Ankyloglossia is defined as an embryological remnant of tissue remnant between the undersurface of the tongue and the floor of the mouth that restricts normal tongue movement.1

It has been linked to latch-on difficulties and maternal pain, which are 2 of the main causes of weaning in our environment.1 Therefore, it could be the source of breastfeeding problems, a circumstance that has not been proven mainly due to the methodological limitations of the published studies.

The Hazelbaker score assessment2 is the tool commonly used for diagnosis. Its main drawback is that it has low inter-rater reliability, especially for the function items.3 The Coryllos classification allows the distinction of different types of ankyloglossia based on the insertion point of the frenulum.

There are no standardised objective clinical criteria for the diagnosis of ankyloglossia, so the prevalence rates observed in the literature are varied (4–10%), with varying study designs that complicate or prevent their comparison.1,4,5 In Asturias there is a reported prevalence of 2% among schoolchildren.6 But the prevalence among newborns at the national and regional level for the rest of Spain is unknown. Our objective was to learn the current prevalence of ankyloglossia in Asturias.

We examined a total of 677 newborns, of whom 82 met the diagnostic criteria for ankyloglossia. The total prevalence was 12.11% (95% CI: 9.58–14.64). Depending on the hospital, the prevalence ranged between 3.4% and 50% (Fig. 2).

Figure 2.

Prevalence of ankyloglossia in the different health districts of the Principality of Asturias. Prevalences expressed as percentages and 95% confidence intervals in parentheses.

(0.12MB).

Of the 82 patients with a short frenulum, 62% were male. Normal delivery was reported in 57% of cases, instrumental delivery in 37% and caesarean in the remaining 6%. There was a family history of ankyloglossia in 1 out of 4 newborns diagnosed with the condition, which in most cases (86%) concerned first-degree relatives: parents or siblings.

According to the Coryllos anatomical classification, 10 patients (12%) had a type I frenulum, 43 patients (54%) a type II, 19 patients (24%) a type III, and 8 patients (10%) a type IV. The distribution by hospital is shown in Table 2.

The 12% prevalence is the highest reported in the literature up to now, between 2 and 3 times greater than expected.1,4,5,8,9 This circumstance reminds us of the importance of examining the oral cavity in newborns, with special emphasis on the assessment for ankyloglossia.

The Coryllos classification allowed the identification by means of palpation of type III and IV frenula that may be missed in a visual inspection.7 On the other hand, we believe that functional criteria must be used in the diagnosis. If we were to use anatomical criteria alone, it would seem, erroneously, as if many newborns had ankyloglossia.

We are aware that it is difficult to assess the range of movement of the tongue correctly, and that there may be inter-rater differences. For these reasons, we used the first 3 function items of Hazelbaker as diagnostic criteria, as they have shown substantial inter-rater agreement (kappa values 0.6 and 0.7) and are easier to assess.3 The cut-off point used in the study (score ≤4), proposed by Amir,3 showed a strong correlation with Hazelbaker's definition of significant ankyloglossia: score <11 out of 14, if all the function items are analysed.2

The observed prevalence was greater than expected. This difference may be due to the digital examination that allowed the identification of cases that could not be detected by visual inspection (34% of types III–IV). The proportion of this type of ankyloglossia was similar in 5 hospitals; in one other hospital up to 65% of frenula fell into this category. This fact may suggest a classification bias resulting from subjectivity and the difficulty involved in establishing anatomical limits. Nevertheless, the results from the 3 hospitals with the largest populations in the study (85% of the sample), were pretty homogeneous, with a prevalence of 10%.

The short, type IV lingual frenulum may be the most difficult to diagnose. There are no data on its prevalence, as this classification is not specified in the published studies. Recently, O’Callahan10 described the performance of 131 frenotomies in newborns younger than one month with breastfeeding problems, with type IV being the most frequently treated (40%).

In our series, 10% of the frenula were classified as type IV, although most of them were diagnosed at a single hospital. Without the participation of this centre, the prevalence of this type of ankyloglossia would be 0.16% and it would account for 1% of the lingual frenula. We do not know whether the criteria used results in underdiagnosis of this type of frenula, so that its prevalence may be even higher than the observed 12%. Still, this once again underscores a subjective factor dependent on the assessor.

There was a family history of ankyloglossia in up to 25% of patients, most concerning first-degree relatives. Genetics play an important role in this condition. Ankyloglossia can be seen in isolation or associated with syndromes such as Ehlers–Danlos, Beckwith–Wiedemann, Simosa, and orofaciodigital syndrome. The higher prevalence in males, along with genetic studies, suggest a pattern of X-linked inheritance.11,12

One of the limitations of the study was interobserver variability. The performance of a practical workshop and an early pilot study allowed us to unify the criteria among researchers. However, these did not suffice, as can be seen in the prevalence disparity of one of the centres. The use of functional criteria lessened this limitation and helped objectivise the subjectivity in diagnosing ankyloglossia, especially for types III and IV.

To our knowledge, this is the first study on the prevalence of ankyloglossia in newborns in Spain. García Pola et al.6 found a prevalence of 2% in our region, but in 6-year-old children. In contrast to other publications, our study employed objective diagnostic criteria and the anatomical classification. The only 2 centres in our region that did not participate account for less than 8% of all births. Furthermore, their socioeconomic factors are equivalent to those of other hospitals that were included, and thus our data are representative of the prevalence of ankyloglossia in Asturias.

To conclude, we found a high prevalence of ankyloglossia. Newborn and infant check-ups should include a routine examination of the oral cavity, focusing on the presence of ankyloglossia, to ensure detection of this condition. Further research is needed in order to unify diagnostic criteria and to elucidate the impact of ankyloglossia on breastfeeding and other problems that may arise later in life.

This study was funded by the Ernesto Sánchez Villares Foundation on the occasion of the XV Convocatoria de Ayuda a la Investigación Clínica y Epidemiológica of year 2012 (project 15/2012).

The authors have no conflicts of interest to declare.