Clinical Implications of Growth Hormone Deficiency for Oral Health in Children: A Systematic Review

Growth hormone (GH) is involved in the regulation of the postnatal dental and skeletal growth, but its effects on oral health have not been clearly defined. This paper aims to provide a review of current clinical knowledge of dental caries, tooth wear, developmental enamel defects, craniofacial growth and morphology, dental maturation, and tooth eruption in growth hormone deficient (GHD) children. A systematic review was carried out using Scopus, MEDLINE-EbscoHost and Web of Science from 2000 to May 2021. PRISMA guidelines for reporting systematic reviews were followed. All the selected studies involved groups under eighteen years of age, covering a total of 465 GHD patients. The studies that were selected provide reliable evidence for delayed dental maturity and orthodontic disturbances in GHD patients. Data on dental hard tissues pathology are scarce and are limited to occurrences of dental caries. GHD children showed abnormal craniofacial morphology with reduced mandibular dimensions, with a resulting tendency towards Angle’s Class II occlusion, which affected up to 31% of patients. Dental age has been shown to be delayed in GHD patients by about 1 to 2 years. Moreover, the risk of dental caries in children with GHD decreases with increasing levels of vitamin D. Hence, further studies would be valuable for evaluating the risk of various oral health problems and to organize targeted dental care for this vulnerable group.


Introduction
Tooth development and eruption, as essential parts of general development, are examples of processes that can be easily disturbed. It is well known that the development of the alveolar bone surrounding the tooth germs is closely coordinated with tooth morphogenesis. At the same time, eruption also depends on precisely regulated bone remodeling [1,2].
Growth hormone (GH) is a critical regulator of the growth process in children. It is secreted by the pituitary gland, mainly during physiological night sleep. GH exerts its function mainly by promoting insulin-like growth factor I (IGF-I) secretion, acting within the GH-IGF-I signaling axis [3,4].
Despite numerous studies and the seemingly clear effect of GH on dental development, this process is still not well understood [5]. Research indicates that growth hormone action is associated with tooth maturation and eruption [6,7]. GH is able to induce proliferation of epithelial stem cells in molar buds, along with preameloblast differentiation and enamel formation [8,9]. GH and IGF-I induce the production of morphogenetic proteins 2 and 4 (BMP-2, BMP-4) and of the transforming growth factor-beta superfamily, affecting odontoblast differentiation and osteodentin and tubular dentine formation [8,10,11]. Furthermore,

Search Strategy
We searched Scopus (2000 to May 2021), MEDLINE-EbscoHost (2000 to May 2021), and Web of Science (2000 to May 2021). As the aim was to present the most recent summary of evidence, the analysis covered publications from the last 20 years, from January 2000 to May 2021, in English or Polish. The present systematic review has been designed to answer the question "Are children diagnosed with GHD, when compared to heathy or GH treated children, more often affected by oral health problems?" formulated according to PICO ("Population", "Intervention", "Comparison", "Outcome") [36]. The search strategy was carried out using MeSH (Medical Subject Headings) synonyms and Boolean logical operators. The following terms were used in search engines: "growth hormone deficiency" AND ("enamel" OR "tooth wear" OR "caries"); "growth hormone deficiency" AND "dental maturity"; "growth hormone deficiency" AND "craniofacial morphology"; "growth hormone deficiency" AND "craniofacial growth"; "growth hormone deficiency" AND "malocclusion". The first search was made on 05 January 2021 and the final one on 02 May 2021.
Studies were selected independently by two review authors (N.T.-W. and K.A.M.) Eligibility was determined by discussion where there were discrepancies. We searched manually for additional studies by cross-checking the reference lists of all the included studies. Duplicate publications were removed. Two of the review authors (N.T.-W. and

