Oral health inequalities in a national sample of Australian children aged 2–3 and 6–7 years

Authors


Ms Nina Lucas
Parenting Research Centre
Level 5, 232 Victoria Parade
East Melbourne VIC 3002
Email: nlucas@parentingrc.org.au

Abstract

Background:  While inequalities in oral health are generally well documented, it is less clear whether such patterns are evident from early childhood. Using four measures of potential inequality, this study examined patterns in oral health for Australian children at ages 2–3 and 6–7 years.

Methods:  Cross-sectional data from two cohorts of children in the Longitudinal Study of Australian Children (LSAC) were used to explore associations between reported oral health and four indicators of social disadvantage: socio-economic position (SEP), residential remoteness, Indigenous status and non-English speaking background.

Results:  For both cohorts, lower SEP and Indigenous status were associated with higher odds of poor oral health on all three indicators, and less accessible location was associated with increased odds for caries. Non-English speaking background was associated with increased odds for caries experience in 2–3 year olds and non-use of dental services in the older cohort. Inequalities were larger in the older cohort for socio-economic position and toothbrushing.

Conclusions:  Marked social disparities in oral health appear as early as 2 years of age and remain evident in school-age children. Interventions to reduce such disparities should start as early as possible.

Abbreviations and acronyms:
ARIA

Accessibility/Remoteness Index of Australia

ATSI

Aboriginal or Torres Strait Islander

LSAC

Longitudinal Study of Australian Children

NESB

non-English speaking background

SEP

socio-economic position

Introduction

Understanding the epidemiology of child oral health and oral health behaviours is important for identifying opportunities to reduce the future burden of dental disease in the community. It is clear that good oral health in childhood is not enjoyed equally. The most recent Australian Child Dental Health Survey found that 66% of 4-year-olds were caries free, but the remaining 34% had on average at least four decayed teeth with 80% being active untreated disease.1 The relationship between children’s social circumstances and their oral health has been demonstrated in numerous studies.2–7 By school age, significant inequalities in caries prevalence are evident in children exposed to greater levels of disadvantage. However, the extent to which oral health outcomes and behaviours are socially patterned during the early years of life is not fully understood.

To date, there has been a lack of nationally representative data on oral health and oral health inequalities for very young Australian children. Samples from specific regional and metropolitan locations or School Dental Service registers may not be generalizable to the broader Australian population, and may result in misleading estimates of the prevalence of problems and inequalities.8–10 Furthermore, many previous studies examining children’s oral health inequalities have used area-level measures of socio-economic status to assess social disadvantage.9,11 However, such measures are not synonymous with individual social circumstances which convey risks for poor oral health that are usually greater than or interact with those conveyed by place of residence.12,13 Studies employing area-level measures alone may result in an underestimation of oral health inequalities. At the individual level, the most common measures of socio-economic status have been family income, parental education or parental occupational prestige.3,4,12 Multi-component indicators of socio-economic position combining these variables address concerns that single indicators may be poor at representing a multi-dimensional phenomenon and offer several advantages. Composite measures take account of multiple intercorrelated sources of social difference, are analytically parsimonious and can be constructed to enable comparisons across different family structures (e.g. one- and two-parent households).

In summary, previous studies have been limited by sampling and measurement methods that may obscure the full extent of oral health inequalities, thereby hindering the development of optimal policies and practices. Further, variations in oral health and oral health behaviours by different types of social disadvantage are also unknown, particularly in relation to the younger population.

The paper addresses these issues using new national Australian data for two cohorts of children aged 2–3 and 6–7 years. This cross-sectional study has three aims. First, to determine whether there are differences in parent-reported dental service use, frequency of toothbrushing and experience of new caries for two cohorts of children aged 2–3 and 6–7 years. Second, to examine variations in these three outcomes by four indicators of social disadvantage: SEP, accessibility, Indigenous status and non-English speaking background. Finally, we compare cohorts to see whether there are any differences in patterns of oral health inequalities in the two age groups.

Materials and Methods

Sample

This paper uses cross-sectional data from the Longitudinal Study of Australian Children (LSAC), an omnibus national study of children’s health and development. LSAC commenced in 2004 (wave 1) with an infant cohort consisting of 5107 infants (aged 3–19 months) and a child cohort consisting of 4983 children (aged 4 years, 3 months to 5 years, 7 months). Oral health data were collected for the first time in wave 2 (2006) when the infant cohort was aged 2–3 years (90.2% retention) and the child cohort was 6–7 years (89.6% retention). Both cohorts were broadly representative of the Australian population at recruitment and wave 2, with some under-representation of less highly educated, single-parent or non-English speaking families, and families living in rental properties.14,15 LSAC data collection was granted approval by the human research ethics committee of the Australian Institute of Family Studies.

