Molar-Incisor-Hypomineralisation (MIH). A retrospective clinical study in Greek children. II. Possible medical aetiological factors

Transcription

1 Molar-Incisor-Hypomineralisation (MIH). A retrospective clinical study in Greek children. II. Possible medical aetiological factors N. A. Lygidakis, G. Dimou, D. Marinou. Dept. of Paediatric Dentistry, Community Dental Center for Children, Athens, Greece. Abstract Aim: This was to examine the potential medical aetiological factors involved in the development of MIH. Methods: During the years , all MIH cases diagnosed according to set criteria were selected from the new patients clinic of a Community Dental Centre for Children (Athens). The age, gender and teeth involved were recorded. A control group of socio-demographically matched controls was also identified. The potential aetiological factors were retrieved through personal interview with the parents and from each child and mother s medical book. Only verified aetiological factors were recorded. Evaluation of the correlation of affected teeth and the timing of the insult was performed in a separate group of 225 affected children aged 8-12 with their entire 12 index teeth erupted. Results: From the 3,518, 5.5 to 12 years old children examined, 360 (10.2%) had MIH. Aetiology of MIH: 44 children (12.2%), presented without any relevant medical history, the remaining 316 (87.8%) recorded various medical problems associated with MIH, compared with 18.9% for controls. Perinatal (163, 33.6%) and postnatal (162, 33.9%) problems were the most frequently found and prenatal the least (33, 8.6%). For 42 children (11.7%) problems occurred in more than one chronological period, mainly during both the perinatal and postnatal period (11.1%). The most common prenatal problem was repeated episodes of high fever (12/33), in the perinatal period birth by Caesarean section (92/163) and other birth complications (34/163). Various respiratory conditions (88/162), repeated episodes of high fever (31/162) and neonatal illness (28/162) were the commonly reported problems in the postnatal period. Many MIH cases presented with more than one medical problem during the peri-and postnatal period. Statistical analysis: Children with MIH recorded 68.9% more frequent medical problems than controls (p<0.0001). A positive correlation (p<0.001) between the total number and type of affected teeth with the timing of the insult was observed in the 225 MIH children with all their index teeth erupted. Conclusion: Children with MIH present with more medical problems than controls during their prenatal, perinatal and postnatal period. The majority of these illnesses may produce hypocalcaemia, hypoxia and pyrexia to the child or the mother. The number of affected teeth was associated with the timing of the possible insult; children with prenatal, perinatal and postnatal problems present more affected teeth in increasing order. Introduction Molar-incisor-hypomineralisation (MIH) is defined as the developmentally derived dental defect that involves hypomineralisation of 1 to 4 permanent first molars (FPM) that is frequently associated with similarly affected permanent incisors. The defect is clinically presented as demarcated enamel opacities of different colour in the affected teeth, occasionally undergoing post-eruptive breakdown due to soft and porous enamel. This may result in atypical cavities or even complete coronal distortion, requiring extensive restorative treatment [Lygidakis et al., 2003; Mathu-Muju and Wright, 2006]. Due to profound sensitivity of affected teeth, children are reluctant to carry out effective oral hygiene (OH) and to accept dental treatment, being at risk of developing dental phobias and presenting behaviour management problems [Javelik and Klingberg, 2002]. Histologically the defective enamel is partly hypomineralised, with well-defined borders between the defective and normal enamel [Javelik and Noren, 2000]. Prevalence. Following the establishment of the presently used diagnostic criteria [Weerheijm et al., 2003], very few well-documented studies have been undertaken concerning the prevalence of MIH [Jasulaityte et al., 2007; Lygidakis et al., 2008]. Previous studies using a variety of criteria have shown that the prevalence of the defect is between 3.6 and 25% in Europe, while there is a relative lack of data concerning the Americas and elsewhere [William et al., 2006; Jasulaityte et al., 2007]. Aetiology. A variety of systematically acting medical factors have been proposed as contributing to or causing MIH, including prenatal, perinatal and postnatal illnesses, low birth weight, antibiotic consumption and toxins from breast-feeding [William et al., 2006]. Children with poor health during the first years of life, the critical period for crown formation of the FPM and incisors, are more likely to be at increased risk for MIH [Jalevik and Noren, 2000]. It also has been proposed that there could be an underlying genetic predisposition that contributes to the risk of developing MIH in at least some cases [Brook and Smith, 1998]. The varying degree of enamel defects in FPM and incisors, that develop at the same time, suggest that not all teeth are equally susceptible to enamel defects and developmental disturbances. Either genetic or spatial differences could play a part in the devel- Key words: molar-incisor-hypomineralisation, clinical study, aetiology Postal address: Dr Nick A. Lygidakis, 2 Papadiamantopoulou St., Athens 11528, Greece. lygidakis@ath.forthnet.gr 207

