Income‐related inequalities in denture‐wearing by Europeans aged 50 and above
Abstract
doi: 10.1111/j.1741‐2358.2011.00590.x Income‐related inequalities in denture‐wearing by Europeans aged 50 and above
Background: Despite its importance for the planning of future treatment needs and an optimised allocation of health care resources, only little is known about socio‐economic inequalities in denture‐wearing by late middle‐aged and elderly generations.
Objectives: To describe income‐related inequalities in denture‐wearing by elderly populations residing in different European countries.
Material and methods: Data from the Survey of Health, Ageing and Retirement in Europe (SHARE Wave 2) were used to assess income‐related inequalities in denture‐wearing by means of Concentration Indices (CI) for populations aged 50+ from 14 different European countries.
Results: We could identify a significant disproportionate concentration of denture‐wearing amongst the poor elderly populations in Denmark (CI = −0.3534, corresponding to the highest level of inequality), Sweden (CI = −0.3479), Switzerland (CI = −0.2013), Greece (CI = −0.1953), the Netherlands (CI = −0.1413), France (CI = −0.1339), Austria (CI = −0.0974), Czech Republic (CI = −0.0959), Belgium (CI = −0.0947), Germany (CI = −0.0762), Ireland (CI = −0.0575) and Spain (CI = −0.0482, corresponding to the lowest level of pro‐poor inequality). Poland became evident as the only country in which individuals from the upper end of the income scale wear more dentures than their peers from the lower end of the income scale (CI = 0.0379). No significant income‐related inequalities were observable in Italy.
Conclusions: There is considerable income‐related inequality in denture‐wearing by several elderly populations in Europe. Future resource planning for prosthetic care should, thus, specifically distinguish between the treatment needs of different socio‐economic groups within elderly populations.
Introduction
Over the past three decades or so, partial and complete edentulism at a given age has been falling in most industrialised countries1-5. Whilst it has been suggested that the number of people in need of dentures would decrease, there is still unclearity about the extent to which population ageing will antagonise such predictions of reduced prosthetic demand6-9. Less ambiguously, however, population ageing is frequently considered a decisive factor for further increasing expenditures in health care10-13. Despite its importance for the planning of future treatment needs and an optimised allocation of health care resources, only little is known about socio‐economic inequalities in denture‐wearing by late middle‐aged and elderly generations. Such information may, however, enable a better insight into socio‐economic gradients regarding the demand for prosthetic care. Finally, this may give valuable guidance for clinicians and health care decision‐makers.
Previous literature, mostly based on evidence from single countries, has documented the existence of a socio‐economic gradient in oral health. In particular, individuals from the lower end of the socio‐economic scale were shown to have poorer oral health than individuals with higher socio‐economic status14-24. Whilst there still remains debate about the exact causes of such a socio‐economic gradient25-27, inequalities in denture‐wearing by late middle‐aged and elderly generations of in Europe have, to the best of our knowledge, never been investigated. The aim of this study is, thus, to explore such disparities in prosthetic utilisation by persons aged 50 and older from different European countries. It was hypothesised that denture use is more prevalent amongst individuals from the lower end of the income scale as compared with individuals from the upper end of the income scale.
Materials and methods
Dataset
The analysis presented here is based on data from Wave 2 of the Survey of Health, Ageing and Retirement in Europe (SHARE). SHARE is the first European data set to combine extensive cross‐national information on socio‐economic status, health and family conditions of late middle‐aged and elderly populations from 14 different European countries. This individual‐level longitudinal survey is modelled closely on the US Health and Retirement Study (HRS). Eligible as study participants were all household members aged 50 and over. Depending on available registers of individuals, different types of survey sample design were used, including simple random sampling from national population registers (e.g. Denmark and Sweden), multi‐stage sampling using regional/local population registers (e.g. Italy and Spain) and single/multi‐stage sampling on the basis of telephone directories followed by screening in the field (e.g. Austria and Switzerland)28.
Data of Wave 2 were collected in 2006–07 on the basis of a computer‐assisted personal interview as well as a self‐completion paper and pencil questionnaire. For standardisation of the fieldwork procedures across countries and to minimise the occurrence of non‐sampling errors (like unit and item non‐response), professional survey agencies were selected in all participating countries. Agencies had to follow a common set of protocols that included, for example, the length of the fieldwork period, the use of advance and follow‐up letters and the set‐up of general rules for the management of the fieldwork. For data quality control, all agencies had to certify that a minimum of 10% of each interviewer’s completed interviews were verified by supervisory personnel. Verification consisted of calling the respondent by telephone and re‐asking questions from the interview. Adherence to these quality control guidelines was ensured by additional checks made by the SHARE coordination team28. More details about the process of data collection are available on the SHARE webpage (see http://www.share‐project.org/).
