Introduction

Recurrent aphthous stomatitis (RAS) or recurrent aphthous ulcers is a very common painful disease of oral mucosa, affecting up to 25% of the general population [1,2,3,4]. It is characterized by recurrent single or multiple ulcers affecting mainly the non-keratinized mucosa [5]. On the basis of the clinical presentation, RAS is classified as minor, major, and herpetifiorm [1,2,3]. The minor form is the commonest type accounting for around 85% of all RAS cases; it presents as a small well-circumscribed ulcer of less than 1 cm in diameter that heals spontaneously within 7–10 days [2, 4]. Although self-limiting, RAS is quite annoying that can interfere with the daily functions such as eating and speaking, thus impairing the patient’s quality of life [5, 6]. The treatment of RAS is quite challenging, and as yet, there is no definitive cure; the primary goal of RAS management is to reduce pain, promote healing, and reduce the frequency of episodes [4,5,6,7].

Despite the extensive research on RAS, the exact etiopathogenesis is still unclear [8]. However, some local and systemic predisposing factors have been suggested. These include genetic predisposition, local trauma, stress, nutritional deficiencies, immunological disturbances, systemic diseases, and vitamin deficiency [5, 8,9,10,11].

Nutritional deficiencies of iron and certain vitamins such as B12, B2, D, and B9 (folic acid) have been implicated in pathogenesis of RAS [9, 10]. Recent data suggest that even the trace elements such as zinc may have a role in pathogenesis of RAS [12,13,14]. In fact, zinc is a vital micronutrient that plays pivotal roles in cell growth/reproduction, collagen synthesis, and wound healing [15]. Additionally, zinc has been reported to have an essential role in regulating the immune system, including innate and adaptive immune responses [15]. Many studies have linked zinc deficiency with several disorders such as rheumatoid arthritis, vitiligo, ectopic dermatitis, and multiple sclerosis [15,16,17,18]. In this regard, several observational studies have investigated the potential association between zinc levels and RAS, and conflicting results were reported [12, 13, 19,20,21,22]. Several studies found a significant association between zinc deficiency and the development of RAS [12,13,14, 20, 23], whereas a few others did not find any association [19, 21, 22]. It seems that there is some controversy regarding the potential role of zinc in pathogenesis of RAS. Therefore, the present meta-analysis sought to assess the available evidence on the potential association between zinc level and the occurrence of RAS.

Materials and methods

Study question

The current meta-analysis followed strictly the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [24]. The study was registered in the PROSPERO database (ID=CRD42020198559). The focused research question was “Is there any association between zinc deficiency and the occurrence of RAS?”

Eligibility criteria

This current systematic review included all observational studies that assessed the potential role of zinc deficiency in developing of RAS, and fulfilled the following criteria: (1) comprised RAS patients and control subjects, (2) RAS patients are systemically healthy ≥ 15 years old, (3) the main outcome measure reported quantitative zinc level (mean ± SD), and (4) the language was either English or Chinese. The following studies were excluded: (1) those without control group, (2) those without adequate quantitative data (mean ± SD), (3) studies that assessed other than serum/plasma zinc levels (e.g., hair), (4) case series, (5) case reports, (6) in vitro studies, and (7) review articles.

Search strategy

PubMed, Scopus, Web of Science, Google Scholar, and China National Knowledge Infrastructure (CNKI) databases/search engines were searched up to August 2020. The following key words were used: (Zinc OR “Zinc deficiency” OR “Zinc level” OR “Zinc levels” OR “serum Zinc” OR “plasma Zinc”) AND (“recurrent aphthous stomatitis” OR “aphthous stomatitis” OR “recurrent aphthous ulcers” OR “recurrent aphthous ulcer” OR “aphthous ulcers” OR “aphthous ulcer” OR “canker sore”). The Chinese translations of these key words were used for the search in the CNKI database. .

Data extraction

The eligible studies were processed for data extraction. Two authors (AS and AA) independently extracted the necessary information from the English studies, and other two authors with good experience the Chinese language (HA and AA) independently extracted the necessary information from the Chinese studies. The following data were extracted from each study: authors and year of the article, country, study design, number of RAS patients and control subjects, gender, age, clinical type of RAS, diagnostic criteria of RAS, assessment methods, and serum/plasma zinc levels.

