Short Root Anomaly in a Hispanic Population: Risk for Orthodontic Root Resorption

While the presentation of Short Root Anomaly (SRA) in Hispanic patients has been described previously, it is not known if this population is predisposed to increased orthodontic root resorption. This study evaluates the response of pre-existing short roots in Hispanic SRA patients to orthodontic treatment. Selected maxillary and mandibular teeth of 40 Hispanic SRA patients (19 male, 21 female) and 40 age and gender matched Caucasian patients (19 male, 21 female) with normal root length were evaluated for root resorption following comprehensive orthodontic treatment. The age range of the subjects was between 10 and 19 years. Relative root length was calculated before and after orthodontic treatment from digital panoramic radiographs. Overall, statistically significant root resorption occurred in the control group, but orthodontic root resorption was not significant in the Hispanic group (p > 0.05). When genders were separated, Hispanic females did experience a mild degree of root resorption in the upper incisors while resorption in Hispanic males was not significant. Caucasian females experienced greater root resorption than Caucasian males. Hispanic SRA patients may be safely treated with comprehensive orthodontics and could be at no more risk of root resorption than Caucasian patients with normal initial root length.


Introduction
There are two main causes of short dental roots: 1) disturbances during root development or 2) resorption of well-developed roots. Developmentally short-rooted permanent teeth can be genetic, known as short root anomaly (SRA) [1,2], or exogenous, including irradiation of the head and neck and/or chemotherapy in childhood during tooth development [3]. Resorption of dental roots as a result of orthodontic treatment is termed Orthodontically Induced Inflammatory Root Resorption (OIIRR) and is believed to result from a combination of genetics, environmental, and mechanical factors. Puranik et al. [4] identified a population of Hispanic orthodontic patients presenting with SRA. Although the Hispanic population only comprises 17.8% of the U.S. population, it is the largest minority ethnicity with 57.5 million people and continues to grow [5].
In cases of pre-existing short root length (Figure 1), orthodontists should be aware if greater root resorption is anticipated as normal orthodontic treatment could sentence these teeth to exfoliation or extraction. Short dental roots may complicate dental treatment planning when considering anchorage or estimating the ability of a tooth to carry masticatory forces. Two recent publications proposed such potential hazards, but did not supply clinical evidence [6,7]. This is the first study, to the authors' knowledge, that has investigated the response of dental roots to orthodontic treatment in a Hispanic SRA population.
Root resorption is a common side effect of orthodontic treatment and is estimated to occur in 73% of orthodontically treated cases [8]. Luckily, in the majority of cases, resorption is classified as mild to moderate [9] and is clinically insignificant. While some studies have found females to experience more resorption [10], others report that gender is unlikely to have an effect on root resorption [11].
Short-root anomaly (SRA) is a developmental phenomenon in which permanent teeth never reach their normal length and appear very short and blunted [2]. Lind [1] first described SRA in 1972 as "abnormally short roots of a characteristically plump shape affecting maxillary central incisors primarily." It is considered to be familial [1] and prior to Lind, SRA was commonly misdiagnosed as root resorption. Interestingly, multiple studies have suggested a primarily autosomal dominant inheritance pattern. In a recent study by Puranik et al. [4], SRA was found to occur more frequently in Latino individuals with the majority having localized SRA and the minority having generalized SRA, in which every tooth in the mouth is affected [4].
Some studies have found that patients with short or blunted roots prior to treatment, undergo significantly more root shortening during orthodontic treatment than patients with no dental anomaly [12][13][14], while others have found a minor or non-significant relationship between blunted roots and root resorption with orthodontic treatment [2,15,16] and still others report that blunted teeth experience less resorption than normal-shaped teeth [11].
Some authors believe that this is due to a change in the position of the tooth's center of resistance, moving it more incisally. Studies that looked specifically at "pipette-shaped" roots, rather than blunted, consistently found greater root resorption [11]. Correlations have also been found between initial tooth length and the amount of resorption (i.e., a longer root was more likely to be resorbed with orthodontic treatment) [11,17]. Mirabella and Artun [17] argued that teeth with longer roots require stronger orthodontic forces to move them; furthermore, the  displacement of the root apex is greater during tipping and torquing movements of longer rooted teeth.
The objective of this study was to assess changes in root lengths that occur with orthodontic treatment by comparing: 1. pre-and post-treatment digital panoramic radiographs of Hispanic patients presenting with short dental root lengths; 2. the experience of Hispanic patients and Caucasian patients presenting with normal dental root lengths; and 3. gender differences within each ethnicity group.