Measures of denture‐wearing and income
The measure of denture‐wearing in this study is a dichotomous variable that reports whether an individual has responded with ‘yes’ or ‘no’ to the SHARE question ‘Do you use dentures?’. This variable neither distinguishes between complete and partial dentures, nor between upper and lower jaw, nor between the existence of one or two dentures per patient. Whilst it may, hence, be considered only a proxy variable for denture status with limitations regarding internal validity, the strength of SHARE is, above all, its external validity (i.e. reliability and representativeness for many European countries). In total, SHARE Wave 2 contains information on utilisation of denture‐wearing by 33 417 individuals aged between 50 and 105 years who represent the late middle‐aged and elderly populations in 14 different European countries. A description of denture‐wearing across populations aged 50+ in Europe on the basis of the same database has already been published before29.
Income (in €) is measured as net monthly household income and specified according to the OECD‐modified equivalence scale, which was first proposed by Hagenaars et al.30 and later adopted by the Statistical Office of the European Union (EUROSTAT). This discount scale takes account of household size and age of household members. Specifically, the household head is assumed to contribute more to household consumption than other household members. Moreover, children aged 14 years or younger are assumed to contribute less to household consumption than household members who are older than 14 years (see formula F.1).
Identification of socio‐economic inequalities
Socio‐economic inequality in denture‐wearing is identified by means of the Concentration Index (CI)31, 32. This index quantifies the degree of relative socio‐economic inequality in a health variable33, 34. It has, for example, been used to measure and compare the degree of socio‐economic‐related inequality in child mortality35, child immunisation36, child malnutrition37, adult health38, health subsidies39 and health care utilisation40, 41.
Formally, the CI can be expressed with reference to the covariance between the health variable and the fractional rank according to the socio‐economic distribution (see formula F.2)31, 32. The value of the CI is bounded between −1 and +1. For ease of interpretation, the definition of the CI for the purpose of our study is as follows: if the CI is zero, it indicates that there is no income‐related inequality regarding denture‐wearing; a positive (negative) value of the CI indicates disproportionate concentration of denture‐wearing amongst the rich (poor) – in other words, a positive (negative) value of the concentration index means that – in comparison with other income groups – more dentures are worn by the rich (poor). This is also referred to as ‘pro‐rich’ (‘pro‐poor’) income‐related inequality. All data analysis of this study was carried out with the software package STATA/SE 10.1 (StataCorp, College Station, TX, USA). CI values are adjusted for age and sex with the exception of Ireland for which SHARE Wave 2 does not offer a weighting variable.
Formula F.1: equivalence income according to the OECD‐modified equivalence scale
where
μ = 1: weighting for the household head.
ϕ = 0.5: weighting for other household members aged older 14 years.
κ = 0.3: weighting for other household members aged 14 years or younger.
Formula F.2: Concentration Index
: population proportion of denture‐wearing.
h i : denture status of individual i.
r i : the fractional income rank of individual i.
Results
Table 1 shows the demographic characteristics of the study sample by respondent’s country of residence. According to numbers of observation, the participation rates in SHARE Wave 2 (release 2.3.1) are as follows: 7.53% of all respondents aged 50+ are from Germany, 8.08% from Sweden, 7.79% from the Netherlands, 6.49% from Spain, 8.71% from Italy, 8.54% from France, 7.54% from Denmark, 9.18% from Greece, 4.23% from Switzerland, 9.19% from Belgium, 8.22% from the Czech Republic, 7.24% from Poland, 3.30% from Ireland and 3.95% from Austria. Average participant’s age ranges between 64.6 years (Poland) and 67.4 years (Austria). Population proportions of women vary between 53.2% (Germany) and 59.0% (Austria).
