- Email: [email protected]
Water hardness and eczema at 1 and 4 y of age in the INMA birth cohort
Environmental Research 142 (2015) 579–585
Contents lists available at ScienceDirect
Environmental Research journal homepage: www.elsevier.com/locate/envres
Water hardness and eczema at 1 and 4 y of age in the INMA birth cohort Laia Font-Ribera a,b,c,d,n, Esther Gracia-Lavedan a,b,c, Ana Esplugues d,e, Ferran Ballester d,e, Ana Jiménez Zabala f,g, Loreto Santa Marina d,f,g, Ana Fernández-Somoano d,h, Jordi Sunyer a,b,c,d, Cristina M. Villanueva a,b,c,d a
Centre for Research in Environmental Epidemiology (CREAL), Barcelona Biomedical Research Park, 88 Doctor Aiguader, Barcelona 08003, Spain Universitat Pompeu Fabra (UPF), 80 Doctor Aiguader, Barcelona 08003, Spain c IMIM (Hospital del Mar Medical Research Institute), 88 Doctor Aiguader, Barcelona 08003, Spain d CIBER Epidemiologia y Salud Pública (CIBERESP), 3-5 Melchor Fernández Almagro, Madrid 28029, Spain e FISABIO-UJI-Universitat de València Joint Research Unit, Av. Catalunya 21, Valencia 46020, Spain f Public Health Division of Gipuzkoa, Basque Government, 4 Av. de Navarra, San Sebastian 20013, Spain g BIODONOSTIA Health Research Institute, Paseo Dr. Beguiristain, San Sebastian 20014, Spain h Universidad de Oviedo, Campus del Cristo, Oviedo, Spain b
art ic l e i nf o
a b s t r a c t
Article history: Received 18 June 2015 Received in revised form 17 July 2015 Accepted 18 July 2015
Background: Exposure to hard water has been suggested as a risk factor for eczema in childhood, based on limited evidence from two ecologic and two cross-sectional studies. Objectives: We evaluate this hypothesis for the first time in early infancy using prospective data from a mother–child cohort study. Methods: We used data from the INMA cohorts in Gipuzkoa, Sabadell and Valencia, Spain (N¼ 1638). Current and ever eczema, bathing frequency and duration and covariables were collected by questionnaires at 14 months (14 m) and 4 years (4 y). Calcium carbonate (CaCO3) level in municipal water was assigned to home addresses at birth, 14 m and 4 y. We calculated Odds Ratio (OR) of eczema related to CaCO3 at home, bath exposure and a combination of both. Results: Prevalence of eczema ever was 18.4% at 14 m and 33.4% at 4 y. Mean CaCO3 ranged from 51.6 to 272.8 mg/L among areas. No association was detected between water hardness at home and current or ever eczema. Adjusted OR was 0.79 (95%CI ¼ 0.45, 1.39) at 14 m and 0.93 (0.56, 1.52) at 4 y among children in the highest vs. lowest tertiles of CaCO3. Bath exposure alone or in combination with water hardness did not increase the OR of eczema at 14 m or 4 y either. Conclusions: We did not find an association between eczema and water hardness at home or bathing exposure during the first four years of life. This first cohort study in a critical age period with improved exposure assessment does not confirm the association suggested among children by previous studies. & 2015 Elsevier Inc. All rights reserved.
Keywords: Calcium carbonate Bathing Atopic dermatitis Early infancy
1. Introduction The prevalence of allergy-related diseases such as eczema, rhinitis and asthma has reached epidemic proportions in highn Corresponding author at: Centre for Research in Environmental Epidemiology (CREAL), Barcelona Biomedical Research Park, 88 Doctor Aiguader, Barcelona 08003, Spain. E-mail addresses: [email protected] (L. Font-Ribera), [email protected] (E. Gracia-Lavedan), [email protected] (A. Esplugues), [email protected] (F. Ballester), [email protected] (A.o.s. Jiménez Zabala), [email protected] (L.o.s. Santa Marina), [email protected] (A. Fernández-Somoano), [email protected] (J. Sunyer), [email protected] (C.M. Villanueva).
http://dx.doi.org/10.1016/j.envres.2015.07.013 0013-9351/& 2015 Elsevier Inc. All rights reserved.
income countries (Eder et al., 2006). Eczema (or atopic dermatitis) is a chronic inflammatory skin disease currently affecting 15–30% of children characterized by a persistent itching (Bieber, 2008). It can seriously impair the quality of life of children and their parents and has a substantial cost for families and health providers (Garg and Silverberg, 2015). Eczema frequently starts early in infancy and is often the prelude to the development of other atopic diseases during childhood (Spergel and Paller, 2003). In this context, understanding and controlling risk factors for the development and progression of eczema during early life will have an important public health impact. Among several other environmental factors, water hardness has been suggested as a risk factor for childhood eczema (Flohrs et al., 2014). Water hardness is caused by a variety of dissolved
580
L. Font-Ribera et al. / Environmental Research 142 (2015) 579–585
polyvalent metallic ions, predominantly calcium and magnesium and is most commonly expressed as calcium carbonate (CaCO3) (World Health Organization, 2011). Water hardness is supposed to be innocuous for health and is not regulated in drinking water, although epidemiological evidence suggests an inverse association between magnesium levels in drinking water and cardiovascular mortality (World Health Organization, 2011; World Health Organization, 2009). Previous evidence on exposure to water hardness and eczema in children is limited to four studies, two ecological studies in UK (McNally et al., 1998) and Japan (Miyake et al., 2004) and two cross-sectional studies in Spain (Arnedo-Pena et al., 2007) and Belgium (Chaumont et al., 2012). These studies found a positive association between residential water hardness and eczema among specific age groups of school-aged children. However, a randomized control trial failed to show an improvement in eczema severity in children 0–16 y of age after the installation of a water softener at home (Thomas et al., 2011). We aim to test this hypothesis for the first time in early infancy with a child birth cohort, considering the individual variation in activities that affect dermal exposure to water, such as the frequency and duration of showering and bathing. We aimed to evaluate the association between water hardness at home, bathing exposure and eczema at 14 m and at 4 y of age in three cohorts of the INMA Spanish birth cohort.