Data Extraction and Management
Two review authors (N.T.-W. and K.A.M.) independently extracted all the relevant data from the eligible studies and recorded it on a specifically designed form. Any discrepancies were resolved by discussion. If no agreement could be reached, arbitration was carried out by a third review author (J.O.-S.) Data extraction included the following: name of authors, country of patients, year of publication, study group, control group, study design, prevalence of dental caries, tooth wear, tooth eruption and dental maturity, enamel defects, malocclusion, and craniofacial growth and morphology in pediatric GHD patients.
To evaluate the risk of bias, reviewers independently (N.T.-W., K.A.M) evaluated the methodological quality of the studies using the adopted version of Newcastle-Ottawa Quality Assessment Scale according to a star-based system [37,38]. Any discrepancies were resolved by the third author (J.O.-S.) Each study was judged on three categories: the selection of the study groups, the comparability of the study groups, and the ascertainment of the outcome. A study could be awarded a maximum of two stars for the comparability category, the ascertainment of the factor and the assessment of outcomes items and a maximum of one star for each other numbered item within the selection and outcome categories.
The following criteria of reliability were used: ≥7 stars represented a low risk of bias (good quality study), 5-6 stars a medium risk of bias (fair quality study), and ≤4 stars a high risk of bias (poor quality study).

Results
After duplicates were manually eliminated, our systematic search of the three medical databases yielded a total of 62 publications meeting the search criteria. An initial selection of these was made using their titles and abstracts.
A total of fifty one articles were excluded because they focused on topics other than dental status, they did not study pediatric GHD patients, or they were reviews. Based on the full text, ten publications were qualified for further analysis; two of these concerned the mineralized tissues of the tooth-dental caries [39,40], two dealt with dental maturity [28,41], two with malocclusion [28,42], and six with craniofacial growth or morphology. One paper was excluded because of inconsistent data [2,28,[43][44][45][46] (Figure 1).
Two articles worked with the same group of patients [39,40]; for this reason, some results were considered only once.

Hard Mineralized Tissue Pathology
Dental caries were examined in two publications presenting the same patient cohort [39,40] (Table 2). The dental examinations were carried out in line with World Health Organization (WHO) criteria for epidemiological studies. The severity of dental caries was assessed using the DMFT index, which identifies those teeth (T) which have cavities (D); are missing (M); or have been filled (F) as a result of caries. A statistically significant effect of vitamin D3 concentration on the DMFT index and its component DT was found among children from rural areas, where an increase in vitamin D3 concentration by ten units resulted in a decrease in the value of DMFT by 0.82 and a decrease in the value of DT component by 0.66. The percentage of these children with active caries was higher than in urban areas, but not statistically significantly [39]. A positive and statistically significant correlation between the duration of GH therapy and DMFT index was, however, observed in patients from urban areas [40]. There was no healthy control group in this study.

Dental Maturity and Malocclusion
Two of the papers we considered describe the prevalence of malocclusion in GHD children [28,42] (Table 3). Both used the relations of first permanent molars (Angle's classification) to detect deviations from Angle's Class I occlusion, where the mesiobuccal cusp of maxillary first molar occludes in the buccal groove of the mandibular first molar. In the study of Kjellberg et al. [28], 29% of the boys in the study group showed Angle's Class II malocclusion, while the remainder were Angle's Class I. Dental crowding of at least 2 mm was recorded in 44% of patients. A large overjet (>6 mm) was seen in 14%, and a large overbite (>5 mm) in 5%.
Hodge et al. [42] observed Angle's Class II in 31% and Angle's Class III in 6% of patients. Increased overjet and deep overbite were each found in up to 37% of subjects, which is a significantly greater prevalence than in Kjellberg et al. [28] However, these discrepancies are probably due to differences in methodology and definitions. In the study of Hodge et al. [42], an overjet greater than 2 mm and an overbite greater than 3 mm were considered abnormal, while Kjellberg et al. [28] noted only more extreme abnormalities. Unlike Hodge et al. [42], Kjellberg et al. [28] used radiographs and plaster models to record relations between the jaws. Dental maturity was evaluated in two studies by Kjellberg et al. [28] and Partyka et al. [41] Each investigator used a different method: the method of Demirjian was employed by Kjellberg and the method of Matiegka and Lukasova by Partyka; both of which were validated. Kjellberg et al. [28] defined dental maturity on the basis of tooth formation recorded on orthopantomograms. The sum of scores for each individual was converted into a dental age in accordance with the instructions given by Demirjian. The method of Matiegka and Lukasova established dental age by identifying the most recently erupted full group of teeth, including incomplete groups. From Matiegka's table for boys and Lukasova's for girls, age corresponding with the number of teeth can be found, giving a result for a specific patient [41]. Both studies showed statistically significant differences between birth age and dental age between the GHD and non-GHD patients and control groups [28], and between birth age and dental age in patients starting treatment [41]. In Kjellberg et al. [28], dental maturity was delayed about one year in both the non-GHD and GHD boys. Partyka et al. [41] reported a delay of 18.82 and 2.70 months (for the group starting treatment and that undergoing treatment, respectively).

