Cross-Sectional Study to Identify Potential Risk Factors for Eczema within the Common Household Environment in Taiwan.

BACKGROUND: Much attention has been focused on environmental risk factors and their roles in eczema development. In this regard, the specific eczema risk factors in Taiwan were relatively unknown. As such, this study investigated the common indoor risk factors present in Taiwanese households. AIMS: To discuss the effects of several indoor risk factors on the prevalence of atopic eczema in Taiwan.
MATERIALS AND METHODS: A cross-sectional, population-based study was performed in Kaohsiung, Taiwan, using both survey investigation and fungal culturing. A total of 998 participants were enrolled in the survey, with 513 participants selected for fungal culture. Risks of atopic eczema were calculated as odds ratios for various risk factors using logistic regression. The correlation between potential risk factors and the fungal level was analyzed with linear regression.
RESULTS: Pet and house plants have an adjusted odds ratio of 1.434 (95% CL: 1.011-2.033) and 1.820 (95% CL: 1.229-2.696), respectively. Additionally, smoking was shown to possess an odds ratio of 1.461 (95% CL: 1.064-2.006). Wood wall has an adjusted odds ratio of 2.143 (95% CL: 1.235-3.658). Frequent bedroom shower use (β = 0.254) and hours of opened windows (β = 0.106) have shown significant positive associations with indoor fungal level.
CONCLUSION: Pets, house plants, and smoking were concluded to be major risk factors for atopic eczema. Wood wall remained controversial due to its limited sample size and possible confounders. Bedroom shower and window-opening have been shown to increase mold growth, but the lack of association with eczema suggested other allergens besides mold to be the primary eczema trigger. Copyright:

Introduction

Over the past few decades, there had been an overall increase in the prevalence of allergic diseases, such as allergic rhinitis (AR), asthma, and eczema.[12] The heightened prevalence of eczema indicated a need to identify possible risk factors in order to develop new preventative measures. Hence, this study aimed to investigate the relationship between eczema prevalence and aspects of indoor household environment, specifically those most prevalent in Taiwan. In this regard, several potential indoor risk factors, both well- and little-understood, were proposed and examined.

One of the primary risk factors was the presence of indoor dampness. Previous studies into home dampness and eczema in European countries and China have indicated some correlations between the two.[3456] Home dampness in the aforementioned studies was often defined by several common characteristics, such as “moldy odor”, “visible mold patches”, and “water leakage”.[34] This study examined all three stated building characteristics in addition to several other little-understood factors that could have contributed to indoor dampness. Taiwan, with its tropical climate, has a relatively high ambient humidity almost all year round, with periods of exceptionally high humidity level due to seasonal rain. Furthermore, most Taiwanese bedrooms followed the en suite layout, in which a bathroom shower is situated within the bedroom, thus serving as a source of high indoor moisture. This results in an indoor environment where the average humidity is unusually high, providing an ideal habitat for mold growth.[78] Therefore, it was believed that the presence and frequent use of shower could result in a high level of mold growth, which contributed to the occupant's eczema development.

Furthermore, a significant portion of Taiwanese bedroom designs was furnished with Japanese-style wood floors and walls. Due to the high humidity in Taiwan and the ease of water absorption by wood material, wooden structures were particularly prone to decay, making them significantly more vulnerable to mold colonization. Being relatively restricted to Taiwan and Japan, these Japanese-style wood floors and walls were not well-understood outside these areas, and their potential as allergen sources was not well-explored.

In addition to indoor dampness, improper ventilation of the home environment was proposed as a potential risk factor for eczema. Previous studies have indicated that poor airflow rate could result in improper removal of airborne allergens,[910] including mold spores and house dust mites (HDM), both of which were primary triggers for eczema.[11] As such, it was hypothesized that poor home ventilation would effectively trap airborne allergen indoor and contributed to eczema development.

Finally, this study also examined various well-known risk factors, including the presence of pets, house plants, and smoking habits. Pets and domestic animals have been clinically recognized as major triggers for eczema.[11121314] For household plants, contact eczema from plants was well documented both clinically and experimentally.[1115] Finally, studies on the effect of tobacco smoke on eczema have highlighted an elevated risk of eczema when exposed to first hand or second-hand smoke.[161718] It is likely tobacco smoke served as a chemical irritant to the skin, resulting in an allergic reaction.