2. Material and methods 2.1. Study design and population A mother–child cohort study was set up in different Spanish areas following a common protocol to constitute the INMA [Infancia y Medio Ambiente (Childhood and the Environment)] Project (Guxens et al., 2012). The present analysis includes the Gipuzkoa, Sabadell, and Valencia cohorts for which water hardness data could be collected. Study subjects were recruited at the 12th week of gestation and eligibility criteria for enrollment were maternal age Z16 y, singleton pregnancy, planning to deliver at the study hospitals, being able to communicate in either of the official languages, and not having followed an assisted reproduction program. The study sample was representative of the target population in terms of prenatal care attendance in the public health system (used by 480% of the pregnant women); however, the educational level was higher than the target population average. From 45% to 68% of the eligible pregnant women agreed to participate, and enrollment periods ranged from November 2003 in Valencia to February 2008 in Gipuzkoa, including 638 in Gipuzkoa, 657 in Sabadell and 827 in Valencia (total 2122). From the initial sample at recruitment, 2019 women (95.2%) were followed until delivery (Guxens et al., 2012). A follow up was conducted at 12– 14 m after delivery in Gipuzkoa (N ¼ 575; 90.1%), Sabadell (582; 88.6%) and Valencia (708; 85.6%). This will be referred as the 14 months (14 m) follow up. Another follow up was conducted at 4 years of age (4 y), with 1596 children included (85.6% of those at 14 m). Study period included years from birth until the 4-y follow up, i.e. 2004–2012 in Valencia and Sabadell, and 2006–2012 in Gipuzkoa. The study was approved by the ethical committees of the participating centers, and all subjects gave written consent at enrollment and delivery. 2.2. Questionnaire data Fig. 1 illustrates data collected in the different follow-ups. Personal information was gathered through parental administered questionnaires in face-to-face interviews. Eczema was assessed (through questionnaire) at 14 m with the question “Has your child
Fig. 1. Scheme of the study design, INMA cohorts (2004–2012). Data was collected during pregnancy, at birth, at 14 m and at 4 y. Exposure variables are indicated in white squares (bathing frequency and duration, CaCO3 at home and a combined score). Outcome variables are shown in dark gray squares (eczema ever at 14 m and at 4 y and current eczema at 4 y). Bathing FM¼ bathing during the first months of life (ascertained at 14 m follow-up).
ever suffered from atopic eczema?”, from the validated ISAAC questionnaire (Ellwood, 2005). The questions “Has your child suffered from atopic eczema during the last 12 months?” and “Has your child used eczema medication during the last 12 months?” were asked at 4 y. “Current eczema” at 4 y was defined as having had eczema or taken medication in the last 12 m. A positive answer in any of the three questions classified children as “Ever having had eczema” at 4 y. The address of residence was collected at pregnancy and changes of residence were ascertained in each followup. The frequency and duration of baths and showers was collected at 14 m and 4 y. The frequency and duration during the first months of life was also asked in the 14 m questionnaire. Weekly minutes bathing or showering was calculated for the first months of life, 14 m and 4 y and was categorized in tertiles. Maternal asthma and eczema (yes, no), education (primary or less, secondary, high), ethnicity (white, other) and age were ascertained in the interview at recruitment. Birth date and sex were collected at delivery. Day care attendance (yes, no) was collected at 14 m questionnaire. Maternal smoking (yes, no), pet contact (yes, no), use of moisturizer (yes, no), swimming pool attendance in the last 12 m (yes, no) was collected at 14 m and 4 y questionnaires. 2.3. Environmental water hardness levels Concentration of calcium carbonate (CaCO3) in municipal water was ascertained in each cohort from local authorities and water companies. CaCO3 measurements were available from 2006 to 2012 (See Supplemental Material, Table S1). Some municipalities had CaCO3 data for different areas within the municipality and geographical variability in the levels within these municipalities was assessed. Accordingly, a water area was assigned to each municipality, with the exception of six municipalities that had two or more water areas, resulting in 55 water areas. Changes in CaCO3 levels over the study period were evaluated in each water area and a statistically significant trend was not detected. Seasonal variability over a year was also assessed in the municipality with more study participants (Sabadell) conducting new CaCO3 measurements using a commercial kit (HI 3812 Hardness Test Kit, HANNA Instruments). Results indicated no significant trends within a year. Accordingly, a single value for the whole study period was assigned to each water area, as the median value of all
L. Font-Ribera et al. / Environmental Research 142 (2015) 579–585
581
measurements collected. For details on temporal and seasonal variability assessment see Supplemental Material, page 2. In total, 450 CaCO3 measurements were used from 45 municipalities, where lived 90.1% of the study population based on address at birth (see Supplemental Material, Table S1). 2.4. Statistical analysis A water area at birth, 14 m and 4 y was assigned to each participant according to the address of residence (Fig. 1). The mean CaCO3 levels between the water area at birth and at 14 m was calculated and referred as “CaCO3 (mg/L) from 0 to 14 m”. The mean CaCO3 levels in the water areas at birth, at 14 m and at 4 y was calculated and referred as “CaCO3 (mg/L) from 0 to 4 y”. A score combining the CaCO3 levels at birth and the duration of bathing at first months and at 14 m was also calculated. The association between CaCO3 from 0 to 14 m and eczema at 14 m and between CaCO3 from 0 to 4 y and eczema at 4 y was assessed. Generalized additive models (GAM) were used to assess non-linear associations between eczema and CaCO3 as a continuous variable. Logistic regression models were fitted to estimate odds ratios (OR) and 95% confidence intervals (95%CI) for CaCO3, bath duration and the combined score for continuous variables and tertiles. Models were built for eczema at 14 m and 4 y adjusted by cohort, sex, maternal allergy and maternal education. A sensitivity analysis was performed among the two cohorts with more CaCO3 measurements (Gipuzkoa and Valencia) using cohort-specific tertiles. Statistical analyses were performed with the statistical package STATA 12.0.
3. Results From the 1865 children with follow-up at 14 m, 47 were excluded for not having data on eczema at 14 m nor at 4 y, leading to 1818 children. From these, 180 (9.9%) did not have data on CaCO3 at home address, resulting in an analyzed sample of 1638 children at 14 m (See Supplemental material, Table S2). The included children had older mothers, Caucasian ethnicity, higher education, more pets at home, less maternal smoking and the proportion of
Fig. 2. Prevalence of eczema in the three INMA birth cohorts. N ¼1638. Current eczema is defined as having eczema in the last 12 m or taking eczema medication in the last 12 m.