| Country (observations) | Age (standard deviation) | Proportion of women |
|---|---|---|
| Germany (n = 2526) | 65.6 (9.4) | 53.2% |
| Sweden (n = 2711) | 67.1 (10.0) | 53.4% |
| The Netherlands (n = 2611) | 64.5 (9.7) | 54.0% |
| Spain (n = 2177) | 67.3 (10.7) | 54.2% |
| Italy (n = 2921) | 66.3 (9.4) | 54.4% |
| France (n = 2865) | 65.9 (10.7) | 56.1% |
| Denmark (n = 2527) | 64.9 (10.4) | 54.1% |
| Greece (n = 3077) | 65.4 (10.3) | 54.8% |
| Switzerland (n = 1420) | 65.6 (10.4) | 55.1% |
| Belgium (n = 3084) | 65.7 (10.3) | 53.9% |
| Czech Republic (n = 2755) | 64.7 (9.6) | 57.0% |
| Poland (n = 2429) | 64.6 (10.0) | 55.9% |
| Ireland (n = 1106) | 64.7 (9.6) | 53.8% |
| Austria (n = 1324) | 67.4 (9.4) | 59.0% |
Table 2 shows country‐specific tertiles of equalised income and population proportions of ‘denture‐wearing’ decomposed by tertiles of equalised income. The income tertile with the lowest level across all countries is found in Poland (average income of 1st tertile = 140.79 €), whereas the highest income tertile is found in Italy (average income of 3rd tertile = 11 334.82 €). Moreover, Table 2 reports the ratio between the average income of the 1st tertile and the average income of the 3rd tertile, which can be considered an indicator of income disparity (the closer to zero the value of the ratio, the higher the level of disparity). According to Table 2, the highest level of income disparity exists in Italy and the lowest level of income disparity in Austria.
| Country (observations) | Equalised income (in €) | ‘Denture‐wearing’ per tertiles within the income distribution | ||||
|---|---|---|---|---|---|---|
| 1st tertile | 2nd tertile | 3rd tertile | 1st tertile | 2nd tertile | 3rd tertile | |
| Germany (n = 2521) | 860.94 | 1571.39 | 10 289.50 | 59.53% | 45.39% | 44.52% |
| Ratio of 1st to 3rd tertile: 0.08 | Ratio of 1st to 3rd tertile: 1.34 | |||||
| Sweden (n = 2704) | 1095.67 | 1664.97 | 4589.49 | 25.72% | 9.36% | 5.10% |
| Ratio of 1st to 3rd tertile: 0.24 | Ratio of 1st to 3rd tertile: 5.04 | |||||
| The Netherlands (n = 2597) | 849.58 | 1466.87 | 6 637.02 | 61.38% | 47.12% | 30.07% |
| Ratio of 1st to 3rd tertile: 0.13 | Ratio of 1st to 3rd tertile: 2.04 | |||||
| Spain (n = 2175) | 320.10 | 653.32 | 5033.54 | 47.26% | 40.58% | 31.53% |
| Ratio of 1st to 3rd tertile: 0.06 | Ratio of 1st to 3rd tertile: 1.50 | |||||
| Italy (n = 2915) | 612.35 | 1334.73 | 11 334.82 | 37.14% | 27.57% | 34.32% |
| Ratio of 1st to 3rd tertile: 0.05 | Ratio of 1st to 3rd tertile: 1.08 | |||||
| France (n = 2823) | 747.37 | 1447.12 | 6782.71 | 44.10% | 34.10% | 21.62% |
| Ratio of 1st to 3rd tertile: 0.11 | Ratio of 1st to 3rd tertile: 2.04 | |||||
| Denmark (n = 2522) | 1164.22 | 1973.53 | 8984.45 | 51.72% | 19.73% | 10.85% |
| Ratio of 1st to 3rd tertile: 0.13 | Ratio of 1st to 3rd tertile: 4.77 | |||||
| Greece (n = 3074) | 406.89 | 723.81 | 4851.87 | 39.52% | 26.11% | 16.90% |
| Ratio of 1st to 3rd tertile: 0.08 | Ratio of 1st to 3rd tertile: 2.34 | |||||
| Switzerland (n = 1415) | 1072.51 | 2204.46 | 11 074.96 | 43.86% | 31.18% | 16.80% |
| Ratio of 1st to 3rd tertile: 0.10 | Ratio of 1st to 3rd tertile: 2.61 | |||||
| Belgium (n = 3082) | 793.46 | 1243.87 | 5091.50 | 64.51% | 55.75% | 43.82% |
| Ratio of 1st to 3rd tertile: 0.16 | Ratio of 1st to 3rd tertile: 1.47 | |||||
| Czech Republic (n = 2751) | 290.16 | 412.26 | 2767.66 | 53.86% | 44.82% | 33.14% |
| Ratio of 1st to 3rd tertile: 0.11 | Ratio of 1st to 3rd tertile: 1.63 | |||||
| Poland (n = 2418) | 140.79 | 260.30 | 1465.28 | 49.87% | 61.07% | 58.72% |
| Ratio of 1st to 3rd tertile: 0.10 | Ratio of 1st to 3rd tertile: 0.85 | |||||
| Ireland (n = 1098) | 620.85 | 1559.20 | 10 677.27 | 62.76% | 58.49% | 45.32% |
| Ratio of 1st to 3rd tertile: 0.06 | Ratio of 1st to 3rd tertile: 1.38 | |||||
| Austria (n = 1322) | 786.15 | 1209.39 | 1896.61 | 71.63% | 60.36% | 49.87% |
| Ratio of 1st to 3rd tertile: 0.41 | Ratio of 1st to 3rd tertile: 1.44 | |||||
Alongside increasing income ranges, Table 2 also shows a decrease in the population proportions wearing a denture in Germany, Sweden, the Netherlands, Spain, France, Denmark, Greece, Switzerland, Belgium, Czech Republic, Ireland and Austria. For Italy and Poland, Table 2 shows less consistent gradients of denture‐wearing alongside increasing income groups: in Italy, the lowest proportion of denture wearers is found in the 2nd income tertile, whilst in Poland it is found in the 1st income tertile. Moreover, Table 2 reports the ratio between the proportion of denture wearers in the 1st income tertile and of denture wearers in the 3rd income tertile. This can be considered an indicator of socio‐economic disparities in denture‐wearing – if the value of the ratio is greater (smaller) than 1, it indicates a higher (lower) proportion of denture wearers in the 1st tertile in comparison with the 3rd tertile. Accordingly, Table 2 indicates that denture‐wearing is found more frequently in the 1st than in the 3rd income tertile in all countries except Poland. The highest level of socio‐economic disparity in denture‐wearing is identified for Sweden, which has a 1st to 3rd tertile ratio of 5.04, indicating a five times higher proportion of denture wearers in the 1st as compared with the 3rd income tertile.