Craniofacial Growth/Morphology
Six articles on craniofacial growth and morphology were included in the systematic review [2,28,[43][44][45][46] (Table 4). Five publications used lateral cephalograms to measure the results [2,28,[43][44][45]. The number of landmarks and the linear and angular measurements differ between the studies, and three articles mentioned the methods used: Kjellberg et al. [28] used the Bjork method, Choi et al. [45] used Pancherz's method, and Segal et al. [46] used the triangulation methods developed by Bookstein. Anterior cranial base length was found to be significantly reduced by Preda et al. [44], Kim et al. [43], and Choi et al. [45], while the posterior cranial base length was shorter in Kjellberg et al. [28], Preda et al. [44], Choi et al. [45], and Kim et al. [43] Total cranial base length was significantly less in Preda et al. [44] and Kim et al. [43] Lower anterior facial height was significantly smaller among boys and girls prior to and during treatment in Funatsu et al. [2] Segal et al. [46] also found smaller vertical proportions, suggesting a deficiency in the lower face. Both mandibular ramus height and corpus length were shorter in boys prior to treatment in Choi et al. [45] Smaller mandibular corpus lengths were noted by Preda et al. [44] among boys and girls, and prior to and during treatment in Funatsu et al. [2], while mandibular ramus lengths were shorted in untreated boys in Kim et al. [43] The measured angles referred to the mandible's and maxilla's retroposition [28,43,44]. Significant differences between the studied group were also apparent in the angle between the maxillary and mandibular planes, which was larger than normal [43][44][45]. Ans-Me, <Gn-Cd,A'-Ptm',Pog'-Go, Cd-Go were significantly smaller in boys and ANS-Me,Gn-Cd, Pog'-Go in girls in untreated group. Cg-Go was significantly larger; SNA, gonial angle were significantly smaller in boys; gonial angle in girls in short-term therapy. N-Me', ANS-Me, A'-Ptm', Pog'-Go, Cd-Go, gonial angle were significantly smaller in boys and ANS-Me, Cd-Go, gonial angle were significantly smaller but A'-Ptm' was significantly larger in girls in the long-term. There was a significant difference between the untreated and long-term therapy in upper facial height, maxillary length and ramus high-scores increased with the duration of GH therapy Choi et al. (2017) [45] Lateral cephalograms at T0 before the treatment, T1 2 years after treatment, Pancherz's method; 9 linear and 7 angular measurements   [44] Lateral cephalograms, 11 linear and 6 angular measurements n-s s-ba n-ba n-sp pm-sp sp-gn gn-go ss-pm ss-ba s-pm pm-ba SNA, SNB, ANB, ML-NL s-n-sm s-n-ss SNA, SNB were significantly smaller (p < 0.001), ANB was higher (p < 0.001). S-n-ss and s-n-sm were significantly lower (p < 0.001). Linear measurements -N-s (p = 0047), sp-gn (p = 0.008), gn-go (p = 0.003), s-ba (p < 0.001), n-ba (p < 0.001) were significantly reduced.
Kim et al.