2 Lygidakis et al. opment and variability observed clinically in MIH. Collectively, the majority of previous studies imply that the aetiology of MIH is complex with undetermined systematic and genetic factors disrupting normal amelogenesis in the affected teeth [Mathu-Muju and Wright, 2006]. The aim of the present study was to assess the potential medical aetiological factors in a group of MIH children diagnosed according to the recently set criteria. A control group was used in order to evaluate whether any findings were different from the normal population. Finally, various correlations in the MIH group were examined for possible interrelations between the number and type of teeth involved and the timing of the insult onset. Material and Methods Population. During the years , all MIH cases were selected from the new patients clinic of the Community Dental Center for Children in Athens, which accepts for dental treatment children up to the age of 12 years. A pilot study undertaken in 2002 set the clinical procedure for the fulfilment of the present study [Lygidakis et al., 2004]. The criteria used for the diagnosis of MIH were those described in a European meeting held in Athens in 2003 [Weerheijm et al 2003]. All children with MIH had at least one FPM erupted, at least partly, during the time of examination as has been suggested as a diagnostic requirement [Jalevik and Noren, 2000]. Recording of defects. The age, gender, teeth involved and the severity of the defect were recorded. In order to evaluate more precisely the potential interrelation between the timing of the medical insult and the total number and type of teeth affected, a separate sub-group of MIH children with their entire 12 index teeth erupted (4 FPM and 8 incisors) was identified. There were 7 children that had 1-3 canines affected in addition to FMPs and incisors and these were regarded as having only incisors and molars affected in order to conform to the MIH criteria. Following a professional cleaning of the teeth, clinical examination of the cases was performed in a dental chair using mirror, probe and dental light, by either one of the first two authors or by both of them, who were previously calibrated for the diagnosis of MIH in the pilot study [Lygidakis et al., 2004]. Inter-examiner reproducibility was calculated and was found to be high in all parameters examined (kappa= ). Aetiology. Possible medical aetiological factors were recorded by a) a detailed personal interview with the parents where all childhood and pregnancy medical history was asked, b) from the child s official State medical book and c) their mother s medical history during pregnancy and delivery as recorded in their medical insurance book. All patients and families were insured with the same Social Security Insurance, covering all privately employed Greeks. All Greek children have an additional medical book in which all their illnesses and treatments provided are recorded, usually up to the age of 6 years. The parents completed a consent form in order to have the medical history retrieved from the medical book. Aetiological factors were divided into prenatal, perinatal and postnatal according to the timing of the insult. Postnatal problems were defined as those appearing up to the age of 1 year, thus including the neonatal (up to1st month) and infancy period. Low birth weight and twinning were classified in the perinatal problems, as delivery problems were also implicated. Only verified aetiological factors, as written in the medical records, were recorded. Control population. A separate group of children was selected from the same child population, examined in the new patients clinic during , as a control group. Accordingly, 360 children examined exclusively for orthodontic problems were included in the control group. In order to form a random and representative control group, children were matched for age, gender, ethnicity and socio-economic family status. In a consistent way every 3rd normal child that was fulfilling the previously mentioned criteria was included in the control group. Medical information of the child and the mother was retrieved the same way as for the MIH children. Statistical analysis. This was performed using a SPSS program for Windows. Parametric and non-parametric statistical test were used and the significance level was set at p<0.05. Results Population. During the three year period, 6,983 children aged 1-12 years, were examined for a first time, from these 3,518 were 5.5 to 12 years old, the remaining being of preschool age. By the end of the third year of the study 360 children with MIH had been identified, the overall prevalence in the study population being 10.2%. The age span of the affected children at the time of examination was years (mean age 8.17±1.38). Further demographic details of the patients have been presented in the Part I of the present clinical trial together with number, type and severity of affected teeth [Lygidakis et al, 2008]. All cases except for 24 belonged to different families, with no relationship between them. The 24 cases consisted of 2 brothers of a diabetic mother, 12 twins (six sets) and 5 sets of first cousins. Distribution of potential aetiological factors in the MIH children. As shown in Table 1, 44 children (12.2%), presented without any type of medical history. The remaining 316 (87.8%) revealed various medical problems that have been associated with MIH. Regarding the time of the insult, perinatal (33.6%) and postnatal (33.9%) problems were the most frequently found, while prenatal were the least (8.6%). Problems in more than one chronological period occurred in 42 affected children (11.7%). The major occurrence in this category was the combination of perinatal and postnatal medical conditions (11.1%). 208

3 MIH Aetiology Table 1. Distribution of the timing of the possible aetiological factors in 360 MIH children, 225 MIH children with all their 12 index teeth erupted and 360 controls in a Greek population. Aetiological factor timing Total MIH cases (360) MIH cases with all their index teeth erupted (225) Controls (360) Number % Number % Number % Unknown Prenatal Perinatal Postnatal Perinatal+postnatal Prenatal+perinatal Concerning the actual type of the medical condition recorded, in the prenatal group cases of repeated episodes of high fever were more frequent (12/33), in the perinatal group Caesarean section (92/163) and in the postnatal group various upper and lower respiratory medical problems (88/162). There were also repeated episodes of high fever due to common cold/coryza (31/162) and neonatal problems during the first month of life (28/162) (Table 2). An important finding was the frequent recording of more than one medical problem during the perinatal and postnatal period. There were 163 children with MIH who reported 185 medical problems in the perinatal period, whilst 162 children with MIH were associated with 203 medical problems in the postnatal period. An