Statistical significance regarding income‐related inequalities in denture‐wearing is, finally, established in Table 3 that shows the according concentration indices (CI). With the exception of Italy and Poland, all CI are statistically significantly negative, indicating pro‐poor income‐related inequality in denture‐wearing. In decreasing order of pro‐poor inequality, the countries rank as follows: Denmark (highest level of inequality), Sweden, Switzerland, Greece, the Netherlands, France, Austria, Czech Republic, Belgium, Germany, Ireland and Spain. The CI for Italy is shown to be negative but not statistically significant, whereas the CI for Poland is statistically significant and has a positive sign. This indicates that, in Poland, more dentures are used by the late middle‐aged and elderly rich than by individuals located at the lower end of the income distribution.
| Country | Concentration index (95% CI) |
|---|---|
| Germany | −0.0762(−0.1023; −0.0501) |
| Sweden | −0.3479(−0.3999; −0.2959) |
| The Netherlands | −0.1413(−0.1675; −0.1151) |
| Spain | −0.0482(−0.0829; −0.0136) |
| Italy | −0.0308 (−0.0622; 0.0007) |
| France | −0.1339(−0.1651; −0.1027) |
| Denmark | −0.3534(−0.3873; −0.3195) |
| Greece | −0.1953(−0.2388; −0.1517) |
| Switzerland | −0.2013(−0.2484; −0.1544) |
| Belgium | −0.0947(−0.1150; −0.0745) |
| Czech Republic | −0.0959(−0.1255; −0.0664) |
| Poland | 0.0379 (0.0161; 0.0597) |
| Ireland | −0.0575(−0.0927; −0.0222) |
| Austria | −0.0974(−0.1243; −0.0704) |
- Values in bold indicate that income‐related inequality is statistically significant at the 5% level; age‐ and sex‐adjusted results (except for Ireland).
Discussion
On the basis of cross‐sectional survey‐based data (SHARE wave 2), this study describes income‐ related inequalities in denture‐wearing by Europeans aged 50 years and above. The findings indicate a higher prevalence of denture‐wearing amongst late middle‐aged and elderly individuals from the lower end of the income scale in Germany, Sweden, the Netherlands, Spain, France, Denmark, Greece, Switzerland, Belgium, Czech Republic, Ireland and Austria. Only Poland becomes evident as a country in which individuals from the upper end of the income scale wear more dentures than their peers from the lower end of the income scale. The present study, thus, gives mostly consistent evidence for income‐related inequalities in the prosthetic demand by late middle–aged and elderly populations residing in Europe.
One potential interpretation of such socio‐economic disparities draws on the assumption that denture‐wearing mirrors edentulism within different socio‐economic groups to a large extent. For example, earlier evidence from New England suggests that about 90% of edentulous elderly are wearing a denture42. Analogously and in line with previous literature on socio‐economic inequalities in oral health14-24, most of our findings could be seen as evidence for a socio‐economic gradient in the dentition of populations aged 50+ in Europe, i.e. that individuals from the lower end of the income scale suffer more from tooth loss than individuals from the upper end of the income scale. However, we also found a disproportionate concentration of denture‐wearing amongst the rich population aged 50+ in Poland. On the one hand, this may be interpreted as evidence for higher tooth loss at the upper end of the income scale in Poland. On the other, such an inverse income‐related disparity (in comparison with other European countries) could also display a reduced probability of denture use despite considerable tooth loss at the lower end of the socio‐economic scale. For reasons still to be explored (edentulism itself is not measured in SHARE), such an accumulation of untreated edentulism may be more distinct in Poland than in other countries. Viewed in this light, socio‐economic inequalities in tooth loss amongst late middle‐aged and elderly populations in Europe may be even more pronounced than those observed for denture‐wearing in our study.
The present paper may contribute to the discussion about dental public health policies for reducing inequalities. On the one hand, the comparisons across countries suggest that income‐related inequality in denture‐wearing does not necessarily need to be increasing with level of income disparity, vice versa. For example, Italy was identified as having a comparably high level of income disparity but, at the same time, did not have any income‐related inequality in denture‐wearing. Countries such as Denmark and Sweden were shown to have comparably low levels of income disparity despite relatively high degrees of income‐related inequality in denture‐wearing. In addition, these two countries had about the same level of income‐related inequality in denture‐wearing despite Denmark exhibiting a considerably greater income disparity than Sweden. These findings could be interpreted in the sense of earlier evidence which suggests that a simple redistribution of income from the rich to the poor may not necessarily lead to a reduction of inequalities43.
On the other hand, specific characteristics of health policy and dental health services in the studied countries could provide potential explanations for the findings of the present paper. Inequalities in denture‐wearing may be due to the extent of dental insurance coverage and features of the dental workforce. By way of example, Sweden, Denmark and Belgium each have a dental workforce density of about 0.8 dentists per 1000 inhabitants44, 45. Whilst Sweden and Denmark have comparably high proportions of patient’s out‐of‐pocket payments in dental care expenditures (Sweden: 63%; Denmark: 69%), in Belgium only 34% of dental care expenditures are covered by out‐of‐pocket payments44, 45. Interestingly, Belgium was shown to have a much lower degree of (pro‐poor) income‐related inequality in denture‐wearing than Denmark and Sweden, the latter two countries themselves both having about the same degree of inequality. Such a finding may suggest that inequalities in oral health increase alongside increasing shares of patient’s out‐of‐pocket payments in dental care expenditures.
Similarly to Denmark, furthermore, 69% of dental care expenditures in Poland are covered by out‐of‐pocket payments44, 45. However, the dental workforce density in Poland amounts to only about 0.3 dentists per 1000 inhabitants44, 45. In the context of the present paper’s results, i.e. Poland having a disproportionate use of dentures amongst individuals located in the upper end of the income scale and the reverse being found for Denmark, this could indeed be interpreted in the sense that a comparably low dental workforce density may translate into a reduced accessibility of dental services and, finally, result in a reduced probability of denture use (despite oral health need) at the lower end of the socio‐economic scale.
The following limitations surrounding the present study should be mentioned. First, the analysis is based on cross‐sectional data and does not have a causal interpretation. Therefore, it is not possible to isolate the exact causation of the disparities that were found in this study. However, as soon as further waves of SHARE become available in the future this may enable better insights into the reasons for such socio‐economic inequalities. Second, the data used are survey‐based and thus may be subject to recall bias. Third, our dependent variable for denture‐wearing may be considered a proxy variable only and a potential source of imprecision in our results. Nevertheless, because there is currently no comparable epidemiological database available, a high degree of uniqueness is attributed to SHARE as a source for investigating oral health disparities across Europe. Despite some potential limitations regarding internal validity, SHARE facilitates a so far unavailable degree of external validity, i.e. comparability across countries.
In conclusion, this study is the first to investigate income‐related inequalities in denture‐wearing by late middle‐aged and elderly populations residing in European countries. The findings suggest considerable income‐related inequality in denture use in the majority of European countries. Future predictions of demands for prosthetic care should specifically distinguish between different socio‐economic groups.
Acknowledgements
This article uses data from SHARE Wave 2 (Release 2.3.1, published July 29th 2010). The SHARE data collection was primarily funded by the European Commission through its 5th and 6th framework programs (project numbers QLK6‐CT‐2001‐00360; RII‐CT‐2006‐062193; CIT5‐CT‐s005‐028857). Additional funding by the US National Institute on Aging (grant numbers U01 AG09740‐13S2; P01 AG005842; P01 AG08291; P30 AG 12815; Y1‐AG‐4553‐01; OGHA 04‐064; R21 AG025169) as well as by various national sources is gratefully acknowledged (see http://www.share‐project.org for a full list of funding institutions).
