Prevalence and risk factors of acne scars in patients with acne vulgaris

Abstract Background Acne scar is a persistent complication of acne vulgaris. However, the prevalence and risk factors are still unclear. This study aimed to assess the global prevalence and risk factors of acne scars in patients with acne. Materials and methods A systematic search of published studies in three databases was performed and the meta‐analyses were conducted. Results Finally, we included 37 studies involving 24 649 acne patients. And, the pooled prevalence of acne scars in these patients was 47% (95% confidence interval [CI]: 38–56%). Besides, the differences in prevalence were observed based on the subgroup analysis for age, gender, acne severity, source of patients, and so on. Subsequently, we quantified the relationship of three risk factors with acne scars: male gender (odds ratio [OR]: 1.58, 95% CI: 1.19–2.09), positive family history of acne (OR: 2.73, 95% CI: 1.26–5.91), and acne severity (OR for moderate acne: 2.34, 95% CI: 1.54–3.57; OR for severe acne: 5.51, 95% CI: 2.45–12.41). Conclusion Herein, we found that 47% of acne patients suffered from acne scars and identified three risk factors: male gender, positive family history of acne, and acne severity. In order to reduce acne scarring, attention and effective therapy early in the course of acne is important.

Due to their negative impact and intractability, acne scars have received increasing attention. Although several observational studies have reported the epidemiological data of acne scars, these studies show considerable variability. [6][7][8] The global prevalence of acne scars in patients with acne remains unknown, and no meta-analysis has yet been conducted for this problem. In addition, the risk factors of acne scars were not thoroughly described. Some risk factors were reported including the worst-ever severity of acne, duration of acne, family history of atrophic acne scars, and lesion manipulation behaviors. 9 However, the relationship between risk factors and acne scars were not quantified. Therefore, through a systematic and comprehensive literature search, we performed a meta-analysis designed to assess the prevalence and risk factors of acne scars in patients with acne.

This meta-analysis was performed in accordance with the Preferred
Reporting Items for Systematic Reviews and Meta-Analyses guidelines.

Study selection and eligibility criteria
Firstly, we removed all records of duplicates. Then, the titles and abstracts of studies were independently examined by two investigators for the initial selection stage. Next, the second selection stage was based on a full-text review by two investigators. The whole selection process was shown in Figure 1.
As for the eligibility criteria, studies were included if they met the

Data analysis
The aim of this study was to determine the prevalence and risk factors of acne scars in patients with acne. Thus, we performed meta-analyses to obtain the pooled prevalence with 95% CIs and the ORs of risk factors with 95% CIs as main results.
Heterogeneity among the studies was examined using the I 2 statistic: low heterogeneity, I 2 < 25%; moderate heterogeneity,

Search history and study characteristics
In the original search, a total of 7523 studies were identified from the PubMed, EMBASE, and Web of Science databases, together with additional sources, using the primary search strategy. In updated search, 887 studies were identified. A total of 4834 nonduplicate records were screened by titles and abstracts, yielding 304 studies for full-text assessment. Finally, 37 studies were selected for this meta-analysis, while the remaining studies were excluded for various reasons. The detailed search history is presented in Figure 1. Moreover, according to the NOS grading system, the quality of all studies with a score ≥7 was evaluated as high. More details about the characteristics of the study populations are shown in Table S1. No publication bias was found according to Egger's test (p > 0.05) and Begg's test (p > 0.05).

Prevalence of acne scars in patients with acne
Overall, quantitative analysis of 37 studies yielded a pooled acne scar prevalence of 47% (95% CI: 38−56%) in patients with acne ( Figure 2). Considering the heterogeneity was high (I 2 = 99.7%), we

of 9
LIU ET AL.

F I G U R E 2
Forest plots of meta-analysis for the pooled prevalence of acne scars in patients with acne. CI, confidence interval.
further processed sensitivity analysis and subgroup analysis. The result for sensitivity analysis confirmed the robustness of the pooled value ( Figure S1). Besides, we did not find the source of high heterogeneity by subgroup analyses listed in Table 1 Abbreviation: CI, confidence interval. † Prevalence of subtype and location is the proportion conducted in patients with acne scar rather than patients with acne.

of 9
LIU ET AL.

F I G U R E 3
Forest plots of meta-analysis for the OR of acne scars by sex. CI, confidence interval; OR, odds ratio.

Pooled ORs of risk factors in acne scars
From the subgroup analysis, we noted that some factors might be related to acne scars. Therefore, we further performed metaanalysis on ORs for risk factors. Finally, we found three risk factors.