Materials and Methods

Study design and sample population

A cross-sectional, population-based study was performed around Kaohsiung City, Taiwan to investigate the correlation between eczema prevalence and potential household risk factors. The study population was randomly selected from visiting individuals, along with any willing accompanying relatives, at Anshing Clinic, Kaohsiung, Taiwan. In accordance with the institutional review board (IRB) regulation, only participants above the age of 20 years without any major disabilities that would hinder the survey investigation were recruited into the study. The study was divided into two parts: survey investigation and fungal cultures. A total of 1009 participants were recruited. However, 11 participants were excluded due to being under the required age, leaving the actual sample size at 998 participants. Additionally, 513 out of the 998 participants volunteered for air sampling and fungal cultures.

Survey investigation

The survey investigation contained two parts. The first part was the assessment of participants' home environment. In this part, the participants were asked about the presence of bedroom shower and the frequency of shower usage, as well as signs of indoor dampness such as mold spots, water leakage, or moldy odor within their homes. Afterward, the participants were inquired about their ventilation habits, such as whether they opened their windows periodically and hours of ventilation. The participants were also asked about the presence of Japanese-style wood wall or floor in their bedrooms. Additionally, the participants were asked about their smoking habits, presence of pets or house plants, ages of the current home, cleaning habits, recent redecorations (within the last 3 months), and periodic bedding takeout.

The second part of the survey assessed the participants' eczema status and severity. The patient-oriented eczema measure (POEM), designed and proposed by Charman et al., was used for the assessment. POEM examined 10 common clinical symptoms of eczema, including itching, soreness or pain, sleep disturbance, redness of the skin, bleeding, oozing, dryness, flaking of the skin, cracking of the skin, and tightness of the skin.[19] The participants were asked about the presence and duration of all 10 symptoms. The answers were then converted into scores in accordance with the POEM scoring scheme. A total of 28 points could be obtained, with increasing points corresponding to more severe eczema symptoms. Using POEM, the participants' were divided into eczema and noneczema, with POEM score higher than 0 being classified as eczema and vice versa. For the survey investigation, POEM was translated from English to Chinese and incorporated into the survey questionnaire.

Air sampling and mold cultures

To acquire information about the indoor fungal levels within the participants' homes, air sampling and mold cultures were done. Air sampling was done through the settle dust method. Three plates of Sabouraud Dextrose Agar were given to each participant to take home. The participants were instructed to place the three plates in the center of their bedrooms, within 2 m distance between each plate. The covers of the plates were then lifted and exposed to ambient air for 3 h before replacing them. The plates were recovered and incubated for 48 h at 30°C. The plates that suffered physical damages, dehydration, infestation by animals or pests, physical contact with a foreign object (i.e., hand touch) were excluded as the growth medium was either contaminated or too damaged to be recognizable. After incubation, the colony-forming units (CFUs) were counted for all three plates, and the averages were calculated. Out of the 513 participants, 6 were excluded due to one of the abovementioned reasons, leaving a total of 507 participants for mold cultures.

Statistical analysis

To examine the correlation between eczema prevalence and various potential risk factors, univariate logistic regression was used. The results were expressed in odds ratios. For correlation between the indoor fungal level and the risk factors, univariate linear regression was used instead. In both cases, multiple logistic and linear regressions were used to control for potential confounders. All statistical analyses were done using Statistical Product and Service Solution (SPSS) package from IBM Corporation, New York, USA.

Results

The demographic summary for the entire sample population can be found in Table 1. In order to understand the relationship between eczema prevalence and potential indoor risk factors, logistic regression was used. The analysis results for the univariate logistic regression are shown below in Table 2. The data were presented as odds ratios, along with 95% confidence intervals and significance levels. As indicated below, “gender” had shown a significantly positive correlation with eczema prevalence (OR = 1.512), in which eczema prevalence was higher among females than males. “Diabetes mellitus” was also shown to have a higher prevalence of eczema (OR = 1.844). In addition, “pets” and “plants” were both shown to have significant positive correlations (OR = 1.141 and 1.638, respectively). For Japanese-style wood walls and floor, only “wood wall” had shown positive correlation (OR = 1.759). Finally, “smoking” was found to be positively correlated with eczema (OR = 1.382). Multiple logistic regression was used in order to adjust and correct for potential confounders. The results can be found in Table 3. After adjusting, it was shown that all risk factors mentioned above continued to be significantly correlated with eczema prevalence.