Sampling design

The LSAC samples were drawn from the federal government’s universal health insurance (Medicare) enrolment database using a two-stage stratified clustered design involving the selection of postcodes and then children within postcodes.16 Postcodes were divided into strata based on Australian state, capital city versus rest-of-state location, and whether the population of children in the postcode contained the desired cluster size.16 Postcodes were selected on a probability proportional to size basis, with around 10% of all Australian postcodes included. Children were selected within postcodes via simple random sampling.

Measures

Data were collected from the child’s primary carer (97% were the child’s biological mother) via face-to-face interviews and self-complete questionnaires. Reported measures of children’s oral health and related behaviours were: experience of caries (cavities, fillings and extractions; coded ‘yes’, if ‘yes’ for any of these) since the last interview (approximately 24 months); frequency of toothbrushing (‘less than twice per day’ vs. ‘twice or more’); and use of dental services in the last 12 months (yes/no).

Socio-economic position (SEP) was derived from standardized scores for: combined annual household income (with natural log transformation); parents’ years of education; and parents’ occupation (main occupation and occupational status). The resulting continuous score was split into quintiles with group 1 representing the least disadvantaged group and 5 representing the most disadvantaged group.

Accessibility of the child’s residence was coded using the Accessibility/Remoteness Index of Australia (ARIA),17 which classified the minimum travelling distance required to access an urban centre containing basic services. After initial modelling, the original five categories were collapsed into two categories due to small numbers of children in the more remote locations: accessible (original categories of ‘highly accessible’ and ‘accessible’) vs. restricted accessibility (‘moderately accessible’, ‘remote’ and ‘very remote’).

Indigenous status indicated whether the child was of Aboriginal or Torres Strait Islander origin (ATSI) and non-English speaking background (NESB) indicated whether the child’s primary carer spoke a language other than English at home.

Dental services data were collected through mail-back questionnaires, with missing data rates of 28% for 2–3 year olds and 26% for 6–7 year olds. Missing data rates were low for toothbrushing and oral health which were collected via interview (0–1% for both cohorts).

Analyses

LSAC wave 2 design weights were applied to all analyses to correct for non-response and attrition bias. These were calculated as the inverse of a child’s probability of selection and benchmarked to stratum totals for state, part-of-state, child gender, and attrition bias at wave 2.14,15 Standard errors were estimated using first-order Taylor linearization to account for the two-stage clustered design. Logistic regression analyses were used to examine relationships between the disadvantage variables (SEP, restricted accessibility, Indigenous, NESB) and oral health outcomes, adjusting for child’s age in months and gender. Initial analyses were completed for boys and girls separately (tables available from the authors); however few gender differences were found. Finally, tests of interaction were used to determine whether the patterns of outcomes by each disadvantage indicator (SEP, accessibility, NESB and Indigenous status) varied by age group. Analyses were performed using SPSS version 16.0 (SPSS Inc. SPSS for Windows, Release 16.0.0. Chicago: SPSS Inc, 2007) and Stata 10.0 (StataCorp. Stata statistical software: Release 10. College Station, TX: StataCorp LP, 2007).

Results

Cohorts were similar in demographic characteristics and as expected there were substantial cohort differences in the rates of dental service use, toothbrushing and caries (Table 1). Older children were much more likely to have attended a dental service in the last 12 months (59% vs. 15%) and to brush their teeth at least twice a day (61% vs. 44%). However, older children were also much more likely to have parent-reported caries in the last 24 months (32% vs. 3%).

Table 1.   Demographic characteristics and number of children experiencing oral health problems by age group
  2–3 years
n = 4606
6–7 years
n = 4464
  1. AUD = Australian dollars.

  2. †Due to small numbers in the moderately accessible and remote/very remote areas, these categories were collapsed for analysis.

  3. ‡Sample size for dental service use was 3432 for 2–3 years and 3291 for 6–7 years.