Material and methods
This was a retrospective study that included male and female Hispanic and Caucasian patients in a university orthodontic clinic. Dental records of 40 Hispanic SRA and 40 Caucasian patients were collected. In this clinic, most of the Hispanic patients are of Mexican heritage. The Hispanic group consisted of 19 male and 21 female adolescent patients and the Caucasian group consisted of 19 male and 21 female adolescent patients. The mean age of subjects was 14.7 + 1.91 years in the Caucasian group and 13.9 + 1.86 years in the Hispanic group. All pre-existing records were taken at the orthodontic clinic for comprehensive orthodontic treatment. All radiographs were taken in the Radiology Clinic. Panoramic radiographs were taken on a standard combined panoramic/cephalometric machine. All radiographs were consistent with the standard of care for all orthodontic patients in the clinic and included pre-and post-treatment panoramic radiographs. For every extraction case in the Hispanic group, an extraction case in the Caucasian group was included. All subjects were de-identified and assigned a coded number. Only the sex, ethnicity, and age in years and months at the start of treatment were retrieved from the patient chart. All subjects with developmental disorders or complicated medical histories were excluded. Individual teeth were excluded if 1) the apex was not closed, 2) the reference points were not clearly visible, 3) there was a history of dental trauma, root canal therapy, orthodontics, or an incisal or full-coverage restoration, 4) attrition or abrasion of the crown was present. In this study generalized and localized SRA patients were grouped together even though their pathogenesis may differ. IRB exemption was granted prior to data collection.
Root lengths were measured on maxillary and mandibular central and lateral incisors, and second premolars, from the apex to the midpoint of the cementoenamel junction (CEJ) on panoramic radiographs. The aforementioned teeth were selected because they are single-rooted teeth (with the exception of the variation in upper second premolar) and the root outline of these teeth can be more clearly visible on radiographs than multi-rooted teeth. Furthermore, SRA in a Mexican cohort most commonly occurred in maxillary central incisors and mandibular second premolars [4]. It is assumed that during orthodontic treatment, the crown length does not change. Therefore, the ratio between the root and crown length should reflect any changes in root length. Crown and root lengths were measured on panoramic radiographs using the Dolphin™ digital caliper (Dolphin Imaging 11.8, Chatsworth, CA).
The Lind method of relative root length calculation was used [1] in which the midpoint of the mesial and distal CEJ (median CEJ) is used to demarcate the transition from crown to root. Root length was measured from the median CEJ (point M) to the tip of the root (point R). Crown length was measured from point M to the incisal edge or tip (point I). Relative root length was calculated by dividing point M to point R length by point I to point M length [1]. Root length guidelines set by Holtta et al. [3] were used in the measurements of all panoramic radiographs.
Statistical analysis was done using IBM SPSS Statistics for Windows (version 22.0, IBM Corp., Armonk NY). Analyses were performed by the Explore function in SPSS to check the raw data distribution by Shapiro-Wilk Tests of Normality. Parametric tests were done based on the results that the majority of the variables showed normal distribution of the data. Student Paired Sample t-tests were used to assess the mean differences between pre and post measurements of all the variables in each of the Hispanic and Caucasian groups with the sample gender in consideration. Independent t-tests were done for the mean comparisons to all study variables between groups. All values were considered statistically significant for a value of p<0.05.

