHIP (Oral Health Impact Profile) according to age group. These observations demonstrate that age group and sex are important factors in oral health inequalities. In agreement with a previous study, the results of the present study show a

early life circumstances on adult health, as well as the influence of adult resources on health (Davey Smith et al., 1998). Income is the best indicator of material living standard, which influences health through a direct effect on material resources (Galobardes, Shaw, Lawlor, Lynch, & Davey Smith, 2006). The results of this study showed a stronger gradient in education than income among the total population. Because oral health problems are cumulative events throughout the life course, the level of education that represents the past socioeconomic position was more appropriate than the level of income reflecting the present socioeconomic position. Epidemiological studies on oral health inequality have applied various oral health indicators in different countries (Ravaghi et al., 2013; Steele et al., 2015; Tsakos et al., 2011). Our findings on income and education gradients as they relate to subjective oral health status are similar to those of previous studies (Ravaghi et al., 2013; Shen et al., 2013; Steele et al., 2015; Tsakos et al., 2011). Interestingly, the present study determined that the social gradients are larger when measured using severe tooth loss compared with chewing difficulties. This result is in line with those obtained by Shen et al. (Shen et al., 2013) and Aida et al. (Aida et al., 2011). In particular, severe tooth loss is the result of cumulative oral health condition resulting from dental caries and periodontitis throughout the life course. 92

Archives of Oral Biology 95 (2018) 89–94

H.-S. Shin

Table 2 Multivariable logistic regression analysis of association between socioeconomic position and oral health outcomes (N = 10,939).  

 

 

Adjusted odds ratio (AOR) (95% confidence interval)

Variable

n

Model 1

Model 2

Model 3

Model 4

1.91 (1.59-2.29) 1.34 (1.15-1.58) 1.05 (0.90-1.23) 1

1.81 (1.50-2.18) 1.44 (1.23-1.69) 1.15 (0.98-1.34) 1

1.77 (1.46-2.14) 1.42 (1.20-1.67) 1.14 (0.97-1.33) 1

1.74 (1.44-2.11) 1.41 (1.20-1.66) 1.13 (0.97-1.32) 1

4.77 (3.99-5.70) 3.23 (2.68-3.90) 1.71 (1.47-2.00) 1

2.37 (1.93-2.90) 1.85 (1.50-2.29) 1.41 (1.20-1.65) 1

2.32 (1.88-2.85) 1.82 (1.47-2.26) 1.36 (1.16-1.60) 1

2.32 (1.88-2.85) 1.81 (1.47-2.24) 1.36 (1.16-1.60) 1

3.16 (2.52-3.97) 1.73 (1.38-2.16) 1.17 (0.91-1.49) 1

2.35 (1.85-3.00) 1.76 (1.38-2.23) 1.30 (0.99-1.70) 1

2.19 (1.73-2.78) 1.68 (1.32-2.13) 1.28 (0.98-1.67) 1

2.14 (1.69-2.72) 1.66 (1.31-2.11) 1.26 (0.96-1.64) 1

12.40 (9.55-16.10) 6.19 (4.65-8.22) 2.51 (1.92-3.27) 1

2.64 (1.98-3.51) 1.98 (1.47-2.70) 1.65 (1.23-2.20) 1

2.30 (1.71-3.09) 1.84 (1.36-2.48) 1.54 (1.15-2.07) 1

2.23 (1.66-3.00) 1.80 (1.33-2.42) 1.52 (1.14-2.03) 1

Outcome variable: chewing difficulty Income I (< 25%) 1941 II (25-49%) 2800 Ⅲ (50-75%) 3035 Ⅳ (> 75%) 3163 Education I (primary school) 2671 II (middle school) 1256 Ⅲ (high school) 3510 Ⅳ (college) 3502 Outcome variable: severe tooth loss Income I (< 25%) 1941 II (25-49%) 2800 Ⅲ (50-75%) 3035 Ⅳ (> 75%) 3163 Education I (primary school) 2671 II (middle school) 1256 Ⅲ (high school) 3510 Ⅳ (college) 3502