excluded was higher for the Valencia cohort (See Supplemental Material, Table S3). Several characteristics of the population varied between the three cohorts (Table 1). The overall ever eczema prevalence was 18% at 14 m and 33% at 4 y (Fig. 2). Eczema at 14 m was higher among males, those with maternal asthma and using a moisturizer (See Supplemental Material, Table S4). Eczema at 4 y was higher among babies with maternal eczema, born preterm and those attending a day care and using a moisturizer at 14 m. The median CaCO3 levels in the different water areas ranged from 47.5 to 272.8 mg/L and varied between and within cohorts (Fig. 3). The mean CaCO3 levels at participant's home at birth ranged from 119.0 mg/L in Gipuzkoa to 240.9 mg/L in Sabadell (Table 2). The proportion of children at the tertiles of CaCO3 levels at home varied largely between cohorts. Median water contact from bathing was 35 min/week (interquartile range (IQR) ¼20–60; range¼ 0–245) during the first months of life, 70 min/week (IQR ¼40–105; range 0–420) at 14 m and 70 min/week (IQR ¼40– 120; range 6–630) at 4 y. Bath duration varied by cohort and season of birth, and was negatively associated with educational level and related variables such as having pets at home, smoking, ethnicity, day care attendance or maternal age (results not shown). CaCO3 at home was not associated with eczema at 14 m or at 4 y in the linear models (results not shown). Results from the GAM
Table 1 Descriptive of the population characteristics at birth, at 14 months (14 m) and at 4 years follow-up (4 y). N¼ 1638. Gipuzkoa
Sabadell
Valencia
Total
N
%
N
%
N
%
N
%
N
539
32.9
519
31.7
580
35.4
1638
100
Newborn Sex, male Ethnicity, other than white Preterm, yes
267 8 17
49.5 1.5 3.1
268 17 16
51.6 3.3 3.1
305 28 32
52.6 4.8 5.5
840 53 65
51.3 3.3 4.0
0.58 o 0.01 0.06
Maternal characteristics Maternal education, primary or less Maternal age, 4 35 y Maternal asthma, yes Maternal eczema, yes
72 98 31 33
13.4 18.2 5.7 6.1
140 88 42 66
27.1 17.0 8.1 12.7
170 95 43 59
29.3 16.4 7.4 10.2
382 281 116 158
23.4 17.2 7.1 9.7
o 0.001 o 0.001 0.30 o 0.01
Baby characteristics at 14 m and 4 y Age (months) at 14 m follow-up, mean (SD) Pet at home at 14 m, yes Day care attendance at 14 m, yes Maternal smoking at 14 m, yes Swimming at 14 m, yes Moisturizing lotion at 14 m, no Age (years) at 4 y follow-up, mean (SD) Swimming at 4 y, yes
14.5 86 250 96 249 45 4.5 372
(0.7) 16.3 47.3 19.0 47.2 8.6 (0.1) 94.2
14.6 170 151 126 323 16 4.4 444
(0.7) 34.0 30.3 25.7 64.7 3.2 (0.2) 96.1
12.0 273 120 170 275 – 4.3 555
(0.8) 47.1 20.9 29.8 47.4 – (0.2) 96.0
13.6 529 521 392 847 61 4.4 1371
(1.4) 32.9 32.5 25.1 52.7 6.0 (0.2) 95.5
o 0.001 o 0.001 o 0.001 o 0.001 o 0.001 o 0.001 o 0.001 0.30
p-Value for Pearson χ2 test, except age for which an ANOVA was performed.
p-Value
582
L. Font-Ribera et al. / Environmental Research 142 (2015) 579–585
Fig. 3. Water hardness in the study population. CaC03 concentration (mg/L) in the 55 water zones (left) and in the birth address of children in the three cohorts (right).
Table 2 Description of the exposure variables: CaCO3 levels at home, bathing exposure and combined score in the three INMA birth cohorts. N ¼ 1638. Gipuzkoa
Sabadell
Valencia
Total
N
N
%
N
%
N
%
240.9 0 99 4210
(30.1) 0 19.1 80.9
182.7 105 403 73
(19.2) 18.1 69.3 12.6
180.2 553 589 496
(60.8) 33.8 36.0 30.3
CaCO3 from 0 to 4 y (mg/L) o173 448 83.1 173–209.1 88 16.3 4209.1 3 0.6
0 91 428
0 17.5 82.5
106 403 71
18.3 69.5 12.2
554 582 502
33.8 35.5 30.7
Bathing at first months of lifea ≦30 282 54.2 31–35 159 30.6 435 79 15.2
(min/week) 191 38.8 171 34.8 130 26.4
89 217 274
15.3 37.4 47.2
562 547 483
35.3 34.4 30.3
CaCO3 from 0 to 14 m Mean (SD) 119.0 o173 448 173–209.1 88 4209.1 3
% (mg/L) (51.9) 83.1 16.3 0.6
Bathing at 14 m (min/week) ≦45 245 48.0 46–80 173 33.9 480 92 18.0
24.1 35.2 40.7
142 204 233
24.5 35.2 40.2
500 542 233
32.1 34.8 33.1
44.2 27.8 28.0
154 148 277
26.6 25.6 47.8
608 354 471
42.4 24.7 32.9
Combined score at first months (CaCO3 mg/L*min/week bathing) o3768 386 73.8 92 18.5 56 9.6 534 3768–7512 91 17.4 190 38.2 259 44.7 540 47512 46 8.8 215 43.3 265 45.7 526
33.4 33.7 32.9
Showering/bathing at o61 250 60–104 78 4104 65
113 165 191
4 y (min/week) 63.6 204 19.8 128 16.6 129
Combined score at 14 m (CaCO3 mg/L*min/week bathing) o7140 354 67.8 66 13.3 114 19.7 7140– 124 23.8 166 33.4 246 42.4 15,435 415,435 44 8.4 265 53.3 220 37.9
534 536
33.4 33.5
529
33.1
a
Bathing at first months of life was assessed at 14 m questionnaire. All exposure variables were distributed differently among cohorts (all p-values for Pearson χ2 testo 0.001).
models confirmed the lack of association (p-values for nonlinearity 40.05) (See Supplemental Material, Fig. S1). In the categorical analysis (Table 3), water hardness at home was also not associated with eczema ever at 14 m (OR¼ 0.79; 95%CI ¼0.45–1.39 for the highest vs. lowest tertile) or at 4 y (OR ¼0.94 95%CI ¼0.57– 1.53 in the highest vs. lowest tertile). The exclusion of the cohort with less CaCO3 data (Sabadell) or the stratification using specificcohort tertiles in Gipuzkoa and Valencia also led to null
associations. Excluding children who did not have eczema at 14 m also led to null associations between CaCO3 at home from 0 to 4 y and eczema at 4 y in all the cohorts and combined (OR¼ 1.32; 95% CI ¼0.65–2.69, results not shown). Self-reported bathing during the first months of life was not associated with eczema at 14 m or at 4 y (Table 4). However, when bathing and eczema were analyzed cross-sectionally (at the same age), a significant negative association was detected. OR of eczema at 14 m was 0.49 (95%CI¼ 0.35–0.69) among children in the highest vs. lowest tertile of bathing exposure at 14 m. The score combining bathing and CaCO3 levels at home led null associations with current eczema at 4 y, but some negative associations were detected with eczema ever at 14 m and at 4 y (Table 5). Stratified analysis by maternal allergy or swimming pool attendance did not modify the results (results not shown).