Table 1

Demographic data for gender, age, household environment, ventilation habit, smoking habit, average POEM score, and average CFUs

Variable	Total numbers = 998 Numbers (Percentage, %) Mean ± Standard Deviation
Gender (Female	617(61.9)
Age (years)	42.020 ± 14.647
DiabetesMellitus(+)	51(5.1%)
Hypertension (+)	98 (9.8%)
Mold Spots(+)	280 (28.1%)
Water Leakage (+)	92 (9.2%)
Moldy Odor (+)	110(11%)
Pet<+)	254 (25.5%)
Plant (+)	174 (17.4%)
Wood Floor(+)	252 (25J%)
Wood Wall (+)	96 (9.6%)
Toilet in bedroom (+)	472 (47.2%)
Bathroom in bedroom (+)	457 (45.*%)
Windows in bathroom (+)	392 (39.3%)
Frequent usage of Showers in Bed room Bathroom (+)	392 (39.3%)
Number of windows	1-32 ±0.767
Frequent openingofwindows(+)	819(82.1%)
Window-op en Time (hour)	3.52= 1.543
Age of House (year)	18.42±9.958
Daily Cleaning (+)	364 (36.5%)
Redecoration (+)	19(1.9%)
Bedding Takeout(+)	391 (392%)
Smoke (+)	384 (38.5%)
Patient-Orient! czema Measure Score	225 = 4.188
POEM score > 0 (+)	399 (40.1%)
48 Hours Colony Numbers(CFl’)	54.61 = 46.1S

Total number = 998. Data are expressed as mean ± standard deviation

Table 2

Different variables affecting eczema (POEM ≥ 1) were analyzed with univariate logistic regression

Variable	OddsRatio	95%ConfMence Interval	P value
Gender (Female)	1.512	1.159-1.973	0.002*
Age (years)	1.003	0.994-1.012	0.518
Diabetes Mellitus(+)	1.844	1.046-3 2 51	0.034*
Hypertension (+)	1.303	0.855-1.985	0218
Mold Spots (+)	1.135	0.S56-1.5O5	0.379
Water Leakage (+)	1.419	0.921-2.186	0.113
Moldy Odor(+)	1.392	0.934-2.076	0.105
Dampness Score	1.169	0.984-1.388	0.076
Pet(+)	1.141	0.852-1.526	0.376*
Plant (+)	1.683	1.207-2.347	0.002*
Wood Floor (+)	1.027	0.768-1.375	0.856
Wood WaU(+)	1.759	1.152-2.684	0.009*
Toilet in bedroom (+)	1.161	0.899-1.499	0254
Shower in bedroom (+)	1.153	0.892-1.490	0.276
Windows in bathroom (+)	1.031	0.794-1.340	0.817
Frequent usage of Showers in Bedroom Bathroom (+)	1.109	0.854-1.441	0.436
Number of windows	1.046	0.883-123 8	0.603
Frequent opening of window s(+)	0.780	0.560-1.085	0.139
Window-open Time	0.973	0.906-1.046	0.461
Age of House	0.994	0.762-1.295	0.409
Daih’ Cleaning(+)	0.994	0.682-1200	0.962
Red ecora lion (+)	0.479	0.173-1.32S	0.157
Bedding Takeout (+)	0.783	0.601-1.019	0.068
Smoke (+)	1.382	1.061-1.800	0.016*

*P < 0.05 compared with control group **P <0.01 compared with control group.

