Inequality of Oral Health in a Life-Course Perspective

Geriatric dentistry, gerodontics, is the branch of dental care dealing with older adults involving the diagnosis, prevention, and treatment of problems associated with normal aging and age-related diseases as part of an interdisciplinary team with other healthcare professionals. Prosthodontics is the dental specialty pertaining to the diagnosis, treatment planning, rehabilitation, and maintenance of the oral function, comfort, appearance, and health of patients with clinical conditions associated with missing or deficient teeth and/or oral and maxillofacial tissues using biocompatible materials. Periodontology, or Periodontics, is the specialty of oral healthcare that concerns supporting structures of teeth, diseases, and conditions that affect them. The supporting tissues are known as the periodontium, which includes the gingiva (gums), alveolar bone, cementum, and the periodontal ligament. Oral biology deals with the microbiota and their interaction within the oral region. Research in oral health and systemic conditions concerns the effect of various systemic conditions on the oral cavity and conversely helps to diagnose various systemic conditions.


Introduction
Viewing health and disease in a life-course perspective has gained scientific interest recently (Kuh & Ben-Shlomo, 1997). Panel studies are scarce but other designs come close to being able to follow health and disease through life (Kuh & Ben-Shlomo, 1997). There are three major perspectives in life-course research: One line emphasizes the importance of life-style and deprivation in childhood for adult chronic disease. This research investigates environmental conditions and experiences through prenatal life, infancy, childhood and adolescence that may make individuals more susceptible to developing adult chronic disease. Pearce et al., 2004 studied the effect of birth-weight, early diet, use of comforter and social status on oral health of young adults, but found only effect of social status in the expected direction. Nicolau and co-workers found a relationship between several biological factors and caries among adolescents (Nicolau et al., 2003(Nicolau et al., , 2007. A second line of research assumes biological programming during critical periods of development either during pregnancy or in early life (Barker, 1994). A third line of research suggests an accumulation of risk through the life course. Accumulation of risk is different from programming in that it does not require the notion of a critical period. This approach explicitly places more emphasis on a greater range of biological and social experiences in childhood, adolescence and in early adulthood than either the life style or programming models. There are reasons to believe that adult oral health is affected through a range of life-course mechanisms. The present work leans on the third perspective arguing that oral health is continuously exposed to environmental and behavioral risks that lead to accumulated plaque in the mouth and diseases in the dental tissues (Fejerskov & Kidd, 2008). On a population basis the vast majority of children are born with a good oral health. Exposure to different life-styles and nutrional and hygienic conditions appears as a threat to oral health through life (Holst & al., 2007). The global picture of oral health is patterned by variation in living conditions and variation in life-course patterns of oral health (Petersen et al., 2006). A number of studies have described oral health of populations in repeated cross-sectional studies (Schuller & Holst, 1998, Kelly et al., 2000Skudutyte-Rysstad &Eriksen 2007, Krustrup et al., 2008. These studies provide valuable information about background related changes in oral health conditions at certain points of time. It has for example been shown that edentulousness has a much lower incidence now than 30 years ago (Petersen et al., 2004;Holst, 2008). The main explanations for this are improved standard of living, availability of fluoride toothpaste and more accessible dental services. Despite the improvement, social status still affects oral health even though recent research indicates that this relationship has become weaker in some countries (Holst, 2008). There is reason to believe that avoiding edentulousness and maintaining oral health requires a life-long attention to healthy diet, oral health promotion, oral hygiene and preventive dental services. In this chapter the influence of social status on clinical aspects of oral health is assessed from childhood through adolescence to adulthood in the same birth-cohorts in Norway. This unique possibility was made possible through a careful design of a series of crosssectional studies in the counties of Trøndelag in Norway (Schuller &Holst, 1998). On a population basis the vast majority of children are born with a good oral health. Exposure to different life-styles and nutrional and hygienic conditions forms a threat to oral health through life (Holst & al., 2007). The global picture of oral health is patterned by variation in living conditions and variation in life-course patterns of oral health (Petersen et al., 2006). A number of studies have described oral health of populations in repeated cross-sectional studies (Schuller & Holst, 1998, Kelly et al., 2000Skudutyte-Rysstad &Eriksen 2007, Krustrup et al., 2008. These studies provide valuable information about background related changes in oral health conditions at certain points of time. It has for example been shown that edentulousness has a much lower incidence now than 30 years ago (Petersen et al., 2004;Holst, 2008). The main explanations for this are improved standard of living, availability of fluoride toothpaste and more accessible dental services. Despite the improvement, social status still affects oral health even though recent research indicates that this relationship has become weaker in some countries (Holst, 2008). There is reason to believe that avoiding edentulousness and maintaining oral health requires a life-long attention to healthy diet, oral health promotion, oral hygiene and preventive dental services.