4. Discussion Water hardness at home and bathing duration were not related to higher eczema at 14 m and at 4 y of age in the Spanish INMA birth cohorts evaluated. This is the first longitudinal study assessing this hypothesis and the first focusing on the first years of life. The results are in agreement with a randomized controlled trial (Thomas et al., 2011) and in disagreement with two ecologic (McNally et al., 1998; Miyake et al., 2004) and two cross-sectional (Arnedo-Pena et al., 2007; Chaumont et al., 2012) studies suggesting a positive association between eczema and water hardness during childhood. Early childhood is a critical age period, and this is the first study to evaluate eczema risk and water hardness at this age. Around 60% of eczema cases begin within the first year of life and 85% before 5 y of age (Bieber 2008). In our population, 36% of children had ever had eczema at 4 y, and among them, 71% already had it at 14 m of age, highlighting the importance of early life factors in the development of the disease. Previous studies (McNally et al., 1998; Miyake et al., 2004; Arnedo-Pena et al., 2007; Chaumont et al., 2012) were focused on older children, and the associations detected varied across age groups. The studies in UK and Castelló (Spain) found positive associations among children at primary school but not among those at secondary school (McNally et al., 1998; Arnedo-Pena et al., 2007). The Japanese study was conducted among children aged 6–12 y and found a positive association among the oldest but not among the youngest children (Miyake et al., 2004). Previous literature was inconsistent in arguing whether water hardness would increase the risk of new eczema cases (Thomas et al., 2011) or would exacerbate the disease (McNally et al., 1998). In the present study we could estimate
L. Font-Ribera et al. / Environmental Research 142 (2015) 579–585
583
Table 3 Association between eczema at 14 m and 4 y and CaCO3 at home (mg/L) in the Gipuzkoa (G), Valencia (V) and Sabadell (S) cohorts. Eczema ever at 14 m CaCO3 from 0 to 14 m
Eczema ever at 4 y CaCO3 from 0 to 4 y
Current eczema at 4 y CaCO3 from 0 to 4 y
Tertiles CaCO3 (mg/L)
N
OR (95%CI)
N
OR (95%CI)
N
OR (95%CI)
All cohorts (G, V, S)
o 173 173–209 4 209 N model
553 589 496
1 1.00 (0.66, 1.51) 0.79 (0.45, 1.39) 1581
554 582 502
1 0.94 (0.66, 1.35) 0.94 (0.57, 1.53) 1459
554 582 502
1 0.79 (0.51, 1.22) 1.25 (0.71, 2.20) 1428
Gipuzkoa, Valenciaa
o 151 151–176 4 176 N model
373 379 367
1 0.93 (0.63, 1.38) 1.06 (0.72, 1.57) 1104
373 375 371
1 0.94 (0.59, 1.48) 1.08 (0.67, 1.73) 987
373 375 371
1 1.05 (0.60, 1.84) 1.08 (0.61, 1.93) 967
Gipuzkoaa
o 54 54–143 4 143 N model
188 177 174
1 0.64 (0.36, 1.12) 0.67 (0.38, 1.18) 529
188 176 175
1 0.73 (0.44, 1.22) 0.88 (0.53, 1.46) 409
188 176 175
1 0.67 (0.36, 1.27) 0.86 (0.47, 1.60) 389
Valenciaa
o 175 175–179 4 179 N model
268 199 113
1 1.32 (0.81, 2.15) 1.27 (0.71, 2.27) 575
262 203 115
1 1.26 (0.84, 1.90) 1.57 (0.98, 2.52) 578
262 203 115
1 1.01 (0.63, 1.62) 1.27 (0.73, 2.19) 578
Models adjusted by cohort, sex, and maternal asthma and education. a
Exposure assessment in Sabadell (S) was based on less CaCO3 measurements and this cohort was excluded from the cohort-stratified analyses.
Table 4 Association between duration of bathing and showering and eczema at 14 m and at 4 y in the three INMA birth cohorts. Eczema ever at 14 m Eczema ever at 4 y Current eczema at 4 y OR (95%CI) OR (95%CI) OR (95%CI) Bathing first months (min/week) o31 1 1 31–35 0.98 (0.71, 1.34) 0.95 (0.72, 1.25) 435 0.76 (0.53, 1.09) 0.85 (0.63, 1.15) N model 1558 1416
1 0.85 (0.61, 1.17) 0.96 (0.68, 1.34) 1389
Bathing 14 m (min/week) o46 1 46–80 0.56 (0.41, 0.78) 480 0.49 (0.35, 0.69) N model 1526
1 0.70 (0.53, 0.93) 0.64 (0.48, 0.86) 1388
1 0.80 (0.58, 1.10) 0.77 (0.56, 1.08) 1361
Bathing/Showering 4 y (min/week) o60 – 1 60–104 0.69 (0.52, 0.92) 4104 0.62 (0.47, 0.82) N model 1420
1 0.72 (0.51, 1.00) 0.71 (0.52, 0.97) 1422
Models adjusted by cohort, sex, maternal asthma and maternal education.
both cumulative incidence and current prevalence at 14 m and at 4 y and distinguish between earlier and later-onset cases, but we did not find an association between any of these eczema indicators and water hardness. The Belgian study among 5–6 y old children found a positive association between water hardness and eczema especially among atopic children (Chaumont et al., 2012). According to the natural history of the disease, eczema usually initiates in early infancy in a nonatopic form due to the impairment of the epidermal barrier (Bieber, 2008). In 60–80% of patients, sensitization to allergens will occur later on leading to a true atopic dermatitis (Bieber, 2008) and facilitating the development of other atopic conditions (Bieber, 2008; Spergel and Paller, 2003). In this scheme, we hypothesize that if water hardness would disrupt the skin barrier
and increase eczema risk during the first year of life it would do so before (or independently) of an allergic sensitization status. We only had IgE measured at 4 y in one of the cohorts, with a prevalence of sensitization of 7% (Gascon et al., 2014). Unfortunately, this small sample size did not allow us to evaluate sensitized and non-sensitized eczema. We stratified by parental atopy as a surrogate and did not find a different association between water hardness and eczema among the two groups. Outcome misclassification is a potential concern, given that eczema was self reported. A clinical visit is not always feasible in large epidemiological studies, and we used a question from the validated ISAAC questionnaire (Ellwood, 2005). This is the most used world-wide for epidemiological studies on atopic diseases in childhood, with similar questions as the previous studies on the topic (McNally et al., 1998; Miyake et al., 2004; Arnedo-Pena et al., 2007; Chaumont et al., 2012), which allows to compare with previous studies. Misclassification can also be a concern for other reported variables, such as bath duration, as is common in epidemiological studies. The inclusion of three birth cohorts in different areas in Spain allowed us to have a wide range of water hardness (from 47.5 mg/L to 272.8 mg/L of CaCO3). Sabadell and Valencia are in the Mediterranean coast where water is hard to very hard, while Gipuzkoa is in the Northern Atlantic coast with softer water. The source of water at home was municipal in 95% of children and only 0.7% reported having decalcifier filters at home. The level of water hardness is comparable to that reported in the study in the UK (118 to 314 mg/L of CaCO3) (McNally et al., 1998), much higher than the study in Japan ( o48 to 76 mg/L) (Miyake et al., 2004), but lower than in the Belgian (o150 to 4350 mg/L) (Chaumont et al., 2012) and Spanish (o200 to 4300 mg/L) (Arnedo-Pena et al., 2007) study. The birth cohort design allowed a prospective measurement of exposures, outcomes and several covariables, having a lower risk of measurement error, bias and reverse causation. Changes over time, for example in home address that affected 17.4% of children, could be considered. The exposure assessment is another improvement of the current study. CaCO3 data was available for 55 different water zones during the study period (2006–2012). It
584
L. Font-Ribera et al. / Environmental Research 142 (2015) 579–585
Table 5 Association betwCen eczema at 14 m and 4 y in the Sabadell (S), Gipuzkoa (G), and Valencia (V) cohorts and a score combining water hardness at home (CaCO3 mg/L) and bath exposure (min/week). G, Va OR (95%CI)
Ga OR (95%CI)
Va OR (95%CI)
Eczema ever at 14 m – Exposure at first months (CaCO3 mg/L*min/week bathing) o3768 1 3768–7512 0.75 (0.51, 1.11) 47512 0.59 (0.39, 0.88) N model 1565
1 0.79 (0.50, 1.26) 0.63 (0.38, 1.06) 1094
1 1.08 (0.59, 1.98) 0.62 (0.23, 1.64) 519
1 0.53 (0.26, 1.08) 0.47 (0.24, 0.95) 575
Current eczema at 4 y – Exposure at 14 m (CaCO3 mg/L*min/week bathing) o7140 1 7140–15,435 1.07 (0.74, 1.54) 415,435 0.89 (0.61, 1.31) N model 1394
1 1.07 (0.70, 1.63) 1.08 (0.68, 1.71) 952
1 1.06 (0.57, 1.95) 1.23 (0.53, 2.90) 374
1 1.06 (0.58, 1.93) 1.04 (0.57, 1.88) 578
Eczema ever at 4 y – Exposure at 14 m (CaCO3 mg/L*min/week bathing) o7140 1 7140–15,435 0.84 (0.62, 1.14) 415,435 0.72 (0.52, 0.99) N model 1421
1 0.81 (0.57, 1.14) 0.82 (0.56, 1.20) 971
1 0.77 (0.46, 1.29) 0.94 (0.45, 1.94) 393
1 0.81 (0.49, 1.33) 0.79 (0.48, 1.29) 578
All cohorts (G, V, S) OR (95%CI)
Models adjusted by cohort, sex, maternal asthma and maternal education. a
Exposure assessment in Sabadell (S) was based on less CaCO3 measurements and this cohort was excluded from the cohort-stratified analyses.