example of multiple problems was the association of 45 from the 92 born by Caesarean sections with problems in the perinatal (complicated delivery/ preterm birth) and postnatal (frequent respiratory illnesses) (Table 2). Distribution of medical problems in the control group. As can be seen in Table 1, in 68 children (18.9%) of the 360 control subjects medical problems of the same type as those recorded in the MIH group were noted. There were no cases of prenatal medical conditions, while more postnatal (45, 12.5%) than perinatal (23, 6.3%) problems were found. Perinatally the more frequent problem was prolonged/difficult delivery (10/23), followed by Caesarean section (8/23) and premature birth (5/23); postnatally respiratory problems, eg. otitis, bronchitis, asthma (25/45). There were also repeated episodes of high fever due to common cold/coryza (13/45) and other less frequently found conditions, such as neonatal illnesses, seizures, encephalitis and urinary infection (8/45). In contrast to the MIH cases, in the control group children there were no cases of multiple illnesses in the same or in more than one chronological period. Comparison between MIH cases and controls. There was a statistically significant difference between affected and control group, in the number of children recording medical problems with 87.7% of MIH cases being associated with a medical complication as compared with only 18.9% in the control group (Table 1) (McNemar test, p<0.0001). That means that MIH children presented 68.9% (95%CI=[60%-77.8%]) more frequent medical conditions than controls. Similar significant differences were found for separate comparisons in the cases of perinatal (p<0.001) and postnatal medical history alone (p<0.001). In the case of perinatal aetiology, MIH children presented 38.9% (95%CI=[31.8%-46.0%]) more frequently with medical conditions than controls, while in the postnatal group, 33.1% (95%CI=[25.8%-40.5%]). There were no cases of prenatal problems in the control group as compared with the MIH cases. Correlations between number of affected teeth in MIH cases with their entire index teeth erupted (age 8-12) and the timing of possible aetiological factor. In order to evaluate these possible correlations, a separate group was formed of affected children aged 8-12 years with their entire 12 index teeth erupted. The distribution of aetiological factors in this sub-group of 225 children is also shown in Table 1. As is apparent, when they are compared with the results from the total MIH cases, both groups reveal similar percentages, indicating the persistence and value of the results. In Figure 1 the distribution of the number of teeth affected in correlation with the timing of each aetiological factor is shown. Excluding cases of unknown aetiology (n=28), children recording medical problems during the combined perinatal/postnatal periods had significantly more affected teeth than those exposed to complications in the postnatal and perinatal period alone; children with problems in the prenatal period presented with significantly less affected teeth (Kruskal Wallis test=23,65, p< 0.001). It is worth mentioning that all 7 children with canines affected in addition to FPM and incisors, although not assessed as noted in Materials and Methods, belonged to the combined perinatal/postnatal group. As it can be seen in Figure 1, correlations were also evaluated regarding the total separate number of incisors and FPM affected. Statistically significant differences were found in both cases (for FMP Kruskal Wallis test=25,83, p=0.001, for incisors Kruskal Wallis test=12,79, p=0.005), the number of af- 209

4 Lygidakis et al. Fig 1. Distribution of MIH affected teeth (total number, molars, incisors) in comparison with the timing of the known potential aetiological factor, in the sub-group of 225 Greek children with all their index teeth erupted. Combined prenatal/perinatal problems were excluded, due to small numbers. Mean (±SD) number of affected teeth fected teeth being as noted previously, greater in cases of combined perinatal/postnatal aetiology followed by those in postnatal, perinatal and prenatal period in descending order. Correlations between the type and location of affected teeth in MIH cases with their entire index teeth erupted (age 8-12) and the timing of possible aetiological factor. Table 3 summarises the relationship between the type and location of affected teeth with the timing of the medical problem. Overall, maxillary teeth in total were more frequently affected than mandibular, irrespective of the insult period (p= ). For posterior teeth, both affected mandibular FPM were statistical significantly less frequently found in cases of prenatal aetiology and more in the cases of combined perinatal/postnatal aetiology (p<0.001), while both maxillary molars were very frequently found (85-96%) in all different cases of aetiology. However, their distribution within the aetiology groups was statistically non-significant (p>0.005) (Table 3). Overall, maxillary and mandibular affected molars were equally found in all MIH children, regardless of aetiology, with the exception of prenatal aetiology, where maxillary molars were more frequently found than mandibular (p=0.013) (Table 3). When anterior teeth were considered non-significant associations (p>0.005) were found in cases of mandibular lateral incisors when correlated with the period of insult, both of them being much less frequently found than the rest of the affected teeth in all cases of aetiology (0-16%) (Table 3). Additionally, both mandibular central incisors were statistical significantly less frequently affected in cases of prenatal aetiology (5%) and more with postnatal aetiology (35.4/36.7%) (p= ) (Table 3). Furthermore, both maxillary central incisors were statistical significantly more frequently affected in cases of combined perinatal/natal aetiology (84/88%) and less in cases of prenatal aetiology (50/55%) (p<0.001). Regardless of aetiology, maxillary incisors were always more frequently affected than mandibular (p<0.005) (Table 3). Discussion MIH is an interesting disorder, although the so called extensive disruption of molar enamel was evident centuries ago, as has been proven recently in sub-adults retrieved from the post-medieval Broadgate cemetery in London [Ogden et al., 2007]. Only in the last decade has the condition attracted the interest of the dental profession. The decline of dental caries in the Western world has allowed researchers to concentrate on problems that were attracting less interest in the past. Concerning aetiology, as early as 1981, Nikiforuk and Fraser, made an attempt to explain the aetiology of what was called at that time hypoplasia of non-genetic or local aetiology. In their pioneer work it was concluded that hypocalcaemia during amelogenesis is a crucial factor that can lead to enamel hypoplasia of environmental aetiology. They studied cases of rickets and hypoparathyroidism and found that enamel hypoplasia was evident only in those cases with hypocalcaemia 210

