ORIGINAL RESEARCH |
https://doi.org/10.5005/jp-journals-10077-3304 |
Assessment of the Reliability of Mixed Dentition Analysis on Adolescents of Tamil Nadu: A Cross-sectional Study
1,4,5Department of Pediatric and Preventive Dentistry, CSI College of Dental Sciences and Research, Madurai, Tamil Nadu, India
2Department of Oral Medicine and Radiology, Vinayaka Mission’s Sankarachariyar Dental College, Salem, Tamil Nadu, India
3Department of Orthodontics and Dentofacial Orthopedics, CSI College of Dental Sciences and Research, Madurai, Tamil Nadu, India
Corresponding Author: Saravanan Ramasamy, Department of Pediatric and Preventive Dentistry, CSI College of Dental Sciences and Research, Madurai, Tamil Nadu, India, Phone: +91 9842279466, e-mail: saravanan_bds@yahoo.com
Received: 23 January 2024; Accepted: 16 April 2024; Published on: 06 September 2024
ABSTRACT
Aim: This research aims to assess the reliability of various mixed dentition analyzes among the adolescents of Tamil Nadu and to devise a new regression equation by comparing the current mixed dentition analyzes.
Settings and design: About 900 adolescents (13–18 years of age) were selected from the state of Tamil Nadu using stratified multistage sampling.
Materials and methods: On the stone casts, the maximum mesiodistal width of permanent lower incisors, canines, premolars, and first permanent molars were measured. The measured values were analyzed and compared with existing mixed dentition analysis methods.
Statistical analysis: Descriptive statistics were performed, and Pearson’s correlation coefficient analysis was done. The p-value is kept at p < 0.05.
Results: The teeth widths show sexual dimorphism and were larger in males than females. Moyer’s (21.3 ± 0.6), Legovic’s (21.1 ± 0.6), and Bachmann’s (20.6 ± 0.6) methods significantly predicted the actual width (21.9 ± 1.3) in maxilla, whereas Moyer’s (20.3 ± 0.5), Gross and Hussund’s (21.0 ± 0.6), and Tanaka and Johnston’s (20.3 ± 0.7) methods significantly predicted the actual width (20.3 ± 1.0) in mandible. Though these methods significantly predicted, they either overestimated or underestimated in the majority of cases. Hence, a regression equation was developed based on the results of the study.
Conclusion: The basis of various mixed dentition analysis methods is the mesiodistal width of the teeth which is subjected to variability due to racial and ethnic factors. Therefore, the reliability of these analyzes should be analyzed in different racial groups. The regression equation developed in the study though holds good for a small set of samples, it should be tested on a large-scale for its reliability and accuracy.
Keywords: Arch analysis, Mixed dentition analysis, Moyer’s mixed dentition analysis, Reliability
How to cite this article: Ramasamy S, Narayanan M, Jaganathan S, et al. Assessment of the Reliability of Mixed Dentition Analysis on Adolescents of Tamil Nadu: A Cross-sectional Study. J South Asian Assoc Pediatr Dent 2024;7(2):77–82.
Source of support: Nil
Conflict of interest: None
INTRODUCTION
The dimension of the dental arch is a major concern for every pediatric dentist as it plays a vital role in ensuring a smooth transition from primary, through mixed to permanent dentition.1 Malocclusion is known to develop during the mixed dentition stage and timely management either eliminates or decreases the severity.2 The mixed dentition space analysis (MDSA) plays a vital role in diagnosing arch length tooth size discrepancies and determining various orthodontic treatments ranging from serial extraction to a mere periodic checkup.3
Calculating the space required for the eruption of permanent canines and premolars into the oral cavity involves the combined mesiodistal width of the permanent incisors. Carey4 was the first person to prove this higher degree of correlation while it was Ballard and Wylie,5 who first introduced the concept of regression equation that predicts the combined widths of canines and premolars. Though various methods like radiographic measurements and regression equations are available, Moyer’s method6 and Tanaka and Johnston’s method7 were majorly used. Tanaka and Johnston reassessed Moyer’s analysis and derived a simpler equation by calculating the sum of the mesiodistal width of four mandibular incisors.