Material and methods
The material comprised data from independent random samples of three birth-cohorts living in the counties of Sør-and Nord-Trøndelag in 1983. The birth-cohorts were 1929-1938, 1939-1948and 1959-1960 1929-1938 and 1959-1960 birth-cohorts in Nord-Trøndelag only, who were then 67-78-year-old and 46-47-year old ( Table 1). The age specific sample size for each of the participating counties were 500 in 1983 and was reduced to 250 in 2006. The sample in the two-year age-group 46-47 was 100 persons. The methods of data collection comprised standardized clinical measurements and selfadministered questionnaires (Baerum et al., 1985;Schuller and Holst, 1998). In 1983 and in 2006 ten and two calibrated dental teams, respectively, collected the data. Two senior researchers (DH and AAS) followed and guided the procedures in order to secure standardized conditions and comparability among the surveys. The first Trøndelag study in 1973 started as part of the first WHO International Collaborative Study survey (Arnljot et al., 1985). The study was also repeated in 1994 but not reported here. Calibration exercises were conducted each study year. Calibration was performed for paired examiners and intraexaminer variability was high for the DMF index (r>0.92). Inter-examiner agreement was exercised until r> 0.85 between all pairs, and the results otherwise found satisfactory (Baerum et al., 1985;Holst et al., 2007). The examinations took place at the public dental clinics of the South-and Nord-Trøndelag counties. Permission was granted by public authorities and by the participants' informed consent. All necessary permissions were given throughout the study period and by the participants' informed consent. In 2006 the study was approved by the Regional ethical committee Middle of Norway and approved by the Norwegian Council of Research. In the present study the outcome variables were number of sound teeth and toothsurfaces (ST, SS) and sound + filled teeth and toothsurfaces (SFT, SFS) and DMFT and DMFS index. DMFT and DMFS are the sums of DT/S, MT/S and FT/S, where DT/S is defined as the number of teeth/toothsurfaces with primary and secondary caries, including root and coronal caries. Only caries with a distinguishable break in the surface was recorded. Missing surfaces is the number of missing tooth surfaces irrespective of cause. FS is the number of surfaces filled, both root and coronal restorations, including all types of filling materials and crowns. The clinical examination comprised recording of the condition of the visible part of the tooth. The analyses were based on 28 teeth excluding third molars. As part of the study in 1983 twelve questions were asked about social and dental conditions when the sampled persons were ten years old. Social status was measured in two ways: By father's and mother's number of natural teeth in three categories: Many own teeth (2), some own teeth (1) and no own teeth (0). The variables were summed and dummy variables constructed (Table 2, column 4). Eleven other questions were asked about oral health environment at age 10 ( Table 2). The questions comprised whether the families had rules for eating sweets, tooth brushing habits, advice about oral health from teachers, school nurse/medical doctor and school dentist, fathers and mothers dental status, visits to a dentist during preschool and school age, parents control of tooth brushing, use of toothpicks and dental floss. Advice from teachers, school nurse/medical doctor and school dentist were collapsed into an index called advice about oral health (Table 2, column 3). Visits to a dentist during pre-school and school age were summarized to yearly and not yearly. The questions comprising whether the families had rules for eating sweets, tooth brushing habits were combined into parents attention and dummy variables constructed. Sex was included in the meaning of a social construct assuming females to be more engaged in health and oral health behaviors. Social status in early adulthood was measured as number of years with formal education. Length of education was divided into four quartiles each comprising 25 % of the samples. The first quartile comprised the 25 per cent of the sample with the shortest education, the second quartile the 25 per cent of the sample with the second shortest education, the third quartile the 25 per cent with the second longest education and the fourth quartile comprised the 25 per cent of the sample with the longest education. The quartiles thus represent the distribution of length of education in equally sized groups. Using quartiles eliminates the problem often faced with measuring length of education that the length of education the population changes over time. Length of education was transformed into dummy variables ( Table 2, column 4). Four question of oral hygiene practices (1983) were added into oral health behavior index and dummy variables constructed.  (0) Toothbrushing habits Twice a day (1) Once or less (0) Advice from teacher Yes (1) no (0) Advice about oral health Advice from doctor/nurse Yes (1) no (0) Advice (1-3) Advice (1) Advice from school /district dentist Yes (1) no (0) No advice (0+0+0) No advice (0) Father's dental status Many teeth (2), few (1) none (0) Parents' dental status Mother's dental status Many teeth (2), few (1) none (0) Many (4) Dummy parents' dental status many Few (1-3) Dummy parents'dental status few None (0) Reference category Visited a dentist pre school Yearly (2) a few times (1) never (0) Dental care at age 10 Visited a dentist during school Yearly (2) a few timel (1) never (0) Yearly (4) Yearly (1), Not yearly (0) A few times (1-3) Never (0) Parents controlled toothbrushing Often/daily (2) a few times (1) never (0) Parents' attention Used toothpicks Often/daily (2) a few times (1) never (0) Daily (5-6) Dummy parents attention daily Used dental floss Often daily/ (2) a few times (1)

Analysis
In the first part of the analysis social status together with the early oral health environment and sex were related to oral health variables in 1983 by multiple regressions. Since all the dependent variables were measured on the same scale (tooth surfaces, range 1-128), the regression coefficients can be interpreted directly as effects of the independent variables in number of surfaces. For the second part of the analysis the data files from 1983 and 2006 were combined to one data file in order to study whether the impact of social status changed during this period. Multiple regression analysis was used, and the level of significance was p= 0.05. Associations nearly reaching significance (0.07>p>0.05) are shown.

Results
The distribution of the independent variables according to age-groups is shown in Table 3.
The table provides a picture of how the oral health environment at age ten years varied between the birth-cohorts. The youngest birth-cohort had the best level of oral health environment.   Table 5 shows the impact of the independent variables on the outcome variables. Having positive oral health behaviors at age ten increased the likelihood of more sound surfaces (SS), more filled surfaces (FS) and more surfaces with caries experience (DMFS) at age 23-24. If parents had many own teeth the 23-24-years-olds had 5.0 more sound surfaces (SS) and 5.0 less surfaces with caries experience (DMFS). Length of education was statistically significantly related to the D-M-F-S variables in the expected direction. Regular dental care was related to mean number of surfaces with untreated decay (DS). Good oral health behavior was statistically significant related to more sound surfaces (SS) and less filled surfaces (FS). The variables in the model explained from 4 -13 % of the variation in the dependent variables. Table 6 shows that parents' dental status and yearly dental visits at age ten had a statistically significant impact on several of the oral health variables. Having parents with many of their own teeth at age ten the 35-44 year olds had more functional surfaces (SFS), less missing surfaces (MS) and surfaces with caries experience (DMFS). Women had less sound surfaces (SS) and more filled surfaces (FS) and DMFS than men when they were 35-44-years-old in 1983. The longer the education, the better the values of the oral health indicators were; the differences between the quartiles of education were big. Dental care last year had a statistical significant influence on DS, SFS, MS and FS. The explained variation varied from 11 -27 %. Table 7 shows that parents' dental status at age ten years had a statistically significant effect on the oral health variables. In addition tooth brushing and dental care at age 10 had a significant effect on untreated caries (DS). Women had more sound surfaces (SS), less untreated decay (DS), and more filled (FS) and DMFS than men had. Length of education had a significant effect on all oral health variables except untreated caries (DS). Oral health behavior had a similar effect, while regular dental care also affected untreated caries (DS).

The influence of social status from 1983 to 2006
For this part of the analysis data files from 1983 and 2006 were combined into one datafile. An interaction term between social status measured as length of education and study year was included in order to see if the effect of length of education was important both years or only one of the years. -cohort 1959-60, age 23-24 and 46-47 in the combined file Table 8 shows that the interaction between length of education and study year was not significant with the exception of the effect on sound surfaces, where the persons in the second lowest education group had kept nearly 15 more sound surfaces than the lowest group. Gender and regular dental care had an independent effect on mean number of sound surfaces (DS). Oral health behavior had a significant effect on oral health variables. Table 9 shows that the effect of social status mesasured as length of education was not dependent upon which year it was measured except that the persons in the second highest education quartile on average had more surfaces with untreated decay compared with persons in the lowest quartile. In this birth-cohort gender and particularly regular dental care had an effect on several of the outcome variables. Functional surfaces increased by nearly 34 surfaces among persons with visits to the dentist compared to those without. Likewise the average number of FS and MS were much higher and lower, respectively, among those with regular dental care.

Discussion
This study has shown that social staus and oral health environment in childhood was important for adults' oral health during the most of the 20. century. The attention from parents and the local environment lead to a better oral health outcome in adulthood. Social status measured by length of education was also a personal resource that guided choices leading to better oral health. The longer the education the better the oral health was. Regular dental visits were important especially for the eldest birth-cohort. Good oral health behaviors early and during adulthood were also important for oral health. Effects of more than 30 surfaces were found on indicators like missing and functional tooth surfaces. When the birth-cohorts were followed from 1983 to 2006 social status had an effect in both 1983 and 2006. Judged by the number of tooth surfaces the difference between social status groups had not increased by 2006. The latter observation deserves a critical comment. The cumulative DMFS measure is sensitive to increased levels of risk factors in the sense that more surfaces can be affected, until saturation is reached. When lower risk levels occur, the DMFS figures cannot decline within the same birth-cohort. The Missing, Filled and Sound indicators (the DMF index) cannot reverse. Only the number of decayed surfaces can reverse (Holst and Schuller, 2000). In the present study the average number of decayed toothsurfaces was significantly reduced and indicated a lower level of recent risk (Holst et al., 2007). This is a serious limitation. With regard to estimating the influence of social status and other explanatory variables, a reduced effect can thus not be shown, and it can only be concluded that the effect of social status did not increase from 1983 to 2006. In a cohort analysis of the relationship between social status and average number of DMFT in 35-44-  year-olds in 1983 and a new cohort of 35-44-year-olds in 2006 from the same material, the relationship between social status and number of present teeth had disappeared in 2006, and the relationship between social status and average number of DMFT and DMFS was significantly reduced (Holst et al., 2007). This shows that the DMF index can be used in cross-sectional research comparing birth-cohorts of the same age; the index has serious limitations in semi-longitudinal research. It is important to draw attention to the different dimensions that the chosen oral health indicators reflect. The indicators SS and SFS reflect oral health and function and high and increasing values represent positive expressions of oral health. MF and FS are negative expressions of oral health and high and increasing values show reduced oral health. These treatment indicators have limitations since they do not include repeated treatment in the same teeth. DMFS (or DMFT) are a summarized expression of untreated and treated disease, and the values may be difficult to interpret because the indicators of the index move in contrasting directions over time. It is important that that researchers in oral epidemiology engage in finding new ways measures of disease activity that are different from measures of treatment activity. There are a number of threats to reliability and validity of the data when surveys are repeated and the same variables are used over time, and different birth-cohorts are exposed to the same procedures. Concepts of behavioral norms and interpretation of clinical symptoms change, and treatment criteria change (Gimmestad& Holst, 2003). Most of the questions in the present surveys, however, were