5 MIH Aetiology Table 2. Distribution of the possible aetiological factors in a group of Greek children affected by MIH according to the onset chronological period. Possible aetiological factor Prenatal aetiology (33 cases) N (%) % in the total MIH group Multiple episodes of maternal high fever/viral infection during last month 12 (36.3) 3.3% Prolonged medication (myometrium spasmolytics) during last month 6 (18.2) 1.6% Maternal diabetes 5 (15.1) 1.38% Prolonged vomiting up to last month 5 (15.1) 1.38% Malnutrition 2 (6) 0.5% Chicken pox last month of pregnancy 1 (0.3) 0.27% Renal deficiency 1 (0.3) 0.27% Maternal hypertension 1 (0.3) 0.27% Perinatal aetiology (163 cases) % in the total N (%) MIH group Postnatal aetiology (162 cases) % in the total N (%) MIH group Caesarean section 92 (50.5) 25.5% Prolonged/complicated delivery 34 (18.3) 9.4% Twins 29 (15.6) 8% Premature birth-low birth weight 29 (15.6) 8% Hemorrhage + detachment during delivery 1 (0.5) 0.27% Repeated (>5) episodes of high fever due to common cold/coryza 31 (15.2) 8.6% Otitis 34 (6.7) 9.4% Bronchitis 21 (10.3) 5.8% Asthma 15 (7.3) 4.1% Bronchiolitis 7 (3.4) 1.9% Laryngitis 6 (2.9) 1.6% Tonsillitis 5 (2.4) 1.38% Neonatal (first month of infancy) (respiratory problems, incubator, exanthematous 28 (13.7) 7.7% disease, high fever, seizures) Prolonged use of medication other than antipyretics 16 (7.8) 4.4% Seizures afebrile/febrile 12 (5.9) 3.3% Urinary infection 9 (4.4) 2.5% Encephalitis 4 (1.9) 1.1% Gastroenteritis 4 (1.9) 1.1% Exanthematous disease 2 (0.9) 0.5% Other (endocrine disturbance, anaemia, operations, cleft palate, salmonellas) 9 (4.4) 2.5% T0TAL ILLNESSES

6 Lygidakis et al. (vitamin dependent rickets and hypoparathyroidism) and not evident in cases of hypophosphataemia (X-linked rickets), revealing normal calcium levels in the blood. In an attempt to explain the possible aetiological factors in the MIH cases it is important to remember that between 28th week in utero and the first 10 days of life ameloblasts initiate amelogenesis in the first permanent teeth to be formed, the FPM, followed by other teeth later in time [Welbury, 1997]. This time schedule has been recorded in previous studies providing the dental development charts, although there are chronology diversions not easily detected as a result of methodology difficulties in neonates and infants [Moorrees et al., 1963]. It has been clearly shown that ameloblasts belong to the most sensitive cells of the human body. If their function is interrupted, temporarily or permanently, then depending upon the time of insult, enamel hypoplasia or hypomineralisation is produced [Simmer, 2001; Fearne et al., 2004]. Experiments have shown that conditions affecting the enamel matrix ph, i.e. respiratory acidosis and abnormal oxygen levels resulting from hypoventilation in various respiratory diseases, inhibit the action of the proteolytic enzymes and the development of the crystal hydroxyapatite resulting to enamel hypomineralisation [Whitford and Angmar- Mansson, 1995; Sui et al., 2003]. Ameloblastic function itself might be additionally affected by low oxygen levels found during the birth of preterm children [Johnsen et al., 1984; Aine et al., 2000]. It has been also suggested that the lack of calcium phosphate in the area of the crystallites might result in reduced calcium deposits and lower ratio of calcium/ phosphorous leading again to enamel hypomineralisation [Van Amerongen and Kreulen, 1995; Jalevik et al., 2001a]. To support this theory, recent experiments in animals using immunocytochemical analysis have shown that hypocalcaemia affects enamel during the late secretory and early maturation stages, interfering with both cellular and extracellular elements resulting in hypomineralisation [Nanci et al., 2000; Yamaguti et al., 2005]. Table 3. Correlations between type of teeth affected with MIH and the chronology of possible aetiological factor, in the sub-group of 225 children with all their index teeth erupted. Perinatal+postnatal Tooth Prenatal (20 cases) Perinatal (72 cases) Postnatal (79 cases) (25 cases) % FDI Present Absent presence Present Absent % presence Present Absent % presence Present Absent % presence Statistics** T * T * T * T T * T T * T T T T * T * Maxillary vs. mandibular teeth Z=-3.066,p=0.002* Z=-4.650, p=0.000* Z=-2.811, p=0.005* Z=-3.115, p=0.002* (total) Maxillary vs mandibular molars Z=-2.495,p=0.013* Z=-1.035, p=0.301 Z=-.354, p=0.723 Z=.000, p=1.000 Maxillary vs mandibular incisors Z=-2.810,p=0.005* Z=-4.604, p=0.000* Z=-2.877, p=0.004* Z=-3.203, p=0.001* (*Statistically significant,**pearson chi-square, z=wilcoxon signed Ranks test) p 212