A few studies8,9 conducted in India using Tanaka, Johnston, and Moyer’s methods found significant differences between actual and predicted values. A study conducted among different racial groups like Norwegian Lapps, Australian aborigines, Japanese, and Americans, concluded that there was a significant difference in the teeth sizes among different races.10 Hence, these prediction techniques carried out on a specific sample are not reliable on different racial groups.
Legovic et al.11 measured mesiodistal widths and vestibulo-oral diameters of erupted central incisors, lateral incisors, and first permanent molars. They used this data to formulate a regression equation that predicts the widths of unerupted premolars and canines in the Zagreb population. Subsequent studies by Bachmann,12 Gross and Hasund,13 and Trankman et al.,14 also followed this approach. This clearly states that a particular prediction model cannot be applied universally and there is a lack of literature assessing the same in the Indian population, particularly in Tamil Nadu. Hence, the current study was carried out to assess the reliability of Moyer’s mixed dentition analysis and compare the different regression models to establish a novel regression equation that predicts the mesiodistal width of unerupted premolars and canines among the adolescents of Tamil Nadu.
MATERIALS AND METHODS
This was a cross-sectional study carried out among 900 adolescents (13–18 years of age) from the State of Tamil Nadu during 2016–2017. A stratified multistage sampling technique was employed for the recruitment of study participants following World Health Organization (WHO) oral health assessment methods. The clusters were the districts of Tamil Nadu accounting for 32 clusters. Each district/cluster was stratified into urban, suburban, and rural areas as applicable. From each stratum, two districts were randomly chosen accounting for a total of six districts. The districts of Coimbatore and Chennai were randomly chosen using the lottery method for the urban strata. Similarly, the districts of Madurai and Salem are for the suburban strata, and the districts of Nagapattinam and Tiruvannamalai are for the rural strata. From each of the six districts, five schools (sampling sites) were chosen randomly from the list of schools obtained from the School Education Department, Tamil Nadu Government. The schools of each stratum were chosen from within the stratum namely, urban, suburban, or rural areas accordingly. According to the WHO oral health assessment methods, a minimum of 25 subjects from each sampling site is recommended. Considering an attrition of 17%, 30 subjects from each site (school) were considered in the current study.
-
Urban strata: 2 districts × 5 schools × 30 subjects = 300.
-
Suburban strata: 2 districts × 5 schools × 30 subjects = 300.
-
Rural strata: 2 districts × 5 schools × 30 subjects = 300.
Thus, the final sample size obtained was 900 adolescents including an attrition of 17% as depicted in Figures 1 and 2.
Fig. 1: Recruitment of study participants
Fig. 2: Stratified multistage sampling technique
Ethical clearance was obtained from the research committees at CSI College of Dental Sciences and Research, and Vinayaka Mission’s Sankarachariyar Dental College and the approval numbers were CSICDSR/IEC/0023P2016 and VMSDC/IEC/Approval No.065, respectively. Relevant authorities of the schools involved were approached to obtain permission to conduct the study and written informed consent was obtained from the head of the school. The children and their parents were informed appropriately in writing and written informed consent was obtained.
A free dental camp was conducted and assisted by Compulsory Rotatory Residential Internship from CSI College of Dental Sciences and Research at all the concerned schools. During the oral health screening, adolescents who met the inclusion criteria were enrolled in the study. The inclusion criteria were all permanent teeth (except the second and third molars) to be completely erupted in both maxilla and mandible, with no active caries, proximal restorations, tooth anomalies, no history of previous or present orthodontic treatment, and with intact mesiodistal and buccolingual surfaces, and well-aligned teeth enabling measurement of mesiodistal width. The candidates must be free of any systemic disease, congenital craniofacial anomalies, and serious health issues.
The recruited adolescents were asked to rinse their mouths twice with tap water, and the alginate impressions (DPI Algitex) of both the maxilla and mandible were made. After disinfecting the impressions, the dental stone mixture was poured immediately. Casts were trimmed and bases were poured with dental plaster. Around 30 dental casts (15 boys and 15 girls) from each school without any impression errors were considered for analysis. A Vernier gauge (Yamayo Digimatic Caliper Absolute, serial no: 1512398) was used for the measurement of the mesiodistal width of maxillary and mandibular incisors, canines, premolars and molars, and the buccolingual width of the maxillary and mandibular incisors and molars. The measurements were repeated twice by a single examiner who was blinded, and the mean was considered for analysis. The sum of the mesiodistal width of canines and premolars (hereafter mentioned as “actual width”) of both arches was measured. The mesiodistal widths of canines and premolars were predicted using Moyer’s method, Tanaka and Johnston’s method, Bachmann’s Method, Gross and Hussund’s method, Trankman’s method, and Legovic’s method (hereafter mentioned as “predicted value”). The actual width was compared with the predicted values obtained using the abovementioned MDSA methods.