Demographic characteristics
Child’s age (months)Mean (SD)34.01 (3.0)80.1 (3.0)
Mother’s age (years)Mean (SD)33.05 (5.7)36.82 (5.60)
Child gender (male)n (%)2359 (51.2)2288 (51.3)
Mother completed year 12n (%)2598 (56.5)2172 (48.7)
Annual household income (AUD)Mean (SD)$34 484 ($17 339.20)$34 325 ($17 252.20)
Accessibility†
 Highly accessible/accessiblen (%)3607 (79.3)3472 (78.6)
 Moderately accessible arean (%)765 (16.8)753 (7.1)
 Remote/very remote arean (%)179 (3.9)192 (4.4)
Indigenousn (%)236 (5.1)167 (3.7)
Non-English speaking backgroundn (%)750 (16.3)658 (14.8)
Oral health measure
Has used a dental service in last 12 months‡n (%)
95% CI
505 (15.2)
(13.8, 16.6)
1955 (59.4)
(57.3, 61.4)
Cleans teeth at least twice per dayn (%)
95% CI
2045 (44.4)
(42.7, 46.1)
2740 (61.4)
(59.5, 63.3)
Caries experience in the last 24 monthsn (%)
95% CI
154 (3.4)
(2.7, 3.9)
1443 (32.4)
(30.8, 33.9)

Table 2 displays the results for the logistic regression analyses assessing associations between SEP and oral health outcomes. For both cohorts, lower SEP was associated with higher odds of parent-reported caries, infrequent toothbrushing and non-use of dental services in the last 12 months. Tests of interaction were used to determine whether the patterns of outcomes by SEP varied by age group. A significant SEP by age group interaction was found for frequency of toothbrushing, which showed a stronger effect among 6–7 year olds than among 2–3 year olds. The most disadvantaged 6–7 year old children had 3.40 times the odds of infrequent toothbrushing (95% CI 2.77, 4.18) compared to the least disadvantaged group, while for 2–3 year olds, the odds was 1.26 (95% CI 1.05, 1.52). There was also a trend towards an SEP by age interaction for caries experience, with the more disadvantaged younger children tending to have higher odds for caries than for similarly disadvantaged older children. There was no evidence of age group differences in the relationship between SEP and non-use of dental services.

Table 2.   Odds ratio and (CI95) of having an oral health problem by SEP quintile
 SEP quintiles‡
2
OR
3
OR
4
OR
5 (Low)
OR
  1. †All models adjusted for age of child in months and child gender.

  2. ‡Quintiles based on household annual income, parents’ education and parents’ occupational status, 1 = highest 20%; 5 = lowest 20%.

  3. §p values for contribution of SEP in prediction of oral health measures.

  4. ¶p values for test of interaction between age group and SEP in predicting oral health measures.

  5. ††Children who had cavities, fillings or extractions.

  6. ‡‡Within the previous 12 months.

Caries experience††
 2–3 years1.05 (0.50, 2.19)1.47 (0.76, 2.88)1.87 (0.99, 3.53)3.29 (1.84, 5.90)<0.00050.08
 6–7 years1.08 (0.86, 1.35)1.32 (1.06, 1.64)1.62 (1.31, 2.01)1.79 (1.46, 2.20)<0.0005
Clean teeth < twice a day
 2–3 years1.04 (0.85, 1.27)0.93 (0.77, 1.13)1.16 (0.96, 1.41)1.26 (1.05, 1.52)0.007<0.001
 6–7 years1.25 (1.00, 1.57)1.64 (1.32, 2.04)2.35 (1.90, 2.90)3.40 (2.77, 4.18)<0.0005
No use of dental services‡‡
 2–3 years1.04 (0.79, 1.37)1.18 (0.89, 1.56)1.96 (1.44, 2.67)2.40 (1.73, 3.33)<0.00050.21
 6–7 years1.41 (1.12, 1.77)1.38 (1.09, 1.73)1.75 (1.39, 2.20)2.43 (1.94, 3.05)<0.0005

Table 3 displays associations between the three dichotomous social indicators (accessibility, Indigenous status and NESB) and oral health outcomes. For accessibility, those children in locations with restricted accessibility had a 1.59 times increased odds of caries (95% CI 1.05, 2.42) at age 2–3 years, and a 1.33 times increased odds (95% CI 1.10, 1.60) at age 6–7 years compared with those in accessible locations. However, there was no relationship between accessibility and toothbrushing or service use, nor were there age interactions for any of the oral health outcomes.

Table 3.   Odds ratio and (CI95)† of having an oral health problem by accessibility, Indigenous status and non-English speaking background
 ORTest for association
Test for age interaction
  1. †All models adjusted for age of child in months and child gender.

  2. §p values for contribution of each indicator of social disadvantage in prediction of oral health measures.

  3. ¶p values for test of interaction between age group and social disadvantage in predicting oral health measures.

