2.1 Eligibility criteria

When selecting published articles, we applied inclusion criteria according to the PECO method, as described below.

- Population: Children in good health, from birth to 6 years-of-age, who had a primary or mixed dentition.

- Exposure: Studies focusing on the presence of NNSHs, such as pacifier use or finger sucking and the impact of NSHs, such as breastfeeding and bottle-feeding.

- Comparison: Comparisons of the prevalence of NNSHs, including pacifier use or finger sucking, with the occurrence of malocclusion. Studies involving the potential impact of NSHs, such as breastfeeding and bottle-feeding, on the development of malocclusion.

- Outcome: Studies focusing on the frequency, intensity, duration and time of cessation of each habit and its relationship with the development of malocclusions in the three planes of space (sagittal, vertical, transversal) was evaluated.

- Studies: Observational cross-sectional studies, randomized and non-randomized cohorts, were included without language or publication time restrictions.

Articles were excluded if they involved children with systemic diseases, craniofacial syndromes or anomalies; we also excluded systematic reviews and meta-analyses.

2.2 Information sources

To conduct this systematic review, we followed the PRISMA protocol: Preferred Reporting Items Protocol for Collecting Systematic Reviews and Meta-Analyses [17, 18].

2.3 Search strategy

A literature search was performed in the following bibliographic databases: PubMed (Medline), The Cochrane Library (CDSR, CENTRAL and DARE), Embase and Scopus until April 2023. Grey literature was also searched.

The search strategy used the following keywords: (malocclusion* OR malocclusion (Mesh)) AND (Breastfed* OR Breastfeeding (Mesh)) AND (Habit* (Mesh) OR Finger sucking (Mesh)) AND (Pacifier (Mesh) OR Orthodon* (Mesh) OR Myofunctional therapy (Mesh)). The search strategy can be found in Table 1: Search strategy.

2.4 Data collection process

Once the bibliographic search had been performed, the selection of studies was carried out in two phases. In the first phase, two of the authors (CS and LA) evaluated the titles and abstracts of the articles independently and selected those that met the inclusion criteria. In the second phase, full texts and articles that did not contain sufficient information in the title or abstract were reviewed to decide whether they should be included. In case of disagreement (in both phases), a third author (LT) was consulted to reach consensus.

2.5 Data items

The type of study, the sample size, the method used to acquire information on the patients and the comparison between the different types of habits and malocclusions resulting from each study are described in Table 2.

2.6 Risk of bias assessment: quality of evidence

To assess the methodological quality/risk of bias of the selected articles, CS and LA independently used the checklist for cross-sectional studies and the checklist for cohort trials of the JBI Systematic Review [19, 20]. The checklist for cross-sectional studies consisted of eight questions while that for cohort studies consisted of 11 questions. Each question was answered with a “yes” or “no”. To evaluate the different studies, we allocated a score of 1 for the answer: “yes” and 0 for “no” and added the scores for comparison. For cross-sectional studies, a total of less than 4 points was considered as high risk, 4 points as moderate bias, and more than 4 points as low risk. For cohort studies, a total of less than 6 points was considered as high risk, 6 points as moderate bias, and more than 6 points as low risk.

2.7 Risk of bias across studies

We used the GRADES system [21] to assess the overall quality of the included articles. This method includes factors that rate the score as risk of bias, imprecision, inconsistency, indirectness and publication bias and other factors that increase the level as a magnitude of effect, dose-response gradient while taking all confounding factors into account.

2.8 Effect measures

The main outcomes assessed were the relationship between nutritive and non-nutritive sucking habits and the development of malocclusions, the influence of NNSHs and other variables, such as the socioeconomic status of the children and the presence of oral respiration, on the apparition of sucking habits. These outcomes were presented in Table 3.

3.1 Characteristics of the studies

Fourteen cross-sectional studies [22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35] and seven cohorts [36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47] were identified. These studies were conducted in Brazil [23, 24, 27, 30, 33, 35, 39, 40, 41, 42], China [22, 25, 32], USA [36, 38], India [26], Slovenia [28], Senegal [29], Puerto Rico [31] and Spain [34].

Children with primary dentition were investigated. According to the included studies, the age range of the participants varied between two and five years [20, 31, 34, 41], three and six years [24, 25, 27, 30, 34], four and five years [23, 36] and between five and six years [28, 29]. In two studies, researchers followed children from birth [40, 41], in one article from 6 months [30] and in another investigation from 12 months [37].

With regards to NSH and NNSH, one article assessed frequency [36], nine assessed the duration of NNSH and NSH [22, 23, 24, 26, 28, 34, 37, 39, 40] and in the remaining 11 studies, both factors were investigated [24, 25, 27, 29, 30, 31, 32, 33, 37, 39, 40]. Fourteen articles also assessed other variables, including socioeconomic level [22, 25, 30, 33, 35, 37, 38, 39, 40] or the presence of oral respiration/the size of the nasal airways [23, 24, 28].