Note. AOR = adjusted odds ratio; CI = confidence interval. Model 1: unadjusted association. Model 2: adjusted for age, Sex, and area of residence. Model 3: adjusted for Model 2 plus smoking, regular dental checkup, and toothbrushing frequency. Model 4: adjusted for Model 3 plus hypertension, obesity, diabetes mellitus. Bold denotes statistical significance at p < 0.05.

clear social gradient in participants aged ≥ 50 years. Our results can be partially explained as follows. All generations have fundamentally different health behaviors, social circumstances, and life experiences and people are detrimental to their oral health outcomes as they get older. In addition, age and sex differences exist in the prevalence of tooth loss (Korea Center for Disease Control & Prevention, 2014) and chewing ability (Kim, Kim, Ahn, Chung, & Kim, 2015; Locker, 2002). In the

40–49-year-old age group, social gradient was only confirmed for chewing difficulty. Similarly, there have been reports of poor self-reported oral health outcomes using the OHIP in younger groups, despite the presence of many teeth (Shen et al., 2013; Slade & Sanders, 2011). Further well-designed longitudinal studies are required to elaborate on age and sex effects. Our study is the first to evaluate age-sex differences that were not

Table 3 Multivariable logistic regression analysis of age- and sex- stratified adjusted association between socioeconomic position and oral health outcomes [reference = Ⅳ(highest)]. Stratum

Income

  I (< 25%) Outcome variable: chewing difficulty Age group 30-39 1.15 (0.55-2.40) 40-49 2.30 (1.37-3.88) 50-59 2.16 (1.49-3.14) ≥60 1.74 (1.33-2.26) Sex Male 1.83 (1.37-2.46) Female 1.62 (1.27-2.06) Outcome variable: severe tooth loss Age group 30-39 5.50 (1.09-27.84) 40-49 0.57 (0.15-2.17) 50-59 2.69 (1.51-4.79) ≥60 2.11 (1.62-2.76) Gender Male 1.90 (1.35-2.69) Female 2.47 (1.80-3.38)

Education II (25-49%)

III (50-75%)

I (primary school)

II (middle school)

III (high school)

1.47 (0.96-2.25) 1.27 (0.87-1.83) 1.28 (0.93-1.77) 1.66 (1.28-2.17)

0.99 0.97 1.26 1.29

(0.66-1.48) (0.68-1.38) (0.94-1.68) (0.96-1.72)

2.02 (0.27-15.19) 1.76 (0.70-4.45) 2.38 (1.58-3.59) 2.09 (1.54-2.84)

2.23 (0.77-6.46) 2.26 (1.26-4.05) 1.90 (1.33-2.71) 1.53 (1.09-2.15)

1.66 (1.17-2.35) 1.40 (1.02-1.90) 1.39 (0.99-1.94) 1.17 (0.87-1.58)

1.65 (1.29-2.16) 1.20 (0.96-1.51)

1.19 (0.93-1.51) 1.07 (0.87-1.32)

2.16 (1.59-2.94) 2.78 (2.07-3.73)

1.69 (1.26-2.28) 2.16 (1.60-2.92)

1.29 (1.03-1.62) 1.51 (1.19-1.93)

1.59 (0.39-6.45) 0.69 (0.24-2.01) 2.32 (1.43-3.77) 1.60 (1.24-2.06)

0.24 (0.03-2.02) 0.50 (0.19-1.32) 1.79 (1.08-2.97) 1.25 (0.92-1.70)

. 0.59 (0.11-3.17) 3.09 (1.52-6.28) 2.20 (1.58-3.06)

. 3.25 (1.14-9.24) 2.25 (1.11-4.56) 1.75 (1.23-2.49)

5.58 (1.36-22.94) 0.57 (0.23-1.40) 1.88 (0.998-3.56) 1.58 (1.14-2.21)

1.63 (1.17-2.25) 1.79 (1.30-2.46)

1.21 (0.85-1.73) 1.35 (0.96-1.91)

1.74 (1.19-2.55) 4.77 (2.59-8.77)

1.83 (1.25-2.68) 3.15 (1.67-5.94)

1.43 (1.02-2.01) 2.71 (1.43-5.12)