  4. ††Children who had cavities, fillings or extractions.

  5. ‡‡Within the previous 12 months.

Accessibility
Caries experience††
 2–3 years1.59 (1.05, 2.42)0.0290.19
 6–7 years1.33 (1.10, 1.60)0.003
Clean teeth < twice a day
 2–3 years1.01 (0.85, 1.20)0.9470.48
 6–7 years1.13 (0.95, 1.34)0.179
No use of dental services‡‡
 2–3 years1.26 (0.96, 1.65)0.0890.38
 6–7 years1.17 (0.95, 1.43)0.138
Indigenous
Caries experience††
 2–3 years1.90 (1.08, 3.34)0.0270.92
 6–7 years2.04 (1.50, 2.78)<0.0005
Clean teeth < twice a day
 2–3 years1.65 (1.25, 2.19)<0.00050.32
 6–7 years2.12 (1.55, 2.90)<0.0005
No use of dental services‡‡
 2–3 years1.95 (0.96, 3.94)0.0640.62
 6–7 years1.58 (1.03, 2.44)0.038
Non-English speaking background
Caries experience††
 2–3 years2.32 (1.49, 3.61)<0.0005<0.001
 6–7 years0.90 (0.72, 1.12)0.234
Clean teeth < twice a day
 2–3 years1.17 (1.00, 1.38)0.0530.48
 6–7 years1.07 (0.91, 1.25)0.415
No use of dental services‡‡
 2–3 years1.26 (0.95, 1.68)0.1110.59
 6–7 years1.50 (1.25, 1.80)<0.0005

Compared to non-Indigenous children, Indigenous children had a 1.5 to two-fold increased odds of caries experience, infrequent toothbrushing and non-use of dental services. There was no evidence for differences in these associations by age group. Compared to children from English speaking families, NESB children from the younger cohort had an increased odds of caries (OR = 2.32; 95% CI = 1.63, 3.10) and a trend towards a higher odds of infrequent toothbrushing (OR = 1.17; 95% CI = 1.00, 1.38), while those from the older cohort had increased odds of non-use of dental services (OR = 1.50; 95% CI = 1.25, 1.80). The NESB by age interaction was significant for caries experience, indicating that the associations between NESB and caries was significantly different between the cohorts, whereas the NESB effects were similar for both cohorts for toothbrushing and non-dental service use.

Discussion

This paper presents Australia-wide data on parent-reported dental service use, toothbrushing and caries in two cohorts of children aged 2–3 and 6–7 years respectively. Overall, it provides a somewhat disappointing picture of early life oral health in Australia. At age 2–3 years, 15% of children had attended a dental visit in the last year and less than half had their teeth brushed twice daily, although only 3% were reported by parents to have experienced caries in the last two years. While older children’s oral health behaviours were better, there was still much room for improvement with nearly half having no dental visit in the last year and over a third failing to meet the recommended practices for twice daily toothbrushing. Moreover, 32% were reported to have experienced caries in the last 24 months. The similarity between our estimate of dental service use for 2–3 year olds (15% in the last year) and that reported by Slack-Smith8 using the 1995 National Health Survey (12% ever used a dental service) suggests little improvement in early childhood dental service access over the last decade.

For parent-reported caries, inequalities were evident across all four measures of social circumstances (SEP, accessibility, Indigenous and NESB). For non-use of dental services, inequalities were evident for three measures of social circumstances (SEP, Indigenous and NESB). It is notable that there was no association between accessibility and use of dental services in this study. Other research has found stronger evidence of elevated caries and unmet dental service needs among rural Australian populations.18 It is likely that the effects reported here are attenuated by the small sample numbers resident in more remote locations (e.g. only 4% were classified remote/very remote) and the consequent combining of data into two broad categories. Slack-Smith8 similarly found no evidence of dental service inequalities by rural status when using a dichotomous measure.

This study is the first to demonstrate the presence of significant social disparities in the oral health outcomes of very young (2–3 year old) children in Australia. Such inequalities have been previously identified around school age. The appearance of inequalities in oral health in these very young children is important given the evidence that childhood oral health inequalities can persist into adulthood irrespective of later changes in social position.19 Unfortunately, the LSAC included no measures of oral health in its first wave of data collection (i.e. at 3–19 months of age). It would be beneficial for future longitudinal studies to include oral health outcomes from birth as it is likely that the disparities seen in this study may emerge even earlier. Confirmation of this would advance the argument for oral health promotion initiatives that engage parents of children very early, e.g. via routine post-natal mother-child health services. The associations reported here between disadvantaged status (as defined by SEP and Indigenous background) and toothbrushing highlight the importance of such services focusing on encouraging oral health promoting behaviours from the earliest ages.