allowed having variability between and within cohorts and checking for temporal trends. Since water hardness is not regulated and therefore not compulsory to register, the available CaCO3 data varied between cohorts. Valencia and Gipuzkoa had the richest data and stratified analysis with cohort-specific tertiles could be performed. The new CaCO3 measurements conducted in Sabadell, the city with more participants, found no seasonal variability in water hardness but also showed that there was geographical variation in water hardness within the water areas, indicating some exposure misclassification in this cohort. However, the exclusion of Sabadell from the analysis did not modify the association between water hardness and eczema. As previous studies, water hardness was measured with CaCO3, the most standard parameter, but levels of other potential irritant indicators such as chlorine or pH in water were not measured. This is the first study to consider individual water contact through showering and bathing. Bath duration alone or in combination with the level of water hardness at home was not related to eczema in the next follow-up, but a negative association was detected in cross-sectional analyses (i.e. water hardness and eczema measured at the same follow-up). This probably indicates reverse causation, suggesting that parents bath their children less often or shorter if children have eczema. It has been suggested that is not water hardess per se but an increased soap use with harder water what could be the responsible for increasing the pH, irritating the skin, and leading to an impairment of the epidermal barrier, inflammation and eczema (Cork et al., 2006; Thomas et al., 2011). However, hard water may have a higher pH itself what could affect the skin pH. The skin has an acidic pH that contributes to the optimal barrier function and some skin diseases such as eczema have been found to have a higher pH (Cork et al., 2006). We did not have information on soap use (nor had the previous studies), and this is something that would be useful to include in future studies. We have shown that there is considerable variability in bath exposure even among preschoolers and that this is socioeconomically patterned. It is worth mentioning that, besides indicating longer water contact, bathing exposure may also reflect higher hygenic practices that in turn could increase the risk of eczema, following the “hygene hypothesis”. This has been suggested in the ALSPAC birth cohort, where high levels of hygiene at 15 m of age were positively associated with atopic eczema between 30 and 42 m (Sheriff et al., 2002). Swimming in pools may
be another route of skin exposure to water that was associated with childhood eczema in a study (Font-Ribera et al., 2009) but not in others (Font-Ribera et al., 2011, 2013, 2014 Jacobs et al., 2012). Fifty-three percent of this population had attended a swimming pool at 14 m and 95% did it at 4 y of age, but controlling or stratifying for this exposure did not affect the results.
5. Conclusions This longitudinal study with improved exposure assessment did not detect an association between water hardness at home and bathing exposure with eczema during the first 4 y of life in the Spanish INMA birth cohort. The results from a randomized control trial (Thomas et al., 2011) also did not support this hypothesis previously suggested by two ecologic and two cross-sectional studies. Therefore, it appears to be a controversial topic that should be assessed in future longitudinal studies that measure not only water hardness at home, but also water pH, and individual variables such as bathing, soap use and other hygienic practices. This is a relevant topic for public health since the evaluated exposure is modifiable and eczema is very prevalent and is often the first step to other atopic conditions (Bieber, 2008).
Acknowledgments The authors would particularly like to thank all the participants for their generous collaboration. We thank the City Councils of the study areas for facilitating the data on water hardness, José Antonio Bataller from Health Department in Valencia for providing information on water characteristics and Anna Navarro, Lourdes Arjona and Megan O’Driscoll for participating in the sampling and measurement of water hardness in Sabadell. A full roster of the INMA Project Investigators can be found at http://www.proyectoinma.org/presentacion-inma/listado-in vestigadores/en_listado-investigadores.html. This study was funded by Grants from UE (FP7-ENV.2012.6.4-3 no 308610, FP7-ENV2011 cod 282957 and HEALTH.2010.2.4.5-1), Instituto de Salud Carlos III (Red INMA G03/176, CB06/02/0041, FIS-PI06/0867, FISPS09/00090, FIS-FEDER 03/1615, 04/1509, 04/1112, 04/1931, 05/ 1079, 05/1052, 06/1213, 07/0314, 09/02647, 11/0178, 11/01007, 11/
L. Font-Ribera et al. / Environmental Research 142 (2015) 579–585
02591, 11/02038, 13/1944, 13/2032, 14/00891, and 14/01687), Spanish Ministry of Health (FIS-PI041436, FIS-PI081151), Generalitat de Catalunya-CIRIT 1999SGR 00241, Department of Health of the Basque Government (2005111093 and 2009111069), the Provincial Government of Gipuzkoa (DFG06/004 and DFG08/001) and the Conselleria de Sanitat, Generalitat Valenciana. The study was approved by the ethical committees of the participating centers, and all subjects gave written consent at enrollment and delivery.
Appendix A. Supplementary materials Supplementary data associated with this article can be found in the online version at doi:10.1016/j.envres.2015.07.013.