7 MIH Aetiology In the present retrospective clinical study a wide spectrum of medical conditions of the mother and the child appear to be associated with MIH. It is known that retrospective studies of this kind reveal problems that may reduce their value, including the lack of definite diagnostic criteria in the past, the use of parents as the only source of medical information during their child s first year of life, the frequent coexistence of more than one medical problem of the subjects, the wide time spectrum of enamel formation of the affected teeth, etc. Taking these factors into account and until the completion of well-organised prospective studies, the present research made an effort to minimise these drawbacks in order to attain valuable results. In the present study, apart from the parents interview, medical information was retrieved from each child s medical book and mother s illnesses during pregnancy was verified from their insurance medical record. The retrieval of this information from these records means that illnesses were recorded only when a doctor had examined a child, thus excluding various everyday illnesses that do not require medical attention. This might underestimate some medical problems, but it is an accurate method to retrieve medical history. Also in the MIH group only children that fulfilled the clinical criteria set recently were included while questionable subjects were excluded. Finally an effort was made that the control group used for comparison was as representative as possible taking into account, apart from age and gender, the socioeconomic status and the ethnic origin of the children. Prenatal medical conditions. In the present study 33/360 of the MIH children (8.6%) revealed maternal prenatal problems, compared with none of the control group. Aetiological factors acting during this period produce fewer cases of MIH compared with all other periods, indicating that children are probably protected in utero. During the last gestational months, multiple maternal episodes of high fever, due to common cold or infections, were frequently found to be associated with MIH. This finding of maternal pyrexia has been shown experimentally to have a detremental influence on amelogenesis, ranging from ameloblastic dysfunction to complete cellular degeneration [Kreshover and Clough, 1953]. Similar to our results were those reported in a study of 33 children with MIH [Jälevik and Noren, 2000], where 15% of them recorded maternal chronic diseases during pregnancy, eg. syphilis, hypertension, elevated blood glucose, and prolonged use of drugs. However, in a later study the same group could not find any association of these conditions with MIH [Jalevik et al., 2001a]. Other prenatal medical problems evident in the present study included maternal diabetes, prolonged vomiting and use of spasmolytic medication in the late gestational weeks. Although these conditions have not been reported before, it is known that maternal diabetes produces hypocalcaemia in the mother and oxygen shortage problems to the infant that may result, as it was discussed before, in enamel hypomineralisation. A recent study has shown that 15% of infants of diabetic mothers suffer from hypocalcaemia and another 15% from oxygen shortage [Alam et al., 2006]. In addition, in cases of prolonged maternal nausea and vomiting, fluids and electrolytes, as well as nutritional status, are jeopardised leading occasionally to foetal biochemical disturbance [Nelson-Piercy, 1998]. If prolonged vomiting is present during the last gestational days neonatal hypoxia might be produced [Cyna et al., 2006]. Finally, although the prolonged use of myometrium spasmolytics has not been clearly implicated with enamel defects in the past, there is evidence that repeated use of b- Adrenergic antagonists during gestation, found in the present study in 6/33 the cases with prenatal problems, might have, among other side effects, maternal nausea and vomiting and foetal hypocalcaemia [Norwitz et al., 1999]. These side-effects might be a reason for MIH in these cases. Some other prenatal problems were detected in the mother s medical book in the present study. These included malnutrition, chicken pox, renal deficiency, hypertension, also associated with hypocalcaemia, indicating again potential links with MIH. Perinatal medical conditions. In the present study 121/360 of MIH children (33.6%) revealed perinatal problems, compared with 6.3% of the controls. Caesarean section, prolonged delivery, premature birth and twining were the most frequently found conditions. Previous studies have shown association of perinatal problems and MIH. In a group of 21 children with MIH, 48% of them recorded perinatal problems similar to those found in this study, e.g. premature birth, prolonged delivery, cyanosis, [Van Amerongen and Kreulen, 1995]. In another group of 40 premature children with low birth weight (<1500 gr), 43% of them were found to have a higher proportion of enamel hypomineralisation, when compared with control children born at due time having normal weight [Seow, 1996]. Additionally, in a group of 32 children with enamel anomalies, 83% were born prematurely and it was concluded that ameloblast function might be affected by low oxygen levels present during birth [Aine et al., 2000], an explanation that has been given before for enamel hypoplasia [Johnsen et al., 1984]. In contrast with the previous reports, two more clinical studies found no correlation between reported perinatal problems and presence of MIH [Beentjes et al., 2002; Jalevik et al., 2001b]. Many of the perinatal problems identified in the present study and in previous studies, appear to be associated with hypocalcaemia and hypoxia. In the present study, of those children with MIH and perinatal medical history, 15.6% were born prematurely with low birth weight and 18.3% were the outcome of difficult and prolonged delivery. It is well documented that early neonatal hypocalcaemia is present in approximately 30-75% of cases of preterm low birth neonates, particularly in those with respiratory distress and birth asphyxia due to complicated, prolonged or difficult, delivery [Rosli and Fanconi, 1973; Behrman and Vaughan, 1987]. 213

