Speech impediment is not only a physical effect of orthodontic treatment but also a socio-psychological impact. In this systematic review, speech impairment was produced irrespective of the type of orthodontic appliances. Nevertheless, there were some differences among the three appliances, including the severity of speech impediments, duration, sound errors, speech rate, etc.
The severity of speech sound disorders
The LA had the least effect on pronunciation among the three appliances. Several previous studies stated that LA had no impact on speech, agreeing with Fraundorf  and Melo et al. . However, the majority of selected studies agreed that LA caused a slight speech impediment during the initial stage, which returned to baseline after one month. In contrast, previous studies stated that LA patients required 2–3 weeks as an adaptation period. Compared with LA, patients with LI experienced more speech impediments at the beginning of treatment. Most studies reviewed reported that LI patients required at least three months to adapt to the change. Only one study compared the LI and OA; both caused equal but slight impairments in speech, while OA caused even greater difficulty in the production of /s/ and /z/ sounds . Nevertheless, this study did not use objective assessment, and the certainty of the evidence was low. What is certain is that OA had more limitations and difficulties in speech than LA. Besides, two studies indicated that patients with aligners spoke more slowly to attempt to better articulate [8, 21]. To date, no study has assessed whether the OA brands have had an impact on speech sound production.
Interestingly enough, LA brands were associated with the severity of speech impediment due to the different designs of brackets. In the study by Khattab et al., the sharp hooks caused more irritation to the tongue compared with the round hooks. Speech performance was also affected by bracket positioning techniques. The Bonding with Equalized Specific Thickness (BEST) positioning system was significantly poorer than the Transfer Optimized Positioning (TOP) procedure. Additionally, the customized brackets were considerably better than the prefabricated brackets. On the whole, the thinner the lingual appliances, the fewer the speech impediments. However, there are no studies comparing the speech effects caused by the self-ligating brackets and conventional brackets.
The pathology speech sound disorders
While the type of articulatory error differed between appliances, the /s/ sound was most commonly affected. The production of these phonemes is determined in part by the lip, incisor, and tongue contact. Leavy et al. attributed speech sound disorders to placement errors. The anterior lingual protrusion, lingual retraction, and lateral lingual protrusion would disrupt the production of lingual-alveolar consonants, such as /s, sh, l/, while the lip excursion would affect the production of labiodental sounds, such as /f/. Understandably, the LI caused tongue space restriction and incisor morphological change, affecting the lingual-alveolar sounds the most. According to Paley et al., speech sound disorders in labial patients are a result of lingual protrusion. Meanwhile, the LA affected the lip position and resulted in articulatory errors with labiodental sounds. Despite its thinness, OA produces both lingual-alveolar sounds and labiodental sounds since it covers the labial-lingual surfaces of the teeth at the same time. Aside from the aforementioned factors, lingual auxiliaries such as TPAs and Nance appliances had an effect on sound production.
In addition to orthodontic appliances, other factors should be taken into consideration. It is apparent that there is a relationship between malocclusion and speech defects [43–45]. A study by Ahmed et al. showed that Class II patients had difficulties with /p, b, m, s/ sounds, while Class III patients had difficulties with /s, z, f, v/ sounds. Furthermore, Paley et al. revealed that the severity of malocclusion would affect the time needed for speech adaptation. However, only seven included studies[12, 13, 19, 20, 26, 29, 31] in the present systematic review limited the types of malocclusions. Different languages also affect speech impairment, which could explain why the results regarding speech defects differ between trials [33, 46]. Age could also affect the speech performance. According to Hohoff’s study, age-related influence can be found only after 57 years old , while the age for Dahan’s study is 80 years old .
Recommendations for clinical practice
Consider the speech impact of an appliance before choosing one for a particular patient . Before treatment, orthodontists should ensure patients are fully informed about the speech impacts caused by different orthodontic appliances, including the severity, the duration, and the type of sound errors. For the fixed appliances, the upper arch and lower arch bonded at two appointments could enhance patient comfort and reduce adverse effects on speech . As for LI, the thinnest is the most effectives. Choosing customized brackets, TOP techniques, and rounded hooks are all excellent choices. Besides, the use of orthodontic wax can reduce the pain in the tongue and mitigate speech problems. The aligner is the only appliance that can be removed by a patient’s will . To avoid poor adherence, it should not be considered for those who take speech impediment very seriously. Sometimes, appropriate psychological intervention should be taken to boost patients' confidence and help them through the early stages of orthodontic treatment.
Limitations and Prospect
The present systematic review updated the speech effects encountered during orthodontics based on the existing evidence. Besides, studies published before 2000 were excluded to reduce bias because of the development of orthodontic appliances and assessment methods. However, there are also some limitations. First, no studies have been conducted to evaluate the speech effects caused by these three appliances simultaneously. Second, seven selected studies[20, 27, 28, 35–38] only used subjective analysis, which reduced the strength of evidence. Due to substantial heterogeneity across studies meta-analysis was not possible.
In conclusion, long-term, large sample size, well-designed, and well-controlled studies are required. Furthermore, future studies should: 1. consider the potential effects of malocclusions and auxiliaries on speech; 2. use more objective acoustic analysis to reduce the inter-individual variability and subjective judgment of speech; and 3. explore the effectiveness of different solutions in improving speech impediments.