References Arnedo-Pena, A., Bellido-Blasco, J., Puig-Barbera, J., Artero-Civera, A., CamposCruanes, J.B., Pac-Sa, M.R., et al., 2007. Domestic water hardness and prevalence of atopic eczema in Castellon (Spain) school children. Salud Publica Mex 49, 295–301. Bieber, T., 2008. Atopic dermatitis. N. Engl. J. Med. 358, 1483–1494. Chaumont, A., Voisin, C., Sardella, A., Bernard, A., 2012. Interactions between domestic water hardness, infant swimming and atopy in the development of childhood eczema. Environ. Res. 116, 52–57. Cork, M.J., Robinson, D.A., Vasilopoulos, Y., Ferguson, A., Moustafa, M., MacGowan, A., Duff, G.W., Ward, S.J., Tazi-Ahnini, R., 2006. New perspectives on epidermal barrier dysfunction in atopic dermatitis: Gene–environment interactions. J. Allergy Clin. Immunol. 118, 3–21. Eder, W., Ege, M.J., von Mutius, E., 2006. The asthma epidemic. N. Engl. J. Med. 355, 2226–2235. Ellwood, P., Asher, M.I., Beasley, R., Clayton, T.O., Stewart, A.W., ISAAC Steering Committee, 2005. The international study of asthma and allergies in childhood (ISAAC): phase three rationale and methods. Int. J. Tuberc. Lung Dis. 9, 10–16. Flohr, C., Mann, J., 2014. New insights into the epidemiology of childhood atopic dermatitis. Allergy 69, 3–16. Font-Ribera, L., Kogevinas, M., Zock, J.P., Nieuwenhuijsen, M.J., Heederik, D.,
585
Villanueva, C.M., 2009. Swimming pool attendance and risk of asthma and allergic symptoms in children. Eur. Respir. J. 34, 1304–1310. Font-Ribera, L., Villanueva, C.M., Ballester, F., Santa, M.L., Tardon, A., Espejo-Herrera, N., et al., 2013. Swimming pool attendance, respiratory symptoms and infections in the first year of life. Eur. J. Pediatr. 172, 977–985. Font-Ribera, L., Villanueva, C.M., Gracia-Lavedan, E., Borras-Santos, A., Kogevinas, M., Zock, J.P., 2014. Indoor swimming pool attendance and respiratory and dermal health in schoolchildren – HITEA Catalonia. Respir. Med. 108, 1056–1059. Font-Ribera, L., Villanueva, C.M., Nieuwenhuijsen, M.J., Zock, J.P., Kogevinas, M., Henderson, J., 2011. Swimming pool attendance, asthma, allergies, and lung function in the Avon Longitudinal Study of Parents and Children cohort. Am. J. Respir. Crit. Care Med. 183, 582–588. Garg, N., Silverberg, J.I., 2015. Epidemiology of childhood atopic dermatitis. Clin. Dermatol. 33, 281–288. Gascon, M., Casas, M., Morales, E., Valvi, D., Ballesteros-Gomez, A., Luque, N., et al., 2015. Prenatal exposure to bisphenol A and phthalates and childhood respiratory tract infections and allergy. J. Allergy Clin. Immunol. 135, 370–378. Guxens, M., Ballester, F., Espada, M., Fernandez, M.F., Grimalt, J.O., Ibarluzea, J., et al., 2012. Cohort profile: the INMA – INfancia y Medio Ambiente – (Environment and Childhood) project. Int. J. Epidemiol. 41, 930–940. Jacobs, J.H., Fuertes, E., Krop, E.J., Spithoven, J., Tromp, P., Heederik, D.J., 2012. Swimming pool attendance and respiratory symptoms and allergies among Dutch children. Occup. Environ. Med. 69, 823–830. McNally, N.J., Williams, H.C., Phillips, D.R., Smallman-Raynor, M., Lewis, S., Venn, A., et al., 1998. Atopic eczema and domestic water hardness. Lancet 352, 527–531. Miyake, Y., Yokoyama, T., Yura, A., Iki, M., Shimizu, T., 2004. Ecological association of water hardness with prevalence of childhood atopic dermatitis in a Japanese urban area. Environ. Res. 94, 33–37. Sheriff, A., Goldin, J., 2002. Alspac study team. Hygiene levels in a contemporary population cohort are associated with wheezing and atopic eczema in preschool infants. Arch. Dis. Child 87, 26–29. Spergel, J.M., Paller, A.S., 2003. Atopic dermatitis and the atopic march. J. Allergy Clin. Immunol. 112, S118–S127. Thomas, K.S., Dean, T., O’Leary, C., Sach, T.H., Koller, K., Frost, A., et al., 2011. A randomised controlled trial of ion-exchange water softeners for the treatment of eczema in children. PLoS Med. 8, e1000395. World Health Organization, 2009. Calcium and Magnesium in Drinking-water. Public health significance. Available at: 〈http://whqlibdoc.who.int/publications/ 2009/9789241563550_eng.pdf〉 (last accessed 18.06.15). World Health Organization. 2011. Hardness in Drinking-water. Background document for development of WHO Guidelines for Drinking-water Quality. Available at: 〈http://www.who.int/water_sanitation_health/dwq/chemicals/hardness. pdf〉 (last accessed 1815).
Contents lists available at ScienceDirect
Environmental Research journal homepage: www.elsevier.com/locate/envres
Water hardness and eczema at 1 and 4 y of age in the INMA birth cohort Laia Font-Ribera a,b,c,d,n, Esther Gracia-Lavedan a,b,c, Ana Esplugues d,e, Ferran Ballester d,e, Ana Jiménez Zabala f,g, Loreto Santa Marina d,f,g, Ana Fernández-Somoano d,h, Jordi Sunyer a,b,c,d, Cristina M. Villanueva a,b,c,d a
Centre for Research in Environmental Epidemiology (CREAL), Barcelona Biomedical Research Park, 88 Doctor Aiguader, Barcelona 08003, Spain Universitat Pompeu Fabra (UPF), 80 Doctor Aiguader, Barcelona 08003, Spain c IMIM (Hospital del Mar Medical Research Institute), 88 Doctor Aiguader, Barcelona 08003, Spain d CIBER Epidemiologia y Salud Pública (CIBERESP), 3-5 Melchor Fernández Almagro, Madrid 28029, Spain e FISABIO-UJI-Universitat de València Joint Research Unit, Av. Catalunya 21, Valencia 46020, Spain f Public Health Division of Gipuzkoa, Basque Government, 4 Av. de Navarra, San Sebastian 20013, Spain g BIODONOSTIA Health Research Institute, Paseo Dr. Beguiristain, San Sebastian 20014, Spain h Universidad de Oviedo, Campus del Cristo, Oviedo, Spain b
art ic l e i nf o
a b s t r a c t
Article history: Received 18 June 2015 Received in revised form 17 July 2015 Accepted 18 July 2015
Background: Exposure to hard water has been suggested as a risk factor for eczema in childhood, based on limited evidence from two ecologic and two cross-sectional studies. Objectives: We evaluate this hypothesis for the first time in early infancy using prospective data from a mother–child cohort study. Methods: We used data from the INMA cohorts in Gipuzkoa, Sabadell and Valencia, Spain (N¼ 1638). Current and ever eczema, bathing frequency and duration and covariables were collected by questionnaires at 14 months (14 m) and 4 years (4 y). Calcium carbonate (CaCO3) level in municipal water was assigned to home addresses at birth, 14 m and 4 y. We calculated Odds Ratio (OR) of eczema related to CaCO3 at home, bath exposure and a combination of both. Results: Prevalence of eczema ever was 18.4% at 14 m and 33.4% at 4 y. Mean CaCO3 ranged from 51.6 to 272.8 mg/L among areas. No association was detected between water hardness at home and current or ever eczema. Adjusted OR was 0.79 (95%CI ¼ 0.45, 1.39) at 14 m and 0.93 (0.56, 1.52) at 4 y among children in the highest vs. lowest tertiles of CaCO3. Bath exposure alone or in combination with water hardness did not increase the OR of eczema at 14 m or 4 y either. Conclusions: We did not find an association between eczema and water hardness at home or bathing exposure during the first four years of life. This first cohort study in a critical age period with improved exposure assessment does not confirm the association suggested among children by previous studies. & 2015 Elsevier Inc. All rights reserved.