Table 3

Independent risk factors of eczema, evaluated by multiple logistic regression analysis

Variable	Odds Ratio	95%Confidence Interval	P value
Gender (Temale)	1.423	1.044-1.940	0.026*
Age (years)	1.000	0.989-1.012	0.943
Hyp ertension (+)	1.304	0.787-2.162	0.303
Diabetes Mellilus (+)	2.101	1.064-1.149	0.033*
Mold Spots(+)	1.129	0.786-1.620	0.511
Water Leakage (+)	1.416	0.836-2.3 9 8	0.196
MoHy Odor (+)	1.082	0.658-1.780	0.756
Pet (+)	1.434	1.011-2.033	0.043*
Plant (+)	1.820	1.229-2.696	0.003*
Wood Floor (+)	0.967	0.674-1.388	0.855
Wood Wall (+)	2.143	1.235-3.6 5 8	0.005*
Toilet in bedroom (+)	1.143	0.485-2.693	0.760
Show er in bed room (+)	1.270	0.535-3.0 1 7	0.588
Windows in bathroom (+)	0.945	0.672-1.331	0.748
Frequent usage of Show ers in Bedroom Bathroom (+)	0.926	0.623-1.377	0.705
Number of windows	1.047	0.849-1292	0.668
Frequent opening of window s(+)	0.971	0.580-1.628	0.912
Window-open Time	0.969	0.858-1.094	0.606
Age of House	0.993	0.977-1.010	0.435
Daily Cleaning (+)	0.982	0.705-1.367	0.914
Redecoration (+)	0232	0.050-1.084	0.063
Bedding Takeout (+)	0.731	0.533-1.002	0.052
Smoke (+)	1.461	1.064-2.006	0.019*

*Odds ratio were adjusted for age, hypertension, mold spots: water leakage, moldy odor, pet, wood floor, wood wall, frequent usage of showers in bedroom bathroom, frequent opening of windows, window-open time, age of house in full model. *P < 0.05 compared with control group; **P < 0.01 compared with control group.

For analyzing the relationship between potential indoor risk factors and CFUs, univariate linear regression was used. The analysis results are shown in Table 4. In regards to CFUs and risk factors, only “frequent use of shower in bedroom” and “frequent opening of windows” have shown a significant positive correlation to CFU counts. Same as eczema prevalence and risk factors, adjustment for possible confounding variables was made with multiple linear regression. The analysis results are shown below in Table 5. After correcting for confounders, only the two factors, “frequent use of shower in bedroom” and “frequent opening of windows,” were significantly correlated with increasing CFUs.

Table 4

The relationship between different variables and the 48-hour colony numbers was analyzed with univariate linear regression

Variable	Standardized coefficient (P)	P Talue
Gender (female)	0.001	0.968
Age (years)	0.008	0.808
Diabets Mellitus(+)	0.069	0.030 *
Hypertension(+)	-0.003	0.924
Mold Spots(+)	-0.011	0.727
Water Leakage (+)	-0.027	0.391
Moldy Odor(+)	0.016	0.624
Pet(+)	-0.005	0.868
Plant (+)	-0.020	0.529
Wood Floor(+)	-0.011	0.727
Wood Wall (+)	-0.011	0.737
Toilet in bedroom (+)	0.008	0.80S
Bathroom in bedroom (+)	0.007	0.830
Windows in bathroom (+)	•0.031	0.311
Frequent usage of Show ers in Bed room Bathroom (+)	0.098	0.005 **
Number of windows	0.009	0.783
Frequent opening of windows (+)	0J1JJ	0246
Window-op en Time	0.087	0.006
Age of House	0.016	0.646
Daily Cleaning (+)	0.016	0.60S
Redecoiation (+-)	-0.051	0.113
BeddingTakeout (+)	0.034	0293
Smoke (+)	-0.027	0.402
Allergic Rhinitis (+)	0.081	0.011 *

*P < 0.05 compared with control group: **P < 0 01 compared with control group

Table 5

Independent risk factors of 48-hour colony numbers, evaluated by multiple linear regression analysis

Variable	Standardized coefficient (P)	P value
Diabetes MelKtus(+)	0.060	0.092
Mold Spots(+)	-0.005	0.884
Water Leakage (+)	-0.052	0.158
Moldy Odor (+)	0.022	0.548
Pet (+)	0.009	0.796
Plant (+)	-0.026	0.468
Wood Floor (+)	-0.028	0.427
Wood Will (+)	-0.009	0.790
Toilet in bedroom (+)	-0.057	0.603
Bathroom in bedroom (+)	-0.136	0.238
Frequent usage of Showers in Bedroom Bathroom (+)	0.254	<0.001 »*
Window-op en Time	0.106	0.003 ••

* Standardized coefficient were adjusted for diabetes mellitus, mold spots, water leakage, moldy odor, pet, plant, wood floor, wood wall, toilet in bedroom, bathroom in bedroom, frequent usage of showers in bedroom bathroom, window-open time in full model. *P < 0.05 compared with control group; **P < 0.01 compared with control group.