8 Lygidakis et al. The reason is that two-thirds of an individual s stores of calcium and phosphorus accumulate during the last trimester of pregnancy and preterm infants miss much of this mineral accretion [Tsang et al., 1976]. Additionally to hypocalcaemia, prolonged and difficult delivery has also been clearly associated with hypoxia of the newborn. In a study of 2,371 births, multivariate analysis showed that prolonged second stage labour was significantly associated with newborn hypoxia [Chandra et al., 1997], suggesting a potential association with MIH with this perinatal condition. An interesting finding of the present study was the high proportion of children born by Caesarean section; 49.7% of those with perinatal medical history and 25.5% of the total MIH group, percentages significantly higher than in the control group. All children of the present study were born during , when the proportion of Caesarean sections in the population was much less than 20% compared with more than 30% during more recent years [Roussos et al., 2003]. Although in some European countries, including Greece, this type of delivery is nowadays considered as normal, it is still an operation with various implications for the mother and the neonate. For example a recent study reported in the obstetric literature has shown that neonatal hypoxia was more frequent in infants following second-stage Cesarean section [Cebekulu and Buchmann, 2006]. Furthermore, in a large Danish study during 2007 it was clearly shown that, compared with newborns delivered vaginally, those delivered by elective Caesarean section, at around full-term had an increased risk of overall and serious respiratory illnesses, conditions often associated with hypoxia [Hansen et al., 2007]. Finally, the commonly used spinal anaesthesia for Caesarean section has a frequent complication of maternal hypotension that can be associated with severe nausea or vomiting which occasionally produces infant hypoxia [Cyna et al., 2006]. Therefore is seems that Caesarean section is associated with oxygen storage problems in the newborn and this way potentially linked with MIH in these children. Twins comprised 15.6% of those children recorded with perinatal medical history in the present study, the majority of them (22/29) being the result of in vitro fertilisation (IVF), while 8/29 of them were delivered by Caesarian section. Previous studies have shown that the postnatal health of twins, particularly those born after IVF, was worse as a result of frequent problems in the neonatal period [Koivurova et al., 2003] and the increased incidence of low birth weight and prematurity [Ludwig et al., 2006]. These problems might be a potential aetiological factor for MIH in these children, together with complicated deliveries and Caesarean section. Postnatal medical conditions. In the present study, 33.9% of MIH children recorded postnatal problems, as compared with 12.5% of the control group. Respiratory illnesses, repeated episodes of high fever, various neonatal period problems and afebrile/febrile seizures were found more frequently compared with controls. The majority of these medical conditions have been reported before in studies on smaller samples of MIH children. In a group of 21 children with MIH, 67% of them revealed various respiratory illnesses such as bronchitis, asthma, pneumonia and upper respiratory track infections [Van Amerongen and Kreulen, 1995]. In a Swedish study that also used a control group, 77 children with MIH presented with more medical problems during their first year of life, most frequently asthma, pneumonia, otitis media and upper respiratory tract infections [Jalevik et al., 2001b]. In a Dutch study 24 children with MIH were compared with 21 controls, revealing statistically different presence of otitis media, pneumonia, infections and high fever, during their first 4 years of life [Beentjes et al., 2002]. The present and the previous studies have shown potential association of various illnesses during the first year of life with MIH, such as various upper and lower respiratory illnesses of viral/bacterial origin and asthma. Hypoxic episodes, a common characteristic in some of the lower respiratory tract illnesses, can be linked as previously discussed with enamel hypomineralisation. Hypoxia is amongst the other symptoms found in many respiratory tract infections caused by certain viruses, such as RSV [McBride, 1999]. In addition to hypoxia, upper respiratory infections caused by certain bacteria, such as meningococcus, may also produce hypocalcaemia, as has been reported in a recent Cohort study where 70% of children admitted to hospital with such an infection, revealed low total and ionised calcium concentrations in the blood [Banes et al., 2000]. Hypocalcaemia is also a frequent finding in other illnesses during infancy. Viral gastroenteritis, found in the present study and usually caused by Rotavirus, is closely linked with hypocalcaemia in infants [Foldenauer et al., 1998]. It is also known that even healthy full-term babies undergo a physiological nadir in serum calcium levels by hours of age, related to the delayed response of parathyroid and calcitonin hormones in a newborn. This nadir may drop to hypocalcaemic levels in high-risk neonates including infants of diabetic mothers, preterm infants and infants with perinatal asphyxia due to complicated delivery [Aggarwal et al., 2001], correlating the various neonatal problems with MIH. In the present study, neonatal problems accounted for a high percentage of infants with MIH (13.7% in the MIH group compared with 1.3% in controls). The majority of them were neonates requiring hospitalisation and time in an incubator due to respiratory and other illnesses related with delivery problems and preterm birth. In addition to previous problems, any infection, particularly these of the respiratory tract, during the neonatal and postnatal period may produce prolonged episodes of high fever. In the present study 31/162 children with postnatal problems recorded repeated episodes of high fever, usually due to common cold/coryza. Only children with more than 5 episodes during their first year of life were recorded, as less than those are normally found in infants up to the 1st year of life [Behrman and Vaughan, 1987]. The rest of respiratory 214

