Among the NFBC 1966 participants, the overall prevalence of erosive tooth wear was high. Almost half of the now middle-aged cohort members were found to be in need of at least preventative measures against further progression of the condition. Of the individuals with erosive lesions, 14.6 % suffered from severe erosive tooth wear. The risk for severe erosive wear was two-fold for males and those with restorative treatment need.
The most recent studies using the BEWE index support the present prevalence figures [4, 6, 20]. In a large scale multicenter study (including Finland), the overall prevalence of at least one erosive lesion in a sample of European young adults was 57.1 % and of the Finnish participants 17.7 % had a maximum single BEWE score of 2 or more . The subjects were younger than here, which may explain the slightly higher prevalence observed in the present study. Furthermore, the predominance of erosive tooth wear in the upper anterior sextant has also been recently reported . Despite the differences in the scoring systems in other recent studies [27, 28] the prevalence values reported are similar to the present study.
According to this study, those with high restorative treatment need were at almost three-fold risk for severe erosive wear. In addition, past caries experience was positively associated with present erosive wear. Both dental caries and erosive tooth wear have common etiological factors, such as high or constant intake of sweetened soft drinks, low saliva secretion and unhealthy dietary habits, which may explain the association. Furthermore, individuals with erosion have suggested having salivary characteristics similar to those of caries-active individuals . A clinical implication might be, that current restorative treatment need status may act as a risk indicator for erosive wear, and vice versa. However, findings in the literature on the issue are equivocal, with many studies reporting statistically significant association between erosive tooth wear and dental caries [27, 30, 31] while others do not [12, 32].
With regard to gender differences, the prevalence of erosive tooth wear was more frequent in males than in females. This is in agreement with many previous reports [7, 13, 15, 28, 32]. Yet, there are still many studies that have reported no gender differences in the prevalence of erosive tooth wear [6, 20]. As a biological explanation for males being more prone to erosive tooth wear, it has been suggested that there are differences between genders in the consumption of carbonated drinks and in the strength of biting forces . Indeed, tooth grinding may be significantly associated with the incidence of erosive tooth wear and may play a major role in erosive lesions – bigger than has been suspected . This needs further investigations.
In the present study, socio-demographic variables were associated with erosive wear among females, and lower education was associated with elevated risk for severe erosive wear in unadjusted regression model. However, education was not found to be statistically significant variable for severe erosive tooth in the adjusted logistic regression model. With regard to other studies, many have reported no significant association between the condition and sociodemographic factors [6, 20, 30]. One study reported erosive tooth wear being more common among individuals rated with a higher socioeconomic status . Both healthy as well as unhealthy diets may contain acidic foods, thus adults from different socioeconomic backgrounds could possess a similar risk for erosive tooth wear. The overall lack of statistical significance with respect to socio-demographic factors may also be explained by the cross-sectional nature of many studies, including the present one. Given that erosive tooth wear usually develops over time, lifestyle factors present for several years may have been responsible for the current erosive tooth wear. Long-term effects of dietary habits in association with erosive tooth wear could be a topic for a future study.
The strength of this study was a big study population, not limited to any subgroup. Thanks to the large enough sample size, our statistical power is good. Also participation rate was relatively high (62 %). Concerning caries experience, our results are in line with Health 2000-survey , thus this sample can be considered to represent well Finnish adults. However, the presence of response bias must always be kept in mind in cohort studies. Another limitation is the fact that differential diagnosis of erosive wear is difficult, as it usually co-exists simultaneously with other types of tooth wear, especially in the ageing dentitions, such as in the present study. Diagnosis of early signs and symptoms of erosive tooth wear is challenging even in dentitions without other forms of wear. Even though we made attempts to distinguish between erosion, attrition and abrasion, the examiner’s individual opinion on the origin of wear may have produced bias in the study. In addition, with respect to caries diagnostics, the ICDAS protocol states that the teeth should be cleaned with at least a toothbrush and floss before conducting an examination . In our case, no professional cleaning was undertaken; however it is a common habit to brush teeth before seeing a dentist in Finland. It is also well known that radiography may give a substantial contribution to the diagnosis of interdental and occlusal caries. PTG radiographs were taken of all individuals, but they are not reliable in caries diagnostics.
The BEWE index itself and the use of the BEWE cumulative score have been validated and shown to be acceptable for recording tooth wear and scoring severity and in prevalence studies [34, 35].
However, in the study of Olley et al. (2014) , the BEWE index was used as a tool for diagnosing tooth wear overall not distinguishing between erosion and mechanical wear. In our study, the examiners had difficulties in differentiating between intact enamel and initial loss of surface texture. Same tendency can be seen in other studies, especially when having multiple examiners [34, 36]. A recent study of the validity and reliability of the BEWE index showed only a moderate inter- and intra-examiner agreement  suggesting that the BEWE scores should be interpreted with some caution. However, the authors concluded that BEWE is an appropriate and reliable tool. Despite the given sufficient accuracy of the BEWE scoring system, the problems in diagnosing mild erosive tooth wear in aged dentitions must be borne in mind. Using the BEWE sum score and the cut-off point ≥9 for analyses is likely to increase the reliability of the study. In addition, the cut-off values have been suggested to be in need of re-evaluation . A study evaluating the reliability and accuracy of the BEWE scoring system in adult population for erosive tooth wear, would be most valuable.
With respect to the BEWE sum scores, the cut-off values were based on the experience and previous studies of one of the authors (AL), and this topic is under on-going consideration and evaluation . Vered et al. (2014)  suggested that the cut-off sum score for high risk could be 7, for medium risk 4–6 and for low risk 1–3. Originally, the cut-off value for high risk was 14, for medium risk 8–13, for low risk 3–6 and for no risk 0–2 . Another option is to use only the highest BEWE score per subject, but it may not be as comprehensive as the BEWE sum score. Whether using the BEWE sum score (0–18) or the highest BEWE score per subject (0–3), the outcomes in the present study remained substantially the same. Therefore, according to the present results, either system can be used.