Discussion

Pet and plant as potential risk factors

Logistic regression highlighted a positive correlation between the presence of pet and eczema prevalence as with the presence of household plants. In regards to pet as risk factors, domestic animals, such as dogs and cats, have long been clinically established as triggers for both eczema and AR. In addition, past studies into similar topics have further produced evidence that supported the notion of pet as a risk factor for eczema.[1113]

For house plants, contact allergy from the plant is a well-known clinical disorder. Previous studies on plant allergy have demonstrated the possibility of plant being a potential allergic trigger, especially through skin contact. Wrangsjo et al. demonstrated how contact allergy from members of the Compositae family, including various common house plants such as orchids, could exacerbate eczema symptoms.[1115] Therefore, it is reasonable to conclude pet and household plants as risk factors for eczema.

Shower in the bedroom and indoor dampness

Frequent use of shower in bedroom was observed to be positively correlated with increasing CFUs. As stated before, Taiwan has a relatively high average humidity year-round, and the architectural design of bedroom shower is relatively common among Taiwanese households. It can be reasonably inferred that excessive moisture produced by bedroom shower, compounded with the already high level of humidity in Taiwan, could potentially result in a high-dampness environment. As mold is critically dependent on humidity for effective growth,[2021] such an environment would serve as an ideal habitat for mold growth.

That being said, no clear conclusion could be drawn on whether elevated mold level is correlated with eczema or not, as eczema prevalence was not found to be significantly related to the frequent use of bedroom shower, nor with known signs of indoor dampness, such as mold patches, moldy odor, or water leakage. While some studies have indicated a correlation between indoor dampness and eczema,[36] the lack of such evidence in our study suggested otherwise.

One potential explanation for this observation was the presence of previously unknown confounders, specifically the existence of dehumidifiers. As it stood, the presence of a dehumidifier in a bedroom meant that excessive moisture produced by the bedroom shower was essentially annulled. This could explain the lack of correlation between bedroom shower and eczema prevalence in the context of our original theory, as the dehumidifier prevented the emergence of indoor dampness. Hence, no mold colonization in the bedroom environment could occur. Nonetheless, further studies need to be performed in order to understand the reasons behind the discrepancy between this study and previous studies.

Wooden structures

Investigations into Japanese-style wood walls and floor have revealed a correlation between wood walls and eczema. It was predicted that wooden structures, being made up of highly water-absorbent material, were especially susceptible to rot and decay in Taiwan's humid environment, allowing easy mold invasion. However, regression results for wood wall and CFUs did not yield any significant correlation. It is, therefore, inconclusive whether eczema was, in fact, triggered by indoor mold. One possible explanation could be that wood wall served as the source for other allergens instead of mold, such as HDM. Additionally, studies on wood-based products and furniture have also identified significant volatile organic compound (VOC) emission, such as formaldehyde and toluene, from selected wood materials, most prominently particleboard/chipboard panels.[222324252627]

Nonetheless, the examination of the wood wall distribution among the sample population revealed that the participants with wood walls consisted only 9.6% of the total study population. The limited sample size for wood wall meant that, should participants with wood wall have, by chance, unusually high proportion of individual with eczema, the probability of eczema would be falsely inflated, thus, depicting an inaccurate relationship between wood wall and eczema. Thus, future studies with a larger sample size of individuals with wood walls are warranted in order to fully explicate the relationship between the presence of wood walls and eczema prevalence.