Note. OR = odds ratio; CI = confidence interval. If there was no value, there were 12 participants aged 30–39 had < 20 teeth and 32 aged 40–49 years had < 20 teeth. All models were adjusted for age, sex, area of residence, smoking, regular dental checkup, toothbrushing frequency, hypertension, obesity, diabetes mellitus, except the stratum. Bold denotes statistical significance at p < 0.05. 93

Archives of Oral Biology 95 (2018) 89–94

H.-S. Shin

previously reported in Korean adults. In addition, the KNHANES database used in this study includes a large sample size of a representative Korean population, and was conducted with quality management of all investigated stages by the KCDC and various academic societies (Kweon et al., 2014). Previous studies did not assess systemic health status that affects the oral health condition except a few studies (Elani et al., 2012; Shen et al., 2013; Steele et al., 2015; Tsakos et al., 2011), and it was unclear exactly how socioeconomic status would affect oral health status. To overcome this limitation, we adjusted for various health-related variables and systemic health status variables in the logistic regression models. Above all, we used both objective clinical (severe tooth loss) and subjective (chewing difficulties) indicators as oral health outcome variables. The current results add to the evidence of an association between socioeconomic position and the two main indicators of oral health status in Koreans. Some limitations of this study should be discussed. First, although the results of our study can be generalized, it is difficult to identify causality because this study was cross-sectional. Second, we did not use a variety of socioeconomic indicators (occupation, childhood socioeconomic position, and parental socioeconomic position). In the future, it will be necessary to analyze various related indicators. Third, the number of medications used (Tan, Lexomboon, Sandborgh-Englund, Haasum, & Johnell, 2018) and some drug classes are known to cause xerostomia and salivary gland hypo-function, including dental caries and oral mucosal soreness (Wolff et al., 2017) in older adults. In further studies, it will be necessary to investigate regular use of medications that affect oral health status. Fourth, significant results are generally obtained when the sample size is large. We conducted an age- and sexstratified analysis because the residual false-positive effect always exists. Lastly, the systemic health variables were categorized by the cutoff point according to international standards. Dichotomization of variables will lead to residual confounding compared with adjusting the continuous variables. Moreover, unmeasured plausible confounders may be residual confounders that may affect the association between socioeconomic status and oral health status indicators. The results of the present study should be interpreted with caution.