The statistical analysis was conducted using IBM Statistical Package for the Social Sciences (SPSS) Statistics, version 16.0 (SPSS Inc., Chicago, Illinois, United States of America). Descriptive statistics were performed followed by Pearson’s correlation coefficient analysis. The p-value is kept at p < 0.05.
RESULTS
Among the 900 study participants, gender was distributed evenly with 450 males and 450 females. The participants were evenly distributed in each of the three strata namely, urban, suburban, and rural with 300 participants from each stratum. There were no significant differences in the measured widths of various teeth between participants from urban, suburban, and rural areas.
The sum of mesiodistal widths of mandibular incisors (LI), the sum of mesiodistal widths of mandibular canine, first and second premolars (UCPM), and the sum of mesiodistal widths of maxillary canine, first and second premolars (LCPM) among the males were 22.38 ± 1.26, 20.66 ± 0.89, and 20.94 ± 0.78, respectively. The sum of mesiodistal widths of LI, the sum of mesiodistal widths of UCPM, and the sum of mesiodistal widths of LCPM among the females were 21.58 ± 1.33, 19.96 ± 1.07, and 20.71 ± 0.94, respectively. The sum of mesiodistal widths of LI, and the sum of mesiodistal widths of UCPM were significantly higher in boys than in girls. No statistically significant difference was noted in the sum of mesiodistal widths of LCPM between boys and girls.
In the maxillary arch, the actual width for boys (20.9 ± 0.8) was close to the predicted values using Moyer’s (21.0 ± 0.6) and Gross and Hussund’s analysis (21.0 ± 0.4). The actual width of girls (20.71 ± 0.9) was close to the predicted values using Legovic’s (20.8 ± 0.3), and Moyer’s analysis (20.9 ± 0.3). Considering the actual width of the maxilla for both genders combined (21.9 ± 1.3), it significantly correlated with the predicted value using Moyer’s (20.9 ± 0.3), Legovic’s (20.8 ± 0.3), and Bachmann’s analysis (21.0 ± 0.04) with a Pearson’s correlation coefficient of 0.225, 0.098, and −0.079, respectively as shown in Table 1. The difference in the maxillary actual width and the predicted values in the increasing order was Legovic’s < Moyer’s < Bachmann’s < Trankman’s, Gross and Hussund’s < Tanaka and Johnston’s of which only Legovic’s, Moyer’s, and Bachmann’s significantly predicted the actual width as shown in Figure 3. Moyer’s, Bachmann’s, Gross and Hussund’s, Trankman’s, Legovic’s, and Tanaka and Johnston’s methods correctly predicted the maxillary actual width in 13, 10, 10.2, 10, 10, and 12% of the study population, respectively. For the rest of the population, they either underestimated or overestimated the sum of the mesiodistal width of canines and premolars.