Frequent opening of windows and open hours

Curiously, a positive correlation between window open hours and CFU level was observed, suggesting that longer window open time corresponded to higher CFUs. It was originally predicted that longer open hours would result in a decrease in CFUs instead. A review of the existing literature on ventilation and CFUs, however, supported our observations. In Chapter 3 of WHO Guideline for Indoor Air (2009), the guideline discussed the advantages and disadvantages of natural ventilation methods, including window opening. One of the disadvantages mentioned indicated that, depending on the geographical region, natural ventilation could, in fact, be counterproductive to remove allergens.[28] The guideline classified regions with relatively high ambient humidity as unsuitable for natural ventilation as longer ventilation hours could, in fact, resulted in an excessive amount of air moisture being let indoors instead.[28] This could explain our observation of higher CFUs, as the longer open hours were actually increasing indoor humidity, providing sufficient dampness for mold growth. However, logistic regression results for window open hours and eczema did not indicate any significant correlation. As such, no definite conclusion could be drawn for the relationship between natural ventilation and eczema. It is unknown why eczema was not observed to be associated with longer window opening hours. Theoretically, even if mold was not the actual primary trigger for eczema, longer window open hours should at least be associated with some changes in the level of other common indoor allergens, such as HDM.[10] One possible explanation could be that natural ventilation was ineffective in increasing airflow within the participants' houses, thus, preventing proper removal of allergen.[28] Nonetheless, more research effort is needed to understand the underlying causes for our current observation.

Smoking as a potential trigger for eczema

The effect of tobacco smoke on eczema development has been a topic of much interest. It was generally believed that tobacco smoke could serve as a potential source of chemical irritants and helped sensitize the skin against other indoor allergens.[161718] Previous studies have been done to investigate the association of eczema with first-hand and second-hand tobacco smoke. Lee et al., in an investigation into the correlation between active/passive smoking and adult-onset eczema, reported a significant increase in the risk of adult-onset eczema among smokers, compared to nonsmokers.[16] Thyssen et al. likewise observed a positive correlation between hand eczema and tobacco smoking, with smokers having a higher prevalence of eczema than nonsmokers.[18] In comparison, our logistic regression results have shown smoking to be positively associated with eczema, with smokers having a higher probability of eczema than nonsmokers. This observation was in accordance with previous study results and provided evidence that smoking constituted a major risk factor for eczema.

Practical implication and significance

As stated before, a rising trend of allergic diseases such as eczema has been observed across various regions worldwide. This rising trend also correlated to increasing costs in eczema treatment. In the US, the total annual cost of illness for eczema across the country was observed to be as high as US$3.8 billion.[29] Taiwan itself has a relatively low annual cost per household in comparison with neighboring countries, such as Singapore, China, and Korea.[30] This was most likely due to Taiwan's national health care system, which covered the majority of the eczema treatment cost. Nonetheless, this meant that a tremendous amount of financial burden was instead transferred to Taiwan's national health care system, and by extension, government public health spending.

Furthermore, eczema has also been shown to severely decrease the quality of life for both the afflicted individuals and their families.[31] This could include sleep deprivation and depression.[193132] Additionally, mental and physical fatigue due to eczema could be debilitating to an individual's productivity, as well as damaging in the long run to the psychological state. The burden of eczema could be especially damaging to young children, both developmentally and psychologically.[31]

Faced with such heavy cost, it has become clear that more permanent and effective measures are needed in order to fundamentally alleviate the negative consequences of eczema, and provide a more cost-effective measure to lessen the financial burden for eczema patients. One such method is through environmental modification, which involves the elimination of common household risk factors for eczema. Removal of risk factors could greatly reduce the exposure to harmful substances or allergens, thus, lowering the probability of triggering eczema. In this regard, the risk factors identified in this study served to add to the current knowledge of known eczema triggers and risk factors. Our study results also further lend credits to previous studies that yielded similar outcomes. Finally, the identification of these risk factors could provide new insights for designing future guidelines in improving home environment, providing a cost-effective method for ameliorating eczema.

All in all, our study results contained great contributions to the betterment of the current eczema treatment. Nonetheless, similar studies are warranted in the future to further replicate and validate our current study results. In addition, it is highly recommended that additional studies should be conducted in the future to expand upon our current understanding of novel risk factors, using our study results as a basis.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.