Quality Assessment and Risk of Bias
All studies were classified in accordance with the Cochrane collaboration guidelines [33]. A control group was used in all publications, although its size and structure was not always consistent. While only three groups were gender-paired [2,28,45], most of them were similar to the study group in terms of age [2,28,39,40,[42][43][44].
The control group in the study by Kjellberg et al. [28], Segal et al. [46], Hodge et al. [42], and Preda et al. [44] consisted of children from previous studies. None of the publications described the blindness of examiners. Intrarater and interrater reliability were calculated in Kjellberg et al. [28] and Choi et al. [45] All papers performed statistical analysis, although not all aspects were statistically analyzed in one of the studies [28].
The publications on malocclusion [28,29] were found to be at medium risk of bias. Two articles on dental caries describe the same cohort of children and present similar conclusions [39,40]. One of them was classified as a good quality study, since it additionally included a sample size calculation [40]. One publication on craniofacial characteristics [44] was at medium risk, and five [2,28,43,45,46] were at low risk. One publication on dental maturity [41] was at medium, one at low risk [28] (Table 5).                                  Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).         Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).

Comparability Outcomes
Appraisal Score  Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).

Comparability Outcomes
Appraisal Score  Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).

Comparability Outcomes
Appraisal Score  Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).

Comparability Outcomes
Appraisal Score      Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).  Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).  Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).

Comparability Outcomes
Appraisal Score  Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).  Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).  Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).                Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).

Comparability Outcomes
Appraisal Score                  Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).  Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).  Table 5. Quality assessment of studies using the Newcastle-Ottawa Scale (✯ -star (point) awarded in the quality assessment).