Keywords: Calcium carbonate Bathing Atopic dermatitis Early infancy
1. Introduction The prevalence of allergy-related diseases such as eczema, rhinitis and asthma has reached epidemic proportions in highn Corresponding author at: Centre for Research in Environmental Epidemiology (CREAL), Barcelona Biomedical Research Park, 88 Doctor Aiguader, Barcelona 08003, Spain. E-mail addresses: [email protected] (L. Font-Ribera), [email protected] (E. Gracia-Lavedan), [email protected] (A. Esplugues), [email protected] (F. Ballester), [email protected] (A.o.s. Jiménez Zabala), [email protected] (L.o.s. Santa Marina), [email protected] (A. Fernández-Somoano), [email protected] (J. Sunyer), [email protected] (C.M. Villanueva).
http://dx.doi.org/10.1016/j.envres.2015.07.013 0013-9351/& 2015 Elsevier Inc. All rights reserved.
income countries (Eder et al., 2006). Eczema (or atopic dermatitis) is a chronic inflammatory skin disease currently affecting 15–30% of children characterized by a persistent itching (Bieber, 2008). It can seriously impair the quality of life of children and their parents and has a substantial cost for families and health providers (Garg and Silverberg, 2015). Eczema frequently starts early in infancy and is often the prelude to the development of other atopic diseases during childhood (Spergel and Paller, 2003). In this context, understanding and controlling risk factors for the development and progression of eczema during early life will have an important public health impact. Among several other environmental factors, water hardness has been suggested as a risk factor for childhood eczema (Flohrs et al., 2014). Water hardness is caused by a variety of dissolved
580
L. Font-Ribera et al. / Environmental Research 142 (2015) 579–585
polyvalent metallic ions, predominantly calcium and magnesium and is most commonly expressed as calcium carbonate (CaCO3) (World Health Organization, 2011). Water hardness is supposed to be innocuous for health and is not regulated in drinking water, although epidemiological evidence suggests an inverse association between magnesium levels in drinking water and cardiovascular mortality (World Health Organization, 2011; World Health Organization, 2009). Previous evidence on exposure to water hardness and eczema in children is limited to four studies, two ecological studies in UK (McNally et al., 1998) and Japan (Miyake et al., 2004) and two cross-sectional studies in Spain (Arnedo-Pena et al., 2007) and Belgium (Chaumont et al., 2012). These studies found a positive association between residential water hardness and eczema among specific age groups of school-aged children. However, a randomized control trial failed to show an improvement in eczema severity in children 0–16 y of age after the installation of a water softener at home (Thomas et al., 2011). We aim to test this hypothesis for the first time in early infancy with a child birth cohort, considering the individual variation in activities that affect dermal exposure to water, such as the frequency and duration of showering and bathing. We aimed to evaluate the association between water hardness at home, bathing exposure and eczema at 14 m and at 4 y of age in three cohorts of the INMA Spanish birth cohort.
2. Material and methods 2.1. Study design and population A mother–child cohort study was set up in different Spanish areas following a common protocol to constitute the INMA [Infancia y Medio Ambiente (Childhood and the Environment)] Project (Guxens et al., 2012). The present analysis includes the Gipuzkoa, Sabadell, and Valencia cohorts for which water hardness data could be collected. Study subjects were recruited at the 12th week of gestation and eligibility criteria for enrollment were maternal age Z16 y, singleton pregnancy, planning to deliver at the study hospitals, being able to communicate in either of the official languages, and not having followed an assisted reproduction program. The study sample was representative of the target population in terms of prenatal care attendance in the public health system (used by 480% of the pregnant women); however, the educational level was higher than the target population average. From 45% to 68% of the eligible pregnant women agreed to participate, and enrollment periods ranged from November 2003 in Valencia to February 2008 in Gipuzkoa, including 638 in Gipuzkoa, 657 in Sabadell and 827 in Valencia (total 2122). From the initial sample at recruitment, 2019 women (95.2%) were followed until delivery (Guxens et al., 2012). A follow up was conducted at 12– 14 m after delivery in Gipuzkoa (N ¼ 575; 90.1%), Sabadell (582; 88.6%) and Valencia (708; 85.6%). This will be referred as the 14 months (14 m) follow up. Another follow up was conducted at 4 years of age (4 y), with 1596 children included (85.6% of those at 14 m). Study period included years from birth until the 4-y follow up, i.e. 2004–2012 in Valencia and Sabadell, and 2006–2012 in Gipuzkoa. The study was approved by the ethical committees of the participating centers, and all subjects gave written consent at enrollment and delivery. 2.2. Questionnaire data Fig. 1 illustrates data collected in the different follow-ups. Personal information was gathered through parental administered questionnaires in face-to-face interviews. Eczema was assessed (through questionnaire) at 14 m with the question “Has your child
Fig. 1. Scheme of the study design, INMA cohorts (2004–2012). Data was collected during pregnancy, at birth, at 14 m and at 4 y. Exposure variables are indicated in white squares (bathing frequency and duration, CaCO3 at home and a combined score). Outcome variables are shown in dark gray squares (eczema ever at 14 m and at 4 y and current eczema at 4 y). Bathing FM¼ bathing during the first months of life (ascertained at 14 m follow-up).