Quality assessment

The quality (risk of bias) of the included studies was assessed using Newcastle-Ottawa Scale (NOS) [25], which is based on three major components: selection of the study groups (0–4 stars), comparability of cases and controls by controlling for relevant factors (0–2 stars), and exposure (0–3 stars). The overall quality of a given study is rated as either high quality, 7 stars or more; moderate quality, 4–6 stars; or poor quality, 0–3 stars. Two independent investigators (AS and AA) assessed the risk of bias of the English studies, and another group of tow investigators (HA and AA) assessed the risk of bias of the Chinese studies; disagreement between the two investigators, if any, was solved via discussion and/or by consulting a third reviewer (EH) until consensus is reached.

Statistical analyses

The most widely used unit of measurement of serum zinc levels is μg/dL. Means and standard deviations of serum zinc levels which presented otherwise were converted to μg/dL in the following website: http://unitslab.com/node/142. Serum zinc levels, which were presented in unusual unit of measurements or, upon conversion, the values were extraordinary, were excluded. Data were analyzed using Comprehensive Meta-Analysis software version 2.2.046 (Biostat, Englewood, NJ, USA). The weighted difference in means between the two groups along with the 95% confidence intervals (CI) was calculated. Heterogeneity was identified using chi-squared test and the I2 statistics. “The random effects model” was used in case the heterogeneity was significant (I2 > 50%); otherwise, “the fixed effects model” was applied. A P value less than 0.05 was considered statistically significant.

Trial sequential analysis

The main aim of rrial sequential analysis (TSA) is to control for increased random error that may arise in meta-analyses. Simply, TSA generates monitoring boundaries and estimates an optimal sample size in cumulative meta-analyses, creating thresholds for declaring significance [26]. Every time a new study is included in a meta-analysis, the significance is tested, along with estimating the confidence intervals. In line with that, the cumulative Z curves that represent the findings of the data are created. When the Z curves surpass the monitoring boundary, the level of evidence is said to be sufficient and further trials are deemed futile. Hence, once one of the boundaries or the optimal sample size in cumulative meta-analyses is reached, the conclusion is said to be conclusive and confirming, and no further trials are needed. Otherwise, more trials are needed to clarify the conclusion [27]. In our study, we used two-sided trial sequential monitoring boundary type, and calculated the required information size (RIS) based on α = 0.05 and β = 0.20. The mean differences and effect sizes were calculated from the included studies. The software TSA version 0.9.5.10 beta was used for the analysis (Copenhagen Trial Unit, Centre for Clinical Intervention Research, www.ctu.dk/tsa).

Publication bias

Publication bias was assessed using funnel plot and Egger’s test. The funnel plot was conducted using Review Manager 5.3 software (RevMan 5.3, The Cochrane Collaboration, Copenhagen, Denmark), while the Egger’s test was conducted using Stata software programs for Windows (Stata Corporation, version 15.1, College Station, TX, USA).

Results

Study selection

The search strategy is illustrated in Fig. 1. A total of 708 studies were identified; 120 studies were duplicates and were thus removed. The titles and abstracts of the remaining 588 studies were screened. Of these, 506 were found to be irrelevant and were thus excluded. The full texts of the remaining 82 articles were screened, and 63 studies were excluded due to various reasons: lack of control group, missing data, studies conducted on children, studies published in languages other than English or Chinese, use of unclear unit of measurement, case reports, and reviews. Eventually, 19 case-control studies [12,13,14, 19,20,21,22,23, 28,29,30,31,32,33,34,35,36,37,38] were included in the current meta-analysis.

Fig. 1
figure 1

Flow chart of methodology according to PRISMA guidelines

Full size image

General characteristics of the included studies

Table 1 presents the characteristics of the included studies. A total of 19 case-control studies (in English language, n = 7; in Chinese, n = 12) were included in the present meta-analysis [12,13,14, 19,20,21,22,23, 28,29,30,31,32,33,34,35,36,37,38]. The sample comprised 2044 subjects (1079 RAS cases and 965 controls). Thirteen studies [12, 23, 28,29,30,31,32,33,34,35,36,37,38] were conducted in China, 3 in Turkey [13, 14, 22], one in Iran [20], one in India [19], and one in Poland [21]. The main age of the included subjects ranged from 15 to 68 years, with around 40% (n = 812) of the subjects being females. All studies assessed serum zinc levels. The assay method of zinc level varied among studies, although the majority used atomic absorption spectrometer method. All studies included healthy RAS patients and control individuals with no history of systemic diseases.

Table 1 General characteristics of the included studies
Full size table

Meta-analysis results

All studies (n = 19) were included in the meta-analysis. Fifteen of the included studies [12,13,14, 20, 23, 28, 30,31,32,33,34,35,36,37,38] found a significant association between low zinc levels and RAS, whereas the remaining 4 studies [19, 21, 22, 29] did not find such association (Table 1).