| Gender | Actual width | Moyer’s method | Bachmann’s method | Gross and Hussund’s method | Trankman method | Legovic method | Tanaka and Johnson |
|---|---|---|---|---|---|---|---|
| Boys (n = 450) mean ± standard deviation (SD) | 20.9 ± 0.8 | 21.0 ± 0.6 | 20.2 ± 0.6 | 21.0 ± 0.4 | 21.3 ± 0.3 | 21.5 ± 0.5 | 22.1 ± 0.6 |
| Girls (n = 450) mean ± SD | 20.7 ± 0.9 | 20.9 ± 0.3 | 21.0 ± 0.4 | 21.2 ± 0.4 | 21.2 ± 0.2 | 20.8 ± 0.3 | 21.5 ± 1.9 |
| Total (n = 900) mean ± SD | 21.9 ± 1.3 | 21.3 ± 0.6 | 20.6 ± 0.6 | 21.1 ± 0.5 | 21.3 ± 0.3 | 21.1 ± 0.6 | 21.3 ± 1.5 |
| Pearson’s correlation coefficient r | 1 | 0.225 | −0.079 | −0.035 | 0.029 | 0.098 | 0.034 |
Fig. 3: The actual width and the mean values for various mixed dentition analyzes in maxilla by gender
In the mandibular arch, the actual width for boys (20.6 ± 0.9) was close to the predicted values using Gross and Hussund’s (20.6 ± 0.5), Legovic’s (20.5 ± 0.5), and Moyer’s (20.7 ± 0.5). The actual width for girls (20.0 ± 1.0) was close to the predicted values using Moyer’s (20.0 ± 0.4), Legovic’s (20.2 ± 0.4), and Tanaka and Johnston’s (20.3 ± 0.6). Considering the actual width of the mandible for both genders combined (20.3 ± 1.0), it significantly correlated with the predicted value using Moyer’s (20.3 ± 0.5), Gross and Hussund’s (21.0 ± 0.6), and Tanaka and Johnston’s (20.3 ± 0.7) with a Pearson’s correlation coefficient of 0.187, −0.234, and 0.142 as shown in Table 2. The difference in the mandibular actual width and the predicted value in the increasing order was Moyer’s > Tanaka and Johnston’s > Legovic’s > Bachmann’s, Trankman’s > Gross and Hussund’s of which only Moyer’s, Gross and Hussund’s, and Tanaka and Johnston’s significantly predicted the actual width as shown in Figure 4. Moyer’s, Bachmann’s, Gross and Hussund’s, Trankman’s, Legovic’s, and Tanaka and Johnston’s methods correctly predicted the mandibular actual width in 13, 10, 9, 10, 13, and 10% of the study population, respectively. For the rest of the population, they either underestimated or overestimated the sum of the mesiodistal width of canines and premolars.
| Gender | Actual width | Moyer’s method | Bachmann’s method | Gross and Hussund’s method | Trankman method | Legovic method | Tanaka and Johnson |
|---|---|---|---|---|---|---|---|
| Boys (n = 450) mean ± SD | 20.6 ± 0.9 | 20.7 ± 0.5 | 21.0 ± 0.7 | 20.6 ± 0.5 | 20.0 ± 0.4 | 20.5 ± 0.5 | 20.9 ± 0.6 |
| Girls (n = 450) mean ± SD | 20.0 ± 1.0 | 20.0 ± 0.4 | 20.4 ± 0.5 | 21.4 ± 0.3 | 19.7 ± 0.4 | 20.2 ± 0.4 | 20.3 ± 0.6 |
| Total (n = 900) mean ± SD | 20.3 ± 1.0 | 20.3 ± 0.5 | 20.7 ± 0.7 | 21.0 ± 0.6 | 19.9 ± 0.4 | 20.4 ± 0.4 | 20.3 ± 0.7 |
| Pearson’s correlation coefficient r | 1 | 0.187 | 0.126 | −0.234 | 0.102 | 0.115 | 0.142 |
| p-value | – | 0.000 | 0.015 | 0.284 | 0.378 | 0.002 | 0.298 |
Fig. 4: The actual width and the mean values for various mixed dentition analyzes in mandible by gender
Most analyzes use the base formula for predicting the sum of mesiodistal width of unerupted canine and premolars as the “sum of mesiodistal width of four mandibular incisors/2.” Considering this and the mean difference in the actual width and the predicted values using various MDSA, the following regression equation for Tamil Nadu was arrived at to predict the sum of mesiodistal widths of unerupted canine and premolars.
For maxilla, X/2+9
For mandible, X/2+7.5
The aforementioned equation exhibits a high degree of accuracy when applied to a fresh set of 20 samples.
DISCUSSION
Most MDSA methods are based on the tooth sizes of Caucasians. This would question their reliability when used in the non-Caucasian population; hence, many studies were conducted to find the same.
Lavelle in his study, proved the existence of tooth size differences between different races.15 Singh and Nanda, found significant differences in the tooth sizes between Indian and Caucasian children.16 Staley et al.,17 found a significant difference in the tooth size between boys and girls; thus, proving the gender differences. Since gender dimorphism in the tooth width was also noted in the present study, the analysis was considered for males and females separately.