3.2 Risk of bias assessment

We identified average methodological and clinical homogeneity among the included studies. According to the JBI Critical Appraisal Checklist [19, 20], three cohort studies [39, 40, 41] and seven clinical trials [22, 24, 25, 34, 35, 36] were of good quality. The principal failures of the selected articles were (1) a lack of measurement of exposure to sucking habits in a valid and reliable manner, (2) a lack of identification of confounding factors (sociodemographic factors, social status, presence of other diseases), and (3) a lack of information on the follow-up time and whether it was long enough for malocclusions to occur. These results were illustrated in Tables 4 and 5.

3.3 Results of bias across studies

The GRADE system [21] was performed to assess the overall quality of the included studies. All transversal studies qualified as a very serious risk of bias and all analytical studies as a serious risk of bias. All articles provided confidence intervals for the effects of treatment and were scored as no serious risk for imprecision. We did not attribute negative punctuation for inconsistency because there was no heterogeneity in the results of the selected studies. Investigations that had no statistically significant results and a low sample size were considered to feature publication bias.

Five studies [27, 31, 33, 34, 36] were considered to have a moderate quality of evidence and eleven studies [22, 23, 24, 25, 28, 29, 30, 37, 42] had a low quality of evidence. The remaining articles [26, 34, 38, 40, 41] had a low quality of evidence. Quality assessment is summarized in Table 6.

3.4 Results specific analyses

4.1 Breast-feeding vs. bottle-feeding

Exclusive breastfeeding between zero and six months is advised by the World Health Organization (WHO) as a public health policy because it reduces the risk of aero-digestive infections. Breastfeeding is defined as one of the foundations of health promotion and the prevention of many diseases and is one of the pillars of correct maxillofacial growth because it favors proper lip sealing, jaw function and the position of the tongue against the palate [43]. Indeed, breastfeeding forces the child to actively extract milk from the mother’s breast through the synergistic action of the tongue and facial muscles [7, 44, 45, 46, 47, 48]. In opposition, bottle feeding requires less effort to drain the milk, so it does not stimulate the functional matrix and favors the development of malocclusions such as a posterior crossbite, anterior open bite, increased protrusion and a class II molar and canine relationship. Similarly, the bottle nipple is usually made of a less flexible material, which can press inside the oral cavity and lead to improper alignment of the teeth and a narrow palate [43]. Furthermore, according to several authors [6, 22, 49, 50, 51, 52, 53], children with a bottle favor a strong tendency to develop a pacifier habit.

4.2 NNSHs

In some developed countries, the use of pacifiers is so culturally established that the prevalence in young children at 12 months can be as high as 42.5%. Pacifier sucking is the most common NNSH and has received considerable research attention for many years [54]. The use of pacifiers can be a dangerous factor in the development of malocclusions. Indeed, the three factors that are associated with the relationship of malocclusion and pacifier use are intensity, frequency and duration of pacifier use. The recommended age for stopping the pacifier sucking habit is two years, and this is considered a prolonged habit if continued pacifier use lasts until four years of age or more [54]. In most cases, this habit stops spontaneously at around five years-of-age. However, in a minority of cases, the habit may continue for several more years, even into adolescence and beyond [55, 56]. This habit can induce deformity of the dental occlusion, and this deformity is produced in direct proportion to the parameters of the presence of the habit [55, 57, 58, 59, 60]. The longer the duration of the force (50% of the time), the greater its impact on teeth. Thus, to produce significant variability in tooth position, durations should be measured in hours per day. There is clear evidence that sucking habits in children are strongly correlated with malocclusions [5].

4.3 Limitations

Arguably, having followed the PRISMA method, there were no limitations in the methodology of this systematic review. Methodological and mean clinical homogeneity was detected among the studies. According to the JBI Critical Appraisal Checklist [19, 20], in the cohort studies, three studies [39, 41] were found to be of good quality. In the clinical trial studies, there were seven [22, 24, 25, 26, 34, 35, 36] articles of good quality. The main failures of the selected articles were as follows. First, the lack of exposure measurement in a valid and reliable way: questionnaires were sent to parents and referred to a child’s nutritive and NNSHs and a clinical examination of the child or on its cast model were performed. The included studies did not specify the reference test used. In some articles, the authors mentioned that in the questionnaires, parents and surrogate informants could be mistaken about feeding habits or answer the option that they considered correct to please the interviewer. Secondly, the lack of identification of confounding factors (sociodemographic factors, social status and the presence of other diseases). Third, the lack of information on the follow-up time of the population and whether this time was sufficient for the results to be produced: The follow-up of the dental condition is important as it can provide information associated with the age of onset of the malocclusion or whether the malocclusion was self-corrected at the time of the final examination. In addition, it is important to mention the large heterogeneity between the sample sizes of each study. The sample size varied from 60 children to 1377 children. This hindered the comparability of the results. Similarly, there was much disparity in the age of the study participants.