The results of the pooled 19 studies revealed that zinc level was significantly lower in RAS patients than in healthy controls (weighted difference in means = − 21.092, 95% CI − 26.695 to − 15.490, I2 = 95.375%, P < 0.001; Fig. 2).

Fig. 2
figure 2

Forest plot of the Meta-analysis

Full size image

Upon subgroup analyses by geographical distribution of the samples (Chinese vs. others), the results remained significant, although the relationship was more pronounced among Chinese populations (weighted difference in means = − 24.279, 95% CI − 30.828 to − 17.731, I2 = 95.733%, P < 0.001) compared with other nationalities (weighted difference in means = − 13.768, 95% CI − 25.824 to − 1.705, I2 = 94.395%, P = 0.025) (Supplementary figure 1).

Trial sequential analysis

TSA showed that the cumulative Z curves crossed the conventional boundary and the trial sequential monitoring boundary, and surpassed the required information size (n = 310) as well (Fig. 3). Thus, the conclusion is reliable and confirming, and further trials are not needed.

Fig. 3
figure 3

Trial sequential analysis

Full size image

Publication bias

On the basis of visualization of Funnel plot (Supplementary Figure 2), there was no evidence of asymmetrical pattern with the included data, indicating no publication bias. This was further confirmed with the quantitative analysis using Egger’s test (P = 0.982).

Quality of the included studies

The results of the quality assessment of the included studies are presented in Table 2. Two [21, 33] of the included studies were of high quality, 14 studies [12,13,14, 20, 22, 23, 28, 30, 31, 34,35,36,37,38] were of moderate quality, and the other three studies [19, 29, 32] were of low quality (Table 2).

Table 2 Results of the quality assessment using NOS for case-control studies
Full size table

Discussion

RAS is a very common oral lesion of unclear etiology, what makes its therapy a challenge [4, 5, 39]. Many studies have investigated the potential association between zinc levels and RAS, and reported conflicting results [12, 13, 20]. Hence, the present meta-analysis was performed to answer the focused question: Is there any association between zinc levels and the development of RAS? The results of the pooled 19 studies showed that patients with RAS have significantly lower zinc levels than healthy controls. Additionally, upon subgroup analysis by geographical distribution of the sample, the results remained significant, although the relationship was more pronounced among Chinese compared with other countries. The results of the current meta-analysis were further confirmed by TSA, which showed that the current sample size is surpassing the required one to draw a definitive conclusion (Fig. 3). This was true even when TSA was conducted for subgroups (based on geographical differences) separately, although this was more pronounced for Chinese studies. Nevertheless, conducting well-designed, large-scale studies targeting different populations is still encouraged [26].

The results of the present meta-analysis corroborate previous systematic reviews that reported a significant association between zinc deficiency and several autoimmune diseases such as ectopic dermatitis, rheumatoid arthritis, multiple sclerosis, and vitiligo [16,17,18]. Zinc is the second most abundant trace element in the human body [40]; it has a vital role in regulating the innate and adaptive immune responses, cell cycle progression, and differentiation [15]. Although the precise etiopathogenesis of RAS is still uncertain, compelling evidence indicates that immunological dysfunction/deregulation is implicated [8]. Hence, considering the pivotal role of zinc in the immune system and the postulated immunological background of RAS, the potential role of zinc in the pathogenesis of RAS can be conceptualized.

In meta-analysis, it is known that the quality of the included studies is key for determining the level of evidence. For this purpose, the quality of all included studies was thoroughly scrutinized using NOS, a well-known quality appraisal tool for observational studies [25]. The NOS-based results revealed that apart from two studies, which showed high quality, most of the included studies showed some methodological flaws. Such methodological flaws may limit the results of the current study.

The present meta-analysis advocates a significant association between zinc deficiency and the risk of RAS. However, there are some limitations that should be considered when interpreting results of the present meta-analysis. The key limitation is that the evidence obtained is solely based on observational studies, and so the potential confounders might have not been controlled. Another important limitation is the marked heterogeneity among the included studies with respect to settings, methods of zinc measurements, type of RAS, gender, and age of participants. Nevertheless, despite these limitations, the study has several strengths that should be acknowledged. First, this is the first study that summarized the evidence on the association between zinc and RAS. Second, this meta-analysis included a relatively huge sample size, and included all potential studies in the English and Chinese literature. Finally, the results of the meta-analysis were further confirmed by secondary analysis using trial sequential analysis.

In conclusion, the present meta-analysis reveals a significant association of low serum zinc and the risk of RAS. Although TSA confirmed this conclusion, conducting well-designed, large-scale studies targeting different populations is highly recommended.