The three methods used for measuring the width of the tooth are brass wire, Boley’s gauge, and Vernier caliper. Out of the three methods, a digital Vernier caliper is found to be the most accurate as suggested by Legovic et al., and hence, used in the present study.11
Various Indian studies conducted by Grover et al.,18 Ravinthar and Gurunathan,19 and Baheti et al.,20 evaluated the reliability of either one or two MDSA methods in the Indian population. The current study is the only comprehensive study comparing and evaluating six different MDSA methods to the Tamil Nadu population and derived an equation true to the population.
The 75th percentile of probability in Moyer’s method was closer to the actual width in the present study and the correlation was high (r = 0.225 for maxilla; r = 0.187 for mandible). However, the same method overpredicted 75.8 and 75.7% of the cases in maxilla and mandible, respectively. These findings were similar to the ones conducted by Doda et al.,21 in the North Indian population, Bhatnagar et al.,22 Kakkar et al.,23 and Kommineni et al.,24 in the Chennai population. This signifies that Moyer’s method is not appropriate for the Indian population.
Tanaka and Johnston’s method overestimated the actual width in 64.5 and 62.5% population in the maxilla and mandible, respectively. This finding was consistent with the works done by Doda et al.,21 Ravinthar and Gurunathan,19 Bhatnagar et al.,22 and Kakkar et al.,23 stating that Tanaka and Johnston’s method is not appropriate for the Indian population.
In line with this evidence, the regression equation developed from this study holds good for a small set of 20 samples tested. These findings underscore the importance of exploring the equation’s applicability in a broader context to glean further insights into its predictive capabilities.
The alginate impressions made in this study are more prone to shrinkage posing a threat to the accuracy of the measured values. Hence, digital models are recommended for greater accuracy. The sample size in the present study could have been larger for better accuracy and prediction. Determination of the mesiodistal widths of caries-affected children and children with malocclusion was not included in the present study. The study is also limited to a small geographic area and further research of the same kind is recommended including samples to represent the whole Indian population.
CONCLUSION
Tooth widths exhibited sexual dimorphism in the Tamil Nadu population and the tooth widths were noted to be larger among males than females. Legovic’s, Moyer’s, and Bachmann’s analyzes significantly predicted the actual width of canine and premolars in the maxilla whereas it was Moyer’s, Gross and Hussund’s, and Tanaka and Johnston’s for the mandibular arch. Though these analyzes significantly predicted the actual width, they were either overestimating or underestimating the widths in the majority of the population. The base of various MDSA methods is the mesiodistal width of the teeth which is subjected to variability due to racial and ethnic factors. Therefore, this mandates a reliability test for different racial groups and the development of a regression equation for that group. Though the regression equation developed in this study holds good for a small set of samples, it should be tested on a large scale for its reliability and accuracy.
ORCID
Saravanan Ramasamy https://orcid.org/0009-0005-5966-2011
REFERENCES
1. Rajbhoj AA, Parchake P, Begnoni G, et al. Dental changes in humans with untreated normal occlusion throughout lifetime: a systematic scoping review. Am J Orthod Dentofacial Orthop 2021;160(3):340.e3–362.e3. DOI: 10.1016/j.ajodo.2021.02.014
2. Chandna A, Gupta A, Pradhan KL, et al. Prediction of the size of unerupted canines and premolars in a North Indian population - an in vitro study. J Indian Dent Assoc 2011;5:329–333.
3. Staley RN, Kerber RE. A revision of the Hixon and Oldfather mixed-dentition prediction method. Am J Orthod 1980;78(3):296–302. DOI: 10.1016/0002-9416(80)90274-2
4. Carey CW. Linear arch dimension and tooth size: an evaluation of the bone and dental structures in cases involving the possible reduction of dental units in treatment. Am J Orthod 1949;35(10):762–775. DOI: 10.1016/0002-9416(49)90148-7
5. Ballard ML, Wylie WL. Mixed dentition analysis-estimating size of unerupted permanent teeth. Am J Orthod 1947;33(11):754–759. DOI: 10.1016/s0096-6347(47)90073-2
6. Moyers RE. Analysis of the dentition and occlusion. Handbook of orthodontics. 1988.
7. Tanaka MM, Johnston LE. The prediction of the size of unerupted canines and premolars in a contemporary orthodontic population. J Am Dent Assoc 1974:88(4):798–801.