Results
The Caucasian group had a statistically significant differences between preorthodontic treatment and post-orthodontic treatment relative root lengths ( Table 1). The post-orthodontic treatment relative root lengths (1.87 ± 0.30) were significantly less than the initial relative root lengths (2.10 ± 0.29), indicating root resorption from the orthodontic treatment in the Caucasian group. The only measurements that did not show statistical significance were the lower left lateral and lower left central incisors.
Overall, there was only a mild difference between Hispanic initial relative root lengths (1.57 ± 0.27) and Hispanic post-orthodontic treatment relative root lengths (1.55 ± 0.29) ( Table 2). Significant root resorption was seen in five teeth: UR2, UR1, UL1, LR1 and LR2. Differences were also seen in the mandibular left second premolars which increased in relative root length after orthodontic treatment. Furthermore, the percentage decrease in relative root length was consistently higher in the Caucasian group (Tables 1 and 2). Genders were separated and post-treatment relative root lengths were compared against pre-treatment relative root lengths in order to further evaluate any significant    root resorption that occurred with orthodontic treatment. There was no difference in post-orthodontic treatment relative root length when sexes were compared within the Hispanic group. However, there was a difference between males and females in postorthodontic treatment relative root length with Caucasian females (1.80 ± 0.28) experiencing a smaller post-orthodontic treatment relative root lengths than Caucasian males (1.95 ± 0.31). Statistically significant differences in post-orthodontic treatment relative root length were found in four teeth when Caucasian male and Caucasian female groups were compared (Tables 3 and 4). Specifically, the relative root lengths of UR1, UL1, LL2, and LL1 were found to be significantly less in the Caucasian female group than the Caucasian male group. Because there was no statistically significant gender difference in pre-orthodontic treatment root lengths in the Caucasian group, this finding suggests that Caucasian females experienced more root resorption from the orthodontic treatment than Caucasian males. In the Caucasian male group, significance was found in five teeth: UR2, UR1, UL1, UL2, LL5 ( Table 3). In all five teeth, the final was smaller than the initial, indicating that these teeth did in fact experience statistically significant orthodontic root resorption.
Significance between pre and post-treatment relative root length was found for nearly every tooth in the Caucasian female group: UR5, UR2, UR1, UL1, UL2, UL5, LL5, LL2, LR2, LR5 ( Table 4). All teeth showed decreased relative root lengths post-treatment, suggesting significant root resorption from the orthodontic treatment. The only teeth that did not show a significant difference between pre and post were the two lower central incisors: the post-treatment mean relative lengths for these teeth were still smaller than the pre-treatment.
When post-treatment relative root lengths were compared to pre-treatment root lengths in Hispanic males, three teeth showed a significant difference: UR5, UL5, and LL5 (Table 5). However, the mean relative root lengths showed that these root lengths increased in apparent length after treatment.   orthodontic treatment in UR2, UR1, UL1, and LR2; the mean relative root length for UL5 is increased in the post group, suggesting an increase in relative root length with orthodontic treatment.

Discussion
There have been many claims that SRA predisposes patients to increased orthodontic root resorption, for example the study of Wang and Feng [6]. However, it is not known if this cohort indeed experiences a higher degree of orthodontic root resorption, or if they experience a similar degree of resorption to unaffected patients, but the appearance is more suggestive due to their pre-existing short roots. If a greater degree of relative root loss does occur, the orthodontic professional should be aware so that treatment complications can be anticipated when treating Hispanic SRA patients.
Post-orthodontic treatment relative root lengths were significantly smaller than the initial relative root lengths for the Caucasian group, indicating root resorption from the orthodontic treatment. Hispanic initial relative root lengths and Hispanic post-orthodontic treatment relative root lengths were similiar. Furthermore, the percentage decrease in relative root length was consistently higher in the Caucasian group, further supporting the finding that the Caucasian subjects experienced OIIRR while the Hispanic subjects had minimal, if any. Because the treating orthodontists in this study were aware of the pre-existing short roots in the Hispanic SRA group, it is likely that these patients were treated more conservatively, with lower forces, and with greater care. Studies have also speculated that shorter roots may experience less root resorption due to 1) shorter teeth requiring less force to move and 2) the root tip being displaced a shorter distance through the bone in second and third-order tipping motions [18,19].
In the Caucasian male group, the maxillary incisors experienced statistically significant orthodontic root resorption. This finding agrees with the literature as the upper incisors are commonly the most affected teeth in orthodontic root resorption. However, in the Caucasian female group, root resorption was found in nearly every tooth measured and to a greater extent. There was no statistically significant root resorption found in the Hispanic male group. Conversely, statistically significant root resorption was found for four teeth in the Hispanic female group: UR2, UR1, UL1, and LR2 but the percentage decrease in relative root length was still mild in comparison to the degree of root resorption that both Caucasian groups (male and female) experienced. These findings that females experience greater OIIRR than males agree with other findings in the literature [10,20].
Most teeth were bilaterally affected; however, resorption of the mandibular second premolar was the tooth that most commonly presented unilaterally. One of the limitations of this study was the assessment of closed apices in the SRA group. SRA roots are characterized by wide pulp chambers and truncated roots which is very similar in appearance to roots that are still undergoing development. Some of the second premolar measurements displayed slightly longer final root lengths than initial. Explanations for this finding are 1) some form of error in measurement, or 2) the second premolar root lengths were measured in the Hispanic group prior to cessation of root development.
Future studies are needed to complete a more detailed analysis regarding the nature of the genetic inheritance and prevalence of the SRA condition in the Hispanic population. In this study, Hispanic SRA patients were found to comprise approximately 2% of the total patient population treated in the university orthodontic clinic. A point of interest would include the geographic origins within Mexico and inheritance patterns of SRA Hispanic patients. Also, while there is