9 MIH Aetiology illnesses found in the present study (otitis media, bronchitis, bronchiolitis, laryngitis, tonsillitis) may also present with high fever, as the great majority of them are complications of upper respiratory infections that have high fever as one of their clinical symptoms [Behrman and Vaughan, 1987]. This association of high fever during infancy with enamel hypomineralsation has recently been proven, as experimental studies showed that a pattern of persistent high fever influenced the process of enamel formation, producing disorientation of enamel prism and crystal-free area [Tung et al., 2006]. Repeated episodes of high fever can also be one of the main clinical symptoms of urinary tract bacterial infection in infants [Behrman and Vaughan, 1987], found in affected cases of the present study and this way explaining their association with MIH. Seizures were also more frequently reported in the MIH group when compared with controls. Febrile seizures are the result of high fever and in this way apparently linked to MIH, and shown before whilst a febrile condition might be frequently associated with infant hypocalcaemia [Scarfone et al., 2000], explaining why an association with enamel hypomineralisation might occur. Lastly, more children with MIH recorded prolonged use of medications other than antipyretics when compared with controls, indicating the inferior infant health in some of the affected children. All children with prolonged use of medication presented in the present study with respiratory or urinary tract problems requiring repeated use of corticosteroids and antibiotics. Although a recent study associated the use of amoxicillin with MIH [Ess et al., 2008], there is little literature to support the link between prolonged infant medication and MIH [Mathu-Muju and Wright, 2006]. More research is needed on this subject and for the present it appears that the defect is probably linked with the illnesses themselves [Jalevik et al., 2001a]. Combinations of medical conditions. In our study, 44 cases (11.7%) of MIH affected children revealed medical problems in more that one chronological period, the majority of them (40 cases) being during the peri- and postnatal period. This was an interesting finding indicating the difficulties of specifying precisely the aetiological factors involved in the development of MIH. In the majority of these cases (28/40) [Table 1, 2], Caesarean section was combined with neonatal problems of the newborn, as it has been discussed before [Hansen et al., 2007]. Additionally as it is clearly shown in Table 2, in the same chronological period of particularly periand postnatal, there was more than one medical problem associated with MIH, as noted by others [Jalevik and Noren, 2000]. Of 163 cases of perinatal aetiology there were 186 medical problems whilst the greatest overlap was during the postnatal period where 162 cases in that period recorded 237 medical problems, indicating these were children with poor health. This finding highlights yet again the difficulty of specific links between certain medical conditions and MIH, but clearly helps any future research to focus more on the patho-physiology of these particular conditions rather than the conditions themselves. No obvious aetiology. Finally the present study found that 12.2% of the children with MIH were not associated with any medical history. Previous studies on much smaller samples of children have found respective percentages of 9.5% [VanAmerogen and Kreulen, 1995] and 24.2% [Jalevik and Noren, 2000]. Limitations in retrospective studies concerning the possible under-estimation of self-reported data, may indeed lead to questionable results. However, considering the detailed investigation into the medical history in the methodology of the present study, and the large sample of children evaluated, it seems reasonable to conclude that this percentage might be a true figure meaning that 1 out of 10 children MIH might have an aetiology, aside from the medical problems studied herein. For example, environmental pollutants, such as dioxins have been implicated in the past with MIH [Alaluusua et al., 1996], although a more recent study has not found this [Laisi et al., 2008]. The presence of an underlying genetic predisposition that contributes to the risk of developing MIH in some cases should not be under-estimated [Brook and Smith, 1998]. In the present study there were 10 cases of children with MIH (3.3%) that belonged to 5 families. The affected children in each family were first cousins and 2 sets of them (4 children) did not reveal any medical history. Additionally from the 29 twins affected with MIH, 12 were the outcome of 6 gestations, all of them being homozygous. This group had 3 sets with no other aetiological factor apart from twinning, 2 sets delivered by Caesarean section and no other medical history, and 1 set delivered varginally but with neonatal and postnatal problems. The remaining 17 individuals had twin offspring not affected by MIH, 5 being homozygotes and 12 heterozygotes, 4/17 delivered by Caesarean section, the remaining 13 with no obvious medical history. Although the sample of twins is small and complex for statistical analysis, it seems that there might be genetic factors implicated. Correlation of the number and type of the affected teeth with the insult period. The results of the present study (Figure 1) revealed that in increasing order children with medical problems in pre-, peri- and postnatal periods present with more MIH affected teeth, whilst the greatest number of affected teeth was found in cases with combined peri-/postnatal problems. Additionally the number of affected incisors increased in the same order, more intensely than that of FPM. This distribution of teeth affected by MIH appears to correlate well with the timing of the insult. Previous studies on dental development chronology [Welbury, 1997] have shown that only molars may have started calcification prenatally; therefore medical factors acting during this period are expected to mainly produce mainly af- 215