8. Dasgupta B, Zahir S. Comparison of two non-radiographic techniques of mixed dentition space analysis and evaluation of their reliability for Bengali population. Contemp Clin Dent 2012;3(Suppl 2):S146–S150. DOI: 10.4103/0976-237X.101069
9. Sholapurmath SM, Benni DB, Mandroli P. Applicability of two mixed dentition analysis in children of Jangam community of Belgaumcity. World J Dent 2012;3(4):324–329. DOI: 10.5005/jp-journals-10015-1183
10. Bailit HL. Dental variation among populations. Dent Clin North Am 1975;19(1):12539.
11. Legovic M, Novosel A, Legovic A. Regression equations for determining mesio-distal crown diameters of canines and premolars: Angle Orthod 2003;73(3):314–318. DOI: 10.1043/0003-3219(2003)073<0314:REFDMC>2.0.CO;2
12. Bachmann S. Voraussage des Platzbedarfs in den Stützzonenmittels multipler Regressions gleichungen. Fortschritte der Kieferorthopädie 1986;47(1):79–86. DOI: 10.1007/BF02167599
13. Gross A, Hasund A. Neuere vergleichende korrelations statistische Unterschungen zur Vorhersage des Platzbedarfs in den Stutzzonendurch multiple Regressions gleichungen. Fortschr Kieferorthop 1989;50(2):109–117. DOI: 10.1007/BF02203067
14. Trankmann J, Mohrmann G, Themm P. Vergleichende Untersuchungen der Stutzzonenprognose. Fortschr Kiefer Orthop 1990;51(3):189–194. DOI: 10.1007/BF02164570
15. Lavelle CL. Secular trends in different racial groups. Angle Orthod 1972;42(1):19–25. DOI: 10.1043/0003-3219(1972)042<0019:STIDRG>2.0.CO;2
16. Singh M, Nanda RS. Prediction of tooth size and its clinical application. J Indian Dent Assoc 1972;44(5):95–98.
17. Staley RN, Hoag JF. Prediction of the mesiodistal widths of maxillary permanent canines and premolars. Am J Orthod 1978;73(2):169–177. DOI: 10.1016/0002-9416(78)90187-2
18. Grover N, Saha S, Tripathi AM, et al. Applicability of different mixed dentition analysis in Lucknow population. J Indian Soc Pedod Prev Dent 2017;35(1):68–74. DOI: 10.4103/0970-4388.199220
19. Ravinthar K, Gurunathan D. Applicability of different mixed dentition analyses among children aged 11-13 years in Chennai population. Int J Clin Pediatr Dent 2020;13(2):163–166. DOI: 10.5005/jp-journals-10005-1736
20. Baheti K, Babaji P, Ali MJ, et al. Evaluation of Moyer’s mixed dentition space analysis in Indian children. J Int Soc Prev Community Dent 2016;6(5):453–458. DOI: 10.4103/2231-0762.184037
21. Doda A, Saraf BG, Indushekhar KR, et al. Evaluation and applicability of Tanaka–Johnston and Moyers’ mixed dentition analysis for North Indian population. World J Dent 2021;12(1):57–63. DOI: 10.5005/jp-journals-10015-1819
22. Bhatnagar A, Jindal MK, Khan SY. Comparison of two different non-radiographic mixed dentition analysis. Pesquisa Brasileira em Odontopediatria e Clínica Integrada. 2019;19:e4374.9. DOI: 10.4034/PBOCI.2019.191.59
23. Kakkar A, Verma KG, Jusuja P, et al. Applicability of Tanaka–Johnston, Moyers and Bernabé and Flores–Mir mixed dentition analyses in school-going children of Sri Ganganagar City, Rajasthan (India): a cross-sectional study. Contemp Clin Dent 2019;10(3):410–416. DOI: 10.4103/ccd.ccd_654_18
24. Kommineni NK, Reddy CV, Chandra NS, et al. Mixed dentition analysis - applicability of two non-radiographic methods for Chennai school children. JISPCD 2014;4(2):133–138. DOI: 10.4103/2231-0762.139847
________________________
© The Author(s). 2024 Open Access. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted use, distribution, and non-commercial reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.