Education and occupational social class: which is the more important indicator of mortality risk? Journal of Epidemiology and Community Health, 52(3), 153–160. Di Bernardi, E. R., Tsakos, G., Sheiham, A., Peres, K. G., & Peres, M. A. (2016). Association of changes in income with self-rated oral health and chewing difficulties in adults in Southern Brazil. Community Dentistry and Oral Epidemiology, 44(5), 450–457. Elani, H. W., Harper, S., Allison, P. J., Bedos, C., & Kaufman, J. S. (2012). Socio-economic inequalities and oral health in Canada and the United States. Journal of Dental Research, 91(9), 865–870. Galobardes, B., Shaw, M., Lawlor, D. A., Lynch, J. W., & Davey Smith, G. (2006). Indicators of socioeconomic position (part 1). Journal of Epidemiology and Community Health, 60(1), 7–12. Giesinger, I., Goldblatt, P., Howden-Chapman, P., Marmot, M., Kuh, D., & Brunner, E. (2014). Association of socioeconomic position with smoking and mortality: The contribution of early life circumstances in the 1946 birth cohort. Journal of Epidemiology and Community Health, 68(3), 275–279. Han, D. H., & Khang, Y. H. (2016). Lifecourse socioeconomic position indicators and tooth loss in Korean adults. Community Dentistry and Oral Epidemiology. Han, D. H., Khang, Y. H., & Lee, H. J. (2015). Association between adult height and tooth loss in a representative sample of Koreans. Community Dentistry and Oral Epidemiology, 43(5), 479–488. International Centre for Oral Health Inequalities Research and Policy (2015). Social inequalities in oral health:From evidence to action. London: University College London. Kim, Y. J., Kim, C. B., Ahn, Y. H., Chung, W. G., & Kim, N. H. (2015). Gender differences in chewing discomfort in older South Koreans. International Journal of Dental Hygiene, 13(4), 273–282. Korea Center for Disease Control and Prevention (2014). National health statistics 2014. Cheongju: Korea Center for Disease Control and Prevention. Korea Center for Disease Control and Prevention (2016). Strategic plan of oral health program in Korea. Cheongju: Korea Center for Disease Control and Prevention. Kweon, S., Kim, Y., Jang, M. J., Kim, Y., Kim, K., Choi, S., et al. (2014). Data resource profile: The korea national health and nutrition examination survey (KNHANES). International Journal of Epidemiology, 43(1), 69–77. Lee, H. J., & Han, D. H. (2016). Early-life socioeconomic position and periodontal status in Korean adults. Community Dentistry and Oral Epidemiology, 44(1), 11–23. Locker, D. (2002). Changes in chewing ability with ageing: A 7-year study of older adults. Journal of Oral Rehabilitation, 29(11), 1021–1029. Pillay-van Wyk, V., & Bradshaw, D. (2017). Mortality and socioeconomic status: The vicious cycle between poverty and ill health. The Lancet Global Health, 5(9), e851–e852. Ravaghi, V., Quinonez, C., & Allison, P. J. (2013). The magnitude of oral health inequalities in Canada: Findings of the Canadian health measures survey. Community Dentistry and Oral Epidemiology, 41(6), 490–498. Royston, P., Altman, D. G., & Sauerbrei, W. (2006). Dichotomizing continuous predictors in multiple regression: A bad idea. Statistics in Medicine, 25(1), 127–141. Sabbah, W., Tsakos, G., Chandola, T., Sheiham, A., & Watt, R. G. (2007). Social gradients in oral and general health. Journal of Dental Research, 86(10), 992–996. Shen, J., Wildman, J., & Steele, J. (2013). Measuring and decomposing oral health inequalities in an UK population. Community Dentistry and Oral Epidemiology, 41(6), 481–489. Shin, H. S. (2017). The number of teeth is inversely associated with metabolic syndrome: A korean nationwide population-based study. Journal of Periodontology, 88(9), 830–838. Slade, G. D., & Sanders, A. E. (2011). The paradox of better subjective oral health in older age. Journal of Dental Research, 90(11), 1279–1285. Steele, J., Shen, J., Tsakos, G., Fuller, E., Morris, S., Watt, R., et al. (2015). The Interplay between socioeconomic inequalities and clinical oral health. Journal of Dental Research, 94(1), 19–26. Tan, E. C. K., Lexomboon, D., Sandborgh-Englund, G., Haasum, Y., & Johnell, K. (2018). Medications that cause dry mouth as an adverse effect in older people: A systematic review and metaanalysis. Journal of the American Geriatrics Society, 66(1), 76–84. Tsakos, G., Demakakos, P., Breeze, E., & Watt, R. G. (2011). Social gradients in oral health in older adults: Findings from the English longitudinal survey of aging. American Journal of Public Health, 101(10), 1892–1899. Wolff, A., Joshi, R. K., Ekstrom, J., Aframian, D., Pedersen, A. M., Proctor, G., et al. (2017). A guide to medications inducing salivary gland dysfunction, Xerostomia, and subjective sialorrhea: A systematic review sponsored by the world workshop on oral medicine VI. Drugs R&D, 17(1), 1–28. World Health Organization (1992). Recent advances in oral health. Retrieved from Geneva.

Funding source This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Conflict of interest The author declares no conflicts of interest. References Aida, J., Kondo, K., Kondo, N., Watt, R. G., Sheiham, A., & Tsakos, G. (2011). Income inequality, social capital and self-rated health and dental status in older Japanese. Social Science & Medicine, 73(10), 1561–1568. Cunha-Cruz, J., Hujoel, P. P., & Nadanovsky, P. (2007). Secular trends in socio-economic disparities in edentulism: USA, 1972-2001. Journal of Dental Research, 86(2), 131–136. Davey Smith, G., Hart, C., Hole, D., MacKinnon, P., Gillis, C., Watt, G., et al. (1998).

94