10 Lygidakis et al. fected FPM and the least total number of affected teeth. Accordingly, during the perinatal period aetiological factors still produce a restricted total number of affected teeth, although more than those in the prenatal period as more FPM and few incisors may have initiated calcification. During the postnatal period, apart from molars, both central and lateral incisors are well into the same process; therefore factors acting during this period produce greater numbers of affected teeth. Finally, in cases of combined perinatal/postnatal problems, all teeth implicated in MIH have initiated calcification producing the greatest number of affected teeth of all types. It was interesting to find that 7 cases that were presented with 1-3 canines affected together with the typical MIH teeth, revealed combined perinatal/postnatal problems indicating the severity of the defect if medical conditions are found during these periods. Conclusion Considering the limitations of the study design it was found that: % of the children with MIH presented with potential medical aetiological factors acting during their prenatal, perinatal or postnatal period. Children with MIH reveal 68.9% more frequently medical problems than controls. 2. Postnatal conditions (33.9%) followed by perinatal (33.6%) were the most frequently found. 11.7% of the affected children revealed combined perinatal/postnatal problems. These children presented with the greatest number of affected teeth. 3. The majority of medical conditions involved may produce hypocalcaemia and hypoxia to the child or the mother. Maternal and postnatal conditions associated with prolonged pyrexia might also be implicated. 4. The number of affected teeth was associated with the timing of the possible insult; children with prenatal, perinatal and postnatal problems present more affected teeth in increasing order. 5. In 12.2% of the cases there was no reported medical history, indicating the potential implication of other systematic or genetic aetiological factors. Acknowledgements The authors wish to thank Dr Polizois Velentzas, Paediatrician MD, Health Center of Markopoulo, for his valuable assistance and Miss Eumorphia M Delicha, BSc, MSc, Biostatistics Consultant for the statistical analysis of the results. References Aggarwal R, Upadhyay M, Deorari AK, Paul VK. Hypocalcemia in the newborn. Indian J Pediatr. 2001; 68(10): Αine L, Backstrom MC, Maki R, et al. Enamel defects in primary and permanent teeth of children born prematurely. J Oral Pathol Oral Med 2000; 29: Alaluusua S, Lukinmaa PL, Vartiainen T, et al. Polychlorinated dibenzo-pdioxins and dibenzofurans via mother s milk may cause developmental defects in the child s teeth. Environ Toxicol Pharmacol 1996; 1: Alam M, Raza SJ, Sherali AR, Akhtar AS. Neonatal complications in infants born to diabetic mothers. J Coll Physicians Surg Pak. 2006; 16(3): Amerongen van WE, Kreulen CM. Cheese molars: A pilot study of the etiology of hypocalcifications in first permanent molars. ASDC J Dent Child. 1995: Banes PB, Thomson AP, Fraser WD, Hart CA. Hypocalcaemia in severe meningococcal infections. Arch Dis Child. 2000; 83(6): Beentjes VEVM, Weerheijm GHJ, Grohen HJ Factors involved in the aetiology of molar-incisor hypomineralisation (MIH). Eur J of Paediatr Dent. 2002;3(1) Behrman RE., Vaughan VC. Nelson s Textbook of Pediatrics. 13th Edition. W.B.Saunders Philadelphia , , Brook AH., Smith JM. Aetiology of developmental defects of enamel: a prevalence and family study in East London, UK. Connect Tissue Res. 1998; 39: Cebekulu L, Buchmann EJ. Complications associated with cesarean section in the second stage of labor. Int J Gynaecol Obstet. 2006; 95(2): Chandra S, Ramji S, Thirupuram S. Perinatal asphyxia: multivariate analysis of risk factors in hospital births. Indian Pediatrics, 1997; 34(3): Cyna AM, Andrew M, Emmett RS, Middleton P, Simmons SW. Techniques for preventing hypotension during spinal anaesthesia for caesarean section. Cochrane Database Syst Rev. 2006;18:(4):CD Review. Ess A, Laisi S, Sahlberg C, Lukinmaa P-L, Alaluusua S. Early Use of Amoxicillin May Cause Molar-Incisor-Hypomineralisation (MIH) Europ Archs Paediatr Dent 2008;9: Fearne J., Anderson P., Davis GR. 3D X-ray microscopic study of the extent of variations in enamel density in first permanent molars with idiopathic enamel hypomineralisation. Br Dent J. 2004;196(10):634-8 Foldenauer A, Vossbeck S, Pohlandt F. Neonatal hypocalcaemia associated with rotavirus diarrhoea. Eur J Pediatr. 1998;157(10): Hansen AK, Wisborg K, Uldbjerg N, Henriksen TB. Risk of respiratory morbidity in term infants delivered by elective caesarean section: cohort study. Br Med J. 2008;336(7635):85-7. Jalevik B, Noren JG. Enamel hypomineralisation of permanent first molars: a morphological study and survey of possible aetiological factors. In J Paediatr Dent 2000;10: Jalevik B., Obelius H., Dietz W., Noren J. Secondary ion mass spectrometry and X-ray microanalysis of hypomineralised enamel in human permanent first molars. Arch Oral Biol. 2001;46(3): Jalevik B, Noren JG, Klingberg G, Barregard L, Etiologic factors influencing the prevalence of demarcated opacities in permanent first molars in a group of Swedish children. Eur J Oral Sci 2001;109: Jalevik B., Klingberg GA. Dental treatment, dental fear and behaviour management problems in children with severe enamel hypomineralisation of their permanent first molars. Int J Paediatr Dent. 2002;12(1): Jasulaityte L, Veerkamp J.S, Weerheijm KL. Molar-incisor-hypomineralisation: review and prevalence data from a study of primary school children in Kaunas (Lithuania). Eur Arch Paediatr Dent. 2007; 8(2): Johnsen D, Krejci C, Hack M, Fanaroff A. Distribution of enamel defects and the association with respiratory distress in very low birthweight infants. J Dent Res. 1984; 63(1): Koivurova S, Hartikainen AL, Sovio U, et al. Growth, psychomotor development and morbidity up to 3 years of age in children born after IVF. Hum Reprod. 2003; 18(11): Kreshover SJ, Clough OW. Prenatal influences on tooth development. II. Artificially induced fever in rats. J Dent Res. 1953; 32(4): Laisi S, Kiviranta H, Lukinmaa P-L, Vartiainen, T, Alaluusua, S. Exposure to PCDD/Fs and PCBs at prevailing levels is not associated with Molar-Incisor-Hypomineralisation. Europ Archs Paediatr Dent 2008:9: Ludwig AK, Sutcliffe AG, Diedrich K, Ludwig M. Post-neonatal health and development of children born after assisted reproduction: a systematic review of controlled studies. Eur J Obstet Gynecol Reprod Biol. 2006;127(1):3-25. Lygidakis NA, Chaliasou A, Siounas G. Evaluation of composite restorations in hypomineralised permanent molars: a four year clinical study. Eur J Paediatr Dent 2003;3: Lygidakis NA, Dimou G, Marinou D, Gouva G. Aetiology of Molar-Incisor Hypomineralisation. A retrospective study of 151 children with the defect (abstract). 7th Congress of the European Academy of Paediatric Dentistry; Barcelona, Spain Lygidakis NA, Dimou G, Briseniou E. Molar-incisor hypomineralisation (MIH). Retrospective clinical study in Greek children. I. Prevalence and defect characteristics.europ Archs Paediatr Dent. 2008;9: