In this review, we have summarized all studies of eczema prevalence among children and adults conducted in China from 1985 to 2015. Among children, there was an increasing trend in most cities. However, prevalence of lifetime-ever eczema and the eczematous population among children aged 13-14-year-olds decreased from 2001 to 2009. Limited data prevented us from discerning time-trends among adults. In general, eczema prevalences in younger children were higher than in older children, and in boys compared to girls, as well as in male adults compared to female adults.
Our findings that childhood eczema prevalences increased among 6-7-year-olds children but decreased among 13-14-year-olds children were partly consistent with those findings from the ISAAC study [2] and with findings of a systematic review of worldwide incidence and prevalence of atopic eczema [4]. Williams et al. found that eczema symptom prevalence decreased from 1995 to 2006 in some previously high prevalence centers in the developed world, whereas prevalences in developing countries with previously low prevalences continued to increase in 13-14-year-olds children, whereas most centers showed increasing prevalences in 6-7-year-olds children [2]. Deckers et al. found that eczema prevalences were increasing in Africa, eastern Asia, western Europe and parts of northern Europe from 1990 to 2010 [4]. The latest lifetime-ever eczema prevalences among 3-6-year-olds children in Beijing and Chongqing in 2012, as well as among 6-7-year-olds in Hong Kong in 2002 and in Taipei in 2007, ranged from 25.0%-35.0% (Fig. 3), higher than the global-averaged level among 6-7-year-olds children from the ISAAC study Phase Three in 2006 and were comparable to the high prevalences in developed countries reported in recent years [3].
Interestingly, we found that the national-averaged prevalence and absolute eczematous population of 13-14-year-olds children in mainland China decreased from 2001 to 2009. We suspect this may be related to the marked improvement in medical conditions and health services in mainland China in the past decades, as well as to the access to basic public health services has become more and more equitable [102]. Those children who were diagnosed with eczema at a younger age may have had better medical care in recent years than before 2001, such that when they reached 13-14-year-olds, eczema prevalences were lower than in 2009. This hypothesis also is supported by our finding that eczema prevalences among younger children substantially increased during the same years, whereas eczema prevalences among 13-14-year-olds children decreased in all specific cities (Figs. 3 and 5). In the present study, we also found that national absolute eczematous populations among young infants/children were substantially higher than in older children. This finding is consistent with several previous systematic reviews or individual studies, which have reported that atopic eczema was more common among younger infants/children than among older children [1, 14, 103,104,105,106,107,108]. Specifically, Kelbore et al. conducted a facility-based cross-sectional study among 477 children aged from three months to 14 years in the Ayder referral hospital in Mekelle, Ethiopia and found that children aged three months to one year had significantly higher risk of atopic dermatitis than older children (odds ratio: 6.8; 95 % confidence interval: 1.1-46.0) [105]. Hong et al. conducted an ISAAC questionnaire-based cross-sectional study among 31,201 Korean children and found that the past 12 months prevalences of dermatitis symptoms among 0-3, 4-6, 7-9, and 10-13-year-olds children were 19.3%, 19.7%, 16.7%, and 14.5%, respectively (p-value for trend <0.001) [108]. In general, eczema is considered to be an early infancy illness that will get better as infants grow older [106, 107]. However, other allergic diseases and/or symptoms are likely to appear in those children who had eczema as infants [109,110,111,112]. Thus, we consider it reasonable that older children had lower prevalences of eczema than young infants since eczema will get better as children grow older [112]. Also, prevalence of eczema among old children (13-14-year-olds) could have decreased trends if the medical conditions and health services improved [102].
There were too few studies of adults to discern time-trends or to compare eczema prevalences in different cities (Fig. 8). There was only one cross-sectional study in three large cities of eczema prevalence among an overall population of adults [74]. However, some studies reported high prevalences among several special populations, for example, nurses and/or doctors [76, 79, 80, 82], soldiers [85,86,87,88,89,90], fruit farmers [92], hairdressers [93], and cement workers [94]. These findings indicate certain populations with special occupations and with particular environmental exposures could had higher likelihood in getting eczema, and these exposures possibly are risk factors for eczema. These findings also indicate that the national burden of eczema in adults, especially among several special populations, could still be heavy even though the medical conditions and living conditions were improved in the past years in China [102].
Our finding that boys had higher eczema prevalences than girls in most cities is consistent with findings in several studies of young children [4, 107, 113,114,115] and adults [4, 106, 107]. However, our findings that male adults generally had higher eczema prevalences than female adults are inconsistent with findings from several previous studies showing that girls’ eczema prevalence catch up to boys’ prevalence around puberty [116,117,118]. Although these inconsistences could be due to the lack of data, more studies are necessary to clarify adult eczema prevalence in China [117, 118].
There are several gaps in the literature, especially for adults. Firstly, most of the information on eczema we collected was based on questionnaires, and the eczema questions differed, especially in those studies that did not use ISAAC questions. Symptoms of atopic eczema overlap with symptoms of other conditions, such as contact dermatitis [4]. In ISAAC-related questionnaire-based studies for children [5, 7,8,9,10,11, 15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44], childhood eczema was defined as an itchy rash coming and going for at least 6 months [3]. Several studies not using the ISAAC question [6, 62,63,64] used the UK diagnostic criteria [119] of atopic eczema: history of itchy skin plus at least three following signs: 1) history of rash in the skin creases (folds of elbows, behind the knees, fronts of ankles or around the neck); 2) history of asthma or hay fever: 3) history of dry skin in the last year; 4) onset under the age of 2-year-olds; and 5) visible flexural dermatitis as defined by a photographic protocol [120]. Chinese non-ISAAC studies [50, 52, 66] based a clinical diagnosis of eczema on definitions from two Chinese textbooks (Clinical Dermatology; Practical Pediatric Dermatology) [121, 122] and a Chinese guidebook for eczema diagnosis [123]. In these Chinese books, the definition of eczema is: 1) mild level: skin lesion and erythema; 2) moderate level: papule or cracked skin and desquamation; 3) severe level: with blister, erosion and scabs. The Chinese criteria are less specific than the UK diagnostic criteria [119]. Nonetheless, both the UK diagnostic criteria [119] and the Chinese criteria [121,122,123] are stricter that in the ISAAC questionnaire [3]. Studies of adults used different clinical diagnostic criteria: two studies [73, 75] used the definition of eczema in “Clinical Dermatology” [121] and one study [72] used the UK diagnostic criteria of atopic eczema [119]. Secondly, ages of both children and adult varied in the available different studies (supplemental S1-S7 Tables). Eczema prevalence varies with age in both children and adults [14, 103,104,105]. Take the ISAAC-based studies in Beijing as an example (Fig. 2), logically, prevalences of eczema among children could not sharply decrease within one year. Since the definition of eczema and method for survey in these studied are the same, the main reason for these illogical trends of eczema prevalences probably is that ages of the children are largely different among these studies: the 10.3% in 1995 was among 6-7-year-olds children [5] and the 3.8% in 1996 was among 13-14 year-old children [7], as well as the 29.4% in 2008 was among 5-11-year-olds children [18] and the 20.6% in 2009 was among 0-14-year-olds children [24]. Thirdly, a large number of the selected studies did not provide sex information for eczema prevalence. Thus, we could not establish and compare the time-trends for boys and girls. Fourthly, most selected studies were conducted in large cities; there were few studies eczema in rural populations in mainland China. Several studies have reported that rural eczema prevalences are substantially lower than urban eczema prevalences [2, 4, 6, 40, 50, 118]. Thus, we have likely overestimated national eczema prevalence. Also, few studies have been done of non-Han Chinese, so we do not know the magnitude or direction of error that this lack of information introduces [3, 18, 26, 113].
To our best knowledge, no systematic review has summarized national time-trends of eczema prevalences or absolute eczematous populations of children and adults in China. The 93 articles [5,6,7,8,9,10,11, 15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100] we summarized in the quantitative synthesis had similar findings regarding eczema prevalences and their time-trends, and so can be said to give a reasonably accurate picture of eczema from 1985 to 2015 in China. Two reviewers independently checked the quality of all 93 full articles [5,6,7,8,9,10,11, 15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100] with consideration of the definition of eczema, the studied locations and populations, as well as the method used for sample and survey. Although there were only 16 articles [5, 7,8,9, 11, 15,16,17,18, 24,25,26,27, 29, 35, 44] for children and for young children, and only one for adults [74] that used the ISAAC questionnaire, most of these articles had large sample sizes and high response rates (Table 1). The ISAAC questions have been shown to provide adequate symptom-derived estimates of eczema prevalence [124, 125]. Another study has shown that the Chinese translated version of the ISAAC core questions provides sufficient information for atopic eczema among children in China [126]. Thus, the ISAAC questionnaire studies likely have high reliability with respect to eczema prevalence. The quantitative synthesis for the national-levels eczema prevalences used only those studies of at least three cities for mainland China [5, 7, 16, 35, 74] and therefore we considered those estimated national prevalence of eczema and absolute eczematous populations could indicate the actual status, at least in cities of mainland China.
In future work, the diagnostic criteria for atopic eczema among infants, children, and adults should be standardized, and the questionnaire designed according to these standards. Since the effectiveness of ISAAC core questions for childhood eczema have been validated [124,125,126], we recommend using this questionnaire for eczema [5]. Nation-wide studies stratified by age-groups and sex, and with standardized reporting format would be useful to compare prevalences and time-trends in cities.
Many studies have tried to find explanations for the changing time-trends in eczema prevalences among children and adults [1, 4, 127,128,129,130,131,132,133,134,135,136,137,138,139]. The prevalence changes could be an artifact of methodology, due to possible changes in diagnostic criteria over time, and differences in the study design [4, 140,141,142]. However, the prevalence changes could be real. Several studies have proposed that changes in household and ambient environment-related exposures [127,128,129,130,131,132,133,134, 139], while others have suggested changes in lifestyle and dietary habits [135,136,137,138], may be causing increased eczema prevalences. Specifically, several studies have presented evidence that indoor air pollution [127, 128] and home dampness-related exposures [69, 130,131,132,133,134, 139] are risk factors for childhood and adult eczema. Outdoor air pollution [129], shorter breastfeeding [133], antibiotic use in infancy [135], early pet-keeping [137], and eating fast foods [136, 138] have been suspected of association with childhood eczema, but findings from different studies have been inconsistent regarding these factors. More well-designed and large-scale studies are warranted to provide explanations for the increase in eczema prevalence among children and adults in China.
This study had some limitations. First, the collected eczema prevalences among children were based on the population-based survey and parents-reported or self-reported questionnaire. The studied eczema was defined using a positive answer to a single question in the most studies that were based on the ISAAC questionnaire. There may be reporting error and recall bias. Second, the definition in different studies could have notable differences, especially for those studies among adults that were not based on the ISAAC questionnaire. Third, in the estimation of absolute eczematous populations, we assumed that the proportions of populations for each specific age were the same within various age groups. This assumption could introduce error in the estimation.