about factual events and clear to the respondents; some memory bias among the respondents with regard to events at age 10 years should be expected. These are measurement errors that increase the variance of the variables and reduce the discriminative ability of the statistical tests. Even though of one of the authors (DH) was present at all the surveys and has acted as the golden standard, it is difficult to avoid flow in the translation of the standardized criteria. The results from the present study have a limited statistical inference with regard to the size of the population the results may be generalized to. On the other hand when it comes to modeling social processes generalization is based on how validly the model catches the specific underlying social processes. It was not the intention to explain all the variation in the dependent variables. It is interesting to notice that R 2 was high in the oldest cohort. It cannot be settled whether this is a cohort or an age effect. Probably it is both, assuming that age reflects the cumulative exposure to plaque during the life-course, and the later born birth-cohorts have experienced a different environment that will result in a better oral health. There are reasons to believe that our data and the model have caught some of those social processes that were important for oral health and its development over time. Other and nationally representative Norwegian data support the finding of a more equally distributed oral health (Holst et al., 2007;Skudutyte andEriksen, 2007, Holst &Skau, 2010). The Trøndelag studies started at a time where data on oral health and its determinants were scarce. In hind-sight these studies have provided opportunities for valuable descriptions and explanations of the changes in oral health. The Norway is considered to have had a homogeneous population compared to many other countries (Krokstad & Westin, 2004). Yet, the demography, the size of the country and the arctic location has resulted in cultural and distributional differences. Living conditions and social disparities have to a large extent affected oral health of the population previously in Norway (Arnljot et al., 1985). During the last decade larger income differences have been observed that might have lead to increased social inequalities in both oral health and demand for dental services (Krokstad 2004). That seems not to have occurred. Crosssectional data will typically focus on cross-sectional social differences and discuss these with limited insight over time. Often will the lead time between exposure and result be overlooked. Panel data and data with the present analytical potential can detect whether or not a social problem is increasing or decreasing. It cannot be ignored that the results of this study can be ascribed to welfare policies across a number of living conditions in Norway. The public dental service with a population responsibility and out-reach services in this country is an example of one such public policy that have contributed to increasing public awareness of oral health as a value. A high level of public awareness may be expected to influence both the promotion of oral health and accessible adequate dental care. A lifecourse perspective provides an opportunity understand oral health over time. The present work supports the assumption that oral health is continuously exposed to environmental and behavioral risks that affect life-time oral health.

Conclusion
The purpose of the work was to study the influence of 1) the oral health environment at age 10, 2) of adolescent and adulthood dental behaviors and of 3) social status on oral health in three birth-cohorts later in life (1983) and in 2006 in Norway. The material comprised data from random samples of three birth-cohorts living in the counties of Sør-and Nord-Trøndelag in 1983. The birth-cohorts were 1929-1938, 1939-1948 and 1959-1960. In 2006 two samples were drawn from the 1929-1938 and 1959-1960 birthcohort. The data collection comprised standardized clinical measurements and selfadministered questionnaires. The early oral health environment and social status and sex were related to oral health in 1983 by multiple regressions. The impact of social status was studied in combined datafiles from 1983 and 2006. The oral health environment in childhood was important for adults' oral health. The attention from parents and the local environment lead to a better oral health outcome in adulthood. Social status affected choices leading to better oral health. Regular dental visits were important especially for the eldest birth-cohort. Good oral health behaviors early and during adulthood were also important for oral health. Judged by number of tooth surfaces the difference between social status groups had not increased by 2006. A life-course perspective provides an opportunity to understand oral health over time. The present study supports the assumption that oral health is continuously exposed to environmental and behavioral risks that lead to accumulated diseases in the dental tissues.