Comparability Outcomes
Appraisal Score

Discussion
Through their influence on bone metabolism, GH and IGF-I are major regulators of postnatal growth and development. GH acts directly on tissues by means of GH receptors, or indirectly by the production of insulin-like growth factor I. Metabolic agents and the growth hormone/insulin-like growth factor-I axis have a strong influence on the metabolism of oral tissues, particularly during the period of growth [8].
There is very little in the literature on oral cavity status in patients with GHD. Our systematic review has shown that some dental topics have not yet been discussed. There has been little to evaluate dental conditions like tooth wear and enamel defects, although we can assume that the effect of growth hormone on the dentition and facial bones is complex [3,8]. Publications on dental status and craniofacial growth in children with GHD are somewhat confined, and their results are not always concordant [6].
Some researchers have suggested a relationship between tooth wear and craniofacial morphology, as well as a correlation between tooth wear and malocclusion [47,48]. Tooth wear is defined as the mechanical or chemical removal of dental hard tissues, resulting in reduced tooth structure. The prevalence and severity of tooth wear in contemporary populations is on the increase, particularly in younger patients [49]. Patients with significant tooth wear have been described as having a characteristic craniofacial morphology. Cephalometric analysis has reported a reduction in lower anterior facial height, a more horizontal mandibular plane angle, a more acute gonial angle, and a greater posterior facial height [47]. It would be useful to examine these features in GHD children.
The size, growth, and osseous maturity of the jaw also play a role in the process of tooth eruption. A strong correlation has been shown between eruption time and dental maturity. The teeth typically erupt when they have reached a 2/3 root length [50], but individual correlation between chronological age and eruption time is inconsistent [50,51]. Research has demonstrated that, in GHD patients, dental age (maturity) is significantly delayed [28,41]. This is consistent with the results of Cantu et al. [27], which indicated a mean delay in dental age of close to one year. Furthermore, they observed no significant effect of GH treatment on dental maturation. The lack of a subsequent therapeutic response would indicate that dental age is less affected by GH than craniofacial growth.
It thus appears that the increasing maturity of teeth and eruption requires further investigation. It can be assumed that the tooth maturation process and the eruptive movements of the tooth after crown formation are endocrinologically controlled [50].
The studies involved in our systematic review report that not only the height of GHD children, but also their craniofacial morphology and growth, are affected [43]. These studies support the previously demonstrated idea that the linear growth of the body is strongly correlated with jaw growth [49], and that the growth of craniofacial skeletal structures is poor in periods of slow longitudinal growth [2,28,[43][44][45][46].
Several linear craniofacial measurements have been found to be shortened in GHD patients, particularly the mandible and the cranial base [2,28,[43][44][45][46]. In children with GH deficiency, it is the mandible that is small, especially the ramus length [2,43]. The most pronounced facial growth retardation is found for posterior face height [43].
Due to mandibular growth retardation, the mandible can be rotated backwards, and the dental-alveolar compensatory mechanism can be activated vertically in the anterior region, in order to maintain incisal contact for as long as possible [6]. It has been observed that males with GHD prior to treatment had a tendency to exhibit skeletal Class II [45]. However, Angle's Class II malocclusion was not as prevalent as expected given the retrognathic mandibular positions and reduced mandibular dimensions seen in many of the boys. This may be explained by the ability of the occlusion to adapt to slow changes during growth [28].
GH therapy induces the most pronounced catch-up growth within the first one to two years. Funatsu et al. [2] stated that GH therapy was started at a younger age in those in the long-term therapy group than in those in the untreated or short-term therapy groups.
It was postulated that the GHD in the long-term therapy group was more severe than in the others. This is in agreement with the conclusion of Cantu et al. [27], which postulated that catch-up may depend not just on growth potential, but also on accumulated growth deficits at the beginning of growth hormone replacement therapy [52].
Dental caries have still not been extensively studied in GH-deficient children. Caries resistance was first suggested by Nikiforuk et al. [53], who concluded that the etiology of this condition most probably lay in the increased maturation time of enamel tissue before the eruption and the reduced exposure to environmental factors. The more recent study by Schroth et al. [54] reports that caries-free children were twice as likely to have optimal 25 (OH)D concentrations (>75 nmol/L), and those with caries presented deficient levels (<35 nmol/L). This was confirmed by Wójcik et al. [39,40], who related the lower prevalence of caries in GHD children the higher 25 (OH)D level. It was concluded in those studies that an increase in vitamin D3 concentration by ten units decreased DMFT by 0.82 and DT by 0.66. The relatively poor range of data on the level of caries in GHD patients suggests the need for further observations.
The occurrence of dental caries and tooth wear should be studied further in cases where the primary teeth remain longer in the oral cavity. After all, we know that the condition of the mineralized teeth tissues in both generally healthy patients and in those with GHD transfers across to the condition of the stomatognathic system. The healthier the masticatory organ, the better the condition of pediatric patients entering adulthood.
Birth age does not reflect fully the physiological development of a child. In order to closely evaluate the process of growth, it is necessary to use other criteria, such as dental age and skeletal age. These parameters are essential for dental providers to provide diagnoses and to plan therapy. In clinical practice, evaluation of both dental and skeletal age would be valuable in all children undergoing dental treatment, especially in those with GHD.
This study has several limitations. The first is the small number of GHD children examined in some of the studies, which makes the results difficult to compare. The studied groups differed in the number of participants, and sometimes the results of a much larger number of children were analyzed in the control group [28,42,44,46]. Although children with various medical conditions were excluded, the control groups did not only contain healthy children, and the comparison was made between the children both at the beginning of treatment and during treatment. The lack of information on the conditions of the dental examinations, especially regarding the qualifications of the examiners, makes the studies prone to bias. Our study search strategy was limited to English and Polish language papers from the last 20 years; this approach may have resulted in the omission of some reports, but it would be difficult to relate the results of studies conducted over twenty years ago to current conditions.

Conclusions
The available studies indicate that children with GHD showed abnormal craniofacial morphology with reduced mandibular dimensions, with a resulting tendency to Angle's Class II occlusion, which affected up to 31% of the patients. Dental age has been shown to be delayed in GHD patients by about 1 to 2 years. Moreover, the risk of dental caries in children with GHD decreases with increasing levels of vitamin D. The data are scarce and further studies would be valuable in evaluating the risk of various oral health problems and in organizing targeted dental care for this vulnerable group.
To gain more of an insight into the effects of this disease and its treatment on oral health and craniofacial structures, data need to be collected both before and after GH administration. Such longitudinal studies could help us to understand the complex endocrine mechanisms regulating the stomatognathic system's development and functions, in order to provide the optimal treatment of GHD-related disturbances.