Access to two age cohorts presented an opportunity to explore the social patterning of oral health outcomes and behaviours across the two age groups of children. Overall, there was little evidence of differences in the patterning of oral health inequalities by age. For NESB and caries, inequalities were lower in the older cohort compared to the younger cohort. In contrast, there appeared to be a widening of inequalities for SEP and toothbrushing with age, with consistently higher odds ratios at each level of increasing disadvantage for the older compared to the younger cohort. It is feasible that children from more advantaged backgrounds not only preferentially attend public dental services but also may be more likely to adhere to the oral health advice they receive compared with those from less advantaged backgrounds, further exacerbating inequalities. This is of concern as infrequent toothbrushing is likely to predict future caries, suggesting that the caries gap may widen as these children get older. Given the pivotal role that regular exposure to fluoridated toothpaste has in caries prevention,20 the inequalities identified in this oral health behaviour represent an obvious target for future health promotion interventions.

A notable limitation of this study was the reliance on parental report of children’s oral health. Several studies have concluded that parent-reported single-item indicators of their children’s oral health have satisfactory construct validity21,22 and are robust across socio-economic circumstances.23 However, it has also been suggested that the accuracy of such reports is better with increasing disease levels,22 with a tendency towards under-reporting children’s oral health problems when children are very young.21 For both these reasons, it is likely that the current study represents an under-estimation of caries prevalence in the younger cohort. There was no opportunity in LSAC to validate parent report of caries with clinical exams, dental records or medical charts. Future early childhood longitudinal studies should consider incorporating dental clinical examinations in order to strengthen the evidence regarding oral health inequalities. Similarly, the data on service utilization needs to be viewed with caution as around 25% of the data were missing. However, at least in terms of the age interaction, this is unlikely to have influenced the results as there was no difference in response rate across the two cohorts.

A second limitation was the separate modelling of the measures of accessibility, Indigenous status and non-English speaking background. There is overlap between these measures; Indigenous families are over-represented in rural areas, and all three characteristics tend to be associated with more disadvantaged socio-economic circumstances. Thus, it is possible that some of the associations reported here arise from underlying socio-economic circumstances rather than from accessibility or cultural background per se. For example, Aida et al.12 found that the association between regional location and caries in Japanese 3-year-olds was fully accounted for by higher rates of social disadvantage in regional areas compared to major cities. In contrast, Australian studies have found SEP, ethnicity/race and location to make both joint and unique contributions to children’s oral health.9 This highlights the importance of examining alternative sources of social disadvantage separately, particularly for informing the future focus of oral health promotion efforts.

A final major limitation of this study is the cross-sectional nature precluding inferences about causality. Longitudinal studies of the oral health of representative samples of Australian children are rare, and LSAC will provide stronger evidence of the potential causal pathways underlying oral health inequalities as further longitudinal data become available.

This study provides a recent picture of oral health not examined nationally for some time. It uses new data from two age cohorts that are broadly representative of the Australian population, overcoming sampling limitations of previous studies. It shows that children from the disadvantaged groups (as defined by SEP, remoteness, Indigenous background and NESB) were less likely to engage in oral health promoting behaviours and were more likely to have caries. Furthermore, the data suggest that not only do significant inequalities exist at a very early age but they do not appear to improve as children get older. Such findings further strengthen calls for early life oral health prevention and promotion efforts. Interventions promoting the early and twice-daily use of fluoridated toothpaste and that optimize access to preventive oral health services should be developed and delivered in a culturally appropriate manner, with particular consideration given to the specific needs of disadvantaged communities.

Acknowledgements

This work was supported by National Health and Medical Research Council Career Development Awards (237149 to NK, 390136 to JN). All Murdoch Childrens Research Institute staff are supported by the Victorian Government’s Operational Infrastructure Program. This paper uses confidentialized unit record data from Growing Up in Australia, the Longitudinal Study of Australian Children. The study is conducted in partnership between the Department of Families, Housing, Community Services and Indigenous Affairs (FaHCSIA), the Australian Institute of Family Studies (AIFS) and the Australian Bureau of Statistics (ABS). The findings and views reported in this paper are those of the authors and should not be attributed to FaHCSIA, AIFS or the ABS. The authors thank Obioha Ukoumunne and Sebastian Misson for advice on statistics and weighting.

Ancillary