DISCUSSION
In this meta-analysis, the prevalence of acne scars among patients with acne was 47%. This pooled value was similar to previous studies on large population. 11,12 Although we observed high heterogeneity, the source of heterogeneity did not appear to be identified through sub- F I G U R E 5 Forest plots of meta-analysis for the OR of acne scars by acne severity. CI, confidence interval; OR, odds ratio.

of 9
LIU ET AL.
We found that positive family history of acne increased the risk of acne scars (OR = 2.73, 95% CI: 1.26−5.91). Acne had a highly heritable trait and some genetic susceptibility loci, including SELL (i.e., Selectin L) and TGFB2 (i.e., Transforming Growth Factor Beta 2), which were implicated in scarring, were identified. 13 Genetic factors might play a role in acne scarring. For innate immunity profiles, the differences between patients prone to scars (PS) and not prone to scars (NPS) were observed even in normal skin. 14 Besides, Type IV delayed hypersensitivity response was found in NPS, while a predominantly adaptive immune response was present in PS. 15 And inflammatory immune processes persisted longer in PS and plasma cells are specifically involved in immune response of evolved-lesions. 16 These clues suggested that there might be a close connection between acne scars and genetic factors, but more studies were needed for further exploration.
In the subgroup meta-analysis, we found that the estimates for prevalence in mild, moderate, and severe acne were 46%, 67%, and 82%, respectively. Our further meta-analysis suggested that acne severity was a risk factor for acne scars, with an OR of 2.34 (95% CI: 1.54−3.57) in moderate versus mild acne and 5.51 (95% CI: 2.45−12.41) in severe versus mild acne. Our results showed that acne scars affected all levels of acne, and that the prevalence of acne scars increased with the severity of acne. Severe acne was often accompanied by prolonged and exacerbated skin inflammation, which was more likely to result in acne scars. However, it was notable that the prevalence of acne scars in mild acne was as high as 46% in our study. We were supposed to recognize that acne was a disease of sebaceous gland unit, 17 and long-term inflammatory response could cause irreversible destruction of sebaceous gland structure, contributing to atrophic scar formation. 16 Although the inflammation in mild acne was limited to some units and the inflammation degree was milder than that in severe acne, 18 the destruction of these units could also cause acne scars.
Therefore, it was a mistake that only patients with severe acne were examined and treated for acne scars. It is also important to treat mild acne to prevent development of acne and formation of scars.
Previous studies have reported that acne scars were more common in men than in women. 19 Our study yielded a prevalence of 58% in men and 46% in women, and further analysis indicated that male gender might be a risk factor for acne scars (OR = 1.58, 95% CI: 1.19−2.09).
Men were more likely to suffer from severe acne than women, 20 which might be associated with high androgen level 21 and special sebaceous gland. 22 Men were found to have cauliflower-shaped sebaceous glands and depth-dependent differences in sebaceous unit areas, 22 and this special shape might lead men to be more prone to severe acne than women, contributing to acne scar formation. Another possible reason was that most women were more self-conscious of their appearance than men and were more likely to seek help from their physicians for acne. As a result of the early and timely treatment, they had a lower chance of developing acne scars. The above factors might contribute to gender differences in acne scarring, but the details needed to be further studied.
We also found that the prevalence of acne scars in the adult acne group (59%) was significantly higher than that in the nonadult acne group (32%). The adult acne was usually mild or moderate in severity.
However, since most of the lesions were inflammatory, they might become resistant to drugs such as antibiotics and isotretinoin, 23 leading to greater scarring. However, it must be noted that in the adult acne group, the prevalence of acne scars in late-onset and persistent acne was unknown. It was also unclear whether the process of scar formation differed between late-onset and persistent acne. These problems require further investigation to explore their distinctions.
The prevalence of acne scars also showed geographical differences.
In comparison with Asia (52%) and Europe (51%), we observed a lower prevalence in Africa (31%), where the majority of the population was dark-skinned. Interestingly, individuals with dark skin were more prone to keloid and hypertrophic scars. 24 However, they were less likely to develop acne scars, which were predominantly of the atrophic type.
This might be related to genetic susceptibility, including major histocompatibility complex genes and SMAD (i.e., SMAD Family Member 2) genes, 25,26 but the exact mechanism needs to be determined. Moreover, in people with dark skin, small atrophic scars might be ignored during visual examination.
This study had some limitations. First, as mentioned above, there was high heterogeneity for the pooled prevalence. Second, small sample size of some included studies might result in unsatisfactory representativity. Third, in terms of source of patients, the prevalence was much higher in clinic patients (53%) than in community patients (27%).
Individuals with more severe acne and embarrassing scars tended to seek medical advice more frequently. Thus, a proportion of patients with mild acne and mini acne scars who did not visit the hospital might have been missed. Because most of our included studies were performed in the clinic, the pooled prevalence might have been overestimated. In addition, the measure of assessment was also different: some were by physician-diagnosed, while others were by self-reported.
Lack of unified measurement standards could lead to significant differences in results. Therefore, further well-designed large prospective studies, taking into account these potential confounders, are required in the future.
In conclusion, based on the published data, this meta-analysis found that 47% of patients with acne suffered from acne scars and male gender, positive family history of acne, and acne severity were risk factors for acne scars. In order to prevent further aggravation of acne and formation of scars, it is important to treat acne at an early stage.

ACKNOWLEDGMENTS
This study was supported by National Natural Science Foundation of China (grant number: n82073462).