ever suffered from atopic eczema?”, from the validated ISAAC questionnaire (Ellwood, 2005). The questions “Has your child suffered from atopic eczema during the last 12 months?” and “Has your child used eczema medication during the last 12 months?” were asked at 4 y. “Current eczema” at 4 y was defined as having had eczema or taken medication in the last 12 m. A positive answer in any of the three questions classified children as “Ever having had eczema” at 4 y. The address of residence was collected at pregnancy and changes of residence were ascertained in each followup. The frequency and duration of baths and showers was collected at 14 m and 4 y. The frequency and duration during the first months of life was also asked in the 14 m questionnaire. Weekly minutes bathing or showering was calculated for the first months of life, 14 m and 4 y and was categorized in tertiles. Maternal asthma and eczema (yes, no), education (primary or less, secondary, high), ethnicity (white, other) and age were ascertained in the interview at recruitment. Birth date and sex were collected at delivery. Day care attendance (yes, no) was collected at 14 m questionnaire. Maternal smoking (yes, no), pet contact (yes, no), use of moisturizer (yes, no), swimming pool attendance in the last 12 m (yes, no) was collected at 14 m and 4 y questionnaires. 2.3. Environmental water hardness levels Concentration of calcium carbonate (CaCO3) in municipal water was ascertained in each cohort from local authorities and water companies. CaCO3 measurements were available from 2006 to 2012 (See Supplemental Material, Table S1). Some municipalities had CaCO3 data for different areas within the municipality and geographical variability in the levels within these municipalities was assessed. Accordingly, a water area was assigned to each municipality, with the exception of six municipalities that had two or more water areas, resulting in 55 water areas. Changes in CaCO3 levels over the study period were evaluated in each water area and a statistically significant trend was not detected. Seasonal variability over a year was also assessed in the municipality with more study participants (Sabadell) conducting new CaCO3 measurements using a commercial kit (HI 3812 Hardness Test Kit, HANNA Instruments). Results indicated no significant trends within a year. Accordingly, a single value for the whole study period was assigned to each water area, as the median value of all
L. Font-Ribera et al. / Environmental Research 142 (2015) 579–585
581
measurements collected. For details on temporal and seasonal variability assessment see Supplemental Material, page 2. In total, 450 CaCO3 measurements were used from 45 municipalities, where lived 90.1% of the study population based on address at birth (see Supplemental Material, Table S1). 2.4. Statistical analysis A water area at birth, 14 m and 4 y was assigned to each participant according to the address of residence (Fig. 1). The mean CaCO3 levels between the water area at birth and at 14 m was calculated and referred as “CaCO3 (mg/L) from 0 to 14 m”. The mean CaCO3 levels in the water areas at birth, at 14 m and at 4 y was calculated and referred as “CaCO3 (mg/L) from 0 to 4 y”. A score combining the CaCO3 levels at birth and the duration of bathing at first months and at 14 m was also calculated. The association between CaCO3 from 0 to 14 m and eczema at 14 m and between CaCO3 from 0 to 4 y and eczema at 4 y was assessed. Generalized additive models (GAM) were used to assess non-linear associations between eczema and CaCO3 as a continuous variable. Logistic regression models were fitted to estimate odds ratios (OR) and 95% confidence intervals (95%CI) for CaCO3, bath duration and the combined score for continuous variables and tertiles. Models were built for eczema at 14 m and 4 y adjusted by cohort, sex, maternal allergy and maternal education. A sensitivity analysis was performed among the two cohorts with more CaCO3 measurements (Gipuzkoa and Valencia) using cohort-specific tertiles. Statistical analyses were performed with the statistical package STATA 12.0.
3. Results From the 1865 children with follow-up at 14 m, 47 were excluded for not having data on eczema at 14 m nor at 4 y, leading to 1818 children. From these, 180 (9.9%) did not have data on CaCO3 at home address, resulting in an analyzed sample of 1638 children at 14 m (See Supplemental material, Table S2). The included children had older mothers, Caucasian ethnicity, higher education, more pets at home, less maternal smoking and the proportion of
Fig. 2. Prevalence of eczema in the three INMA birth cohorts. N ¼1638. Current eczema is defined as having eczema in the last 12 m or taking eczema medication in the last 12 m.
excluded was higher for the Valencia cohort (See Supplemental Material, Table S3). Several characteristics of the population varied between the three cohorts (Table 1). The overall ever eczema prevalence was 18% at 14 m and 33% at 4 y (Fig. 2). Eczema at 14 m was higher among males, those with maternal asthma and using a moisturizer (See Supplemental Material, Table S4). Eczema at 4 y was higher among babies with maternal eczema, born preterm and those attending a day care and using a moisturizer at 14 m. The median CaCO3 levels in the different water areas ranged from 47.5 to 272.8 mg/L and varied between and within cohorts (Fig. 3). The mean CaCO3 levels at participant's home at birth ranged from 119.0 mg/L in Gipuzkoa to 240.9 mg/L in Sabadell (Table 2). The proportion of children at the tertiles of CaCO3 levels at home varied largely between cohorts. Median water contact from bathing was 35 min/week (interquartile range (IQR) ¼20–60; range¼ 0–245) during the first months of life, 70 min/week (IQR ¼40–105; range 0–420) at 14 m and 70 min/week (IQR ¼40– 120; range 6–630) at 4 y. Bath duration varied by cohort and season of birth, and was negatively associated with educational level and related variables such as having pets at home, smoking, ethnicity, day care attendance or maternal age (results not shown). CaCO3 at home was not associated with eczema at 14 m or at 4 y in the linear models (results not shown). Results from the GAM
Table 1 Descriptive of the population characteristics at birth, at 14 months (14 m) and at 4 years follow-up (4 y). N¼ 1638. Gipuzkoa
Sabadell
Valencia
Total
N
%
N
%
N
%
N
%
N
539
32.9
519
31.7
580
35.4
1638
100
Newborn Sex, male Ethnicity, other than white Preterm, yes
267 8 17
49.5 1.5 3.1
268 17 16
51.6 3.3 3.1
305 28 32
52.6 4.8 5.5
840 53 65
51.3 3.3 4.0
0.58 o 0.01 0.06
Maternal characteristics Maternal education, primary or less Maternal age, 4 35 y Maternal asthma, yes Maternal eczema, yes
72 98 31 33
13.4 18.2 5.7 6.1
140 88 42 66
27.1 17.0 8.1 12.7
170 95 43 59
29.3 16.4 7.4 10.2
382 281 116 158
23.4 17.2 7.1 9.7
o 0.001 o 0.001 0.30 o 0.01
Baby characteristics at 14 m and 4 y Age (months) at 14 m follow-up, mean (SD) Pet at home at 14 m, yes Day care attendance at 14 m, yes Maternal smoking at 14 m, yes Swimming at 14 m, yes Moisturizing lotion at 14 m, no Age (years) at 4 y follow-up, mean (SD) Swimming at 4 y, yes
14.5 86 250 96 249 45 4.5 372
(0.7) 16.3 47.3 19.0 47.2 8.6 (0.1) 94.2
14.6 170 151 126 323 16 4.4 444
(0.7) 34.0 30.3 25.7 64.7 3.2 (0.2) 96.1
12.0 273 120 170 275 – 4.3 555
(0.8) 47.1 20.9 29.8 47.4 – (0.2) 96.0
13.6 529 521 392 847 61 4.4 1371
(1.4) 32.9 32.5 25.1 52.7 6.0 (0.2) 95.5
o 0.001 o 0.001 o 0.001 o 0.001 o 0.001 o 0.001 o 0.001 0.30
p-Value for Pearson χ2 test, except age for which an ANOVA was performed.
p-Value
582
L. Font-Ribera et al. / Environmental Research 142 (2015) 579–585
Fig. 3. Water hardness in the study population. CaC03 concentration (mg/L) in the 55 water zones (left) and in the birth address of children in the three cohorts (right).
Table 2 Description of the exposure variables: CaCO3 levels at home, bathing exposure and combined score in the three INMA birth cohorts. N ¼ 1638. Gipuzkoa
Sabadell
Valencia
Total
N
N
%
N
%
N
%
240.9 0 99 4210
(30.1) 0 19.1 80.9
182.7 105 403 73
(19.2) 18.1 69.3 12.6
180.2 553 589 496
(60.8) 33.8 36.0 30.3
CaCO3 from 0 to 4 y (mg/L) o173 448 83.1 173–209.1 88 16.3 4209.1 3 0.6
0 91 428
0 17.5 82.5
106 403 71
18.3 69.5 12.2
554 582 502
33.8 35.5 30.7
Bathing at first months of lifea ≦30 282 54.2 31–35 159 30.6 435 79 15.2
(min/week) 191 38.8 171 34.8 130 26.4
89 217 274
15.3 37.4 47.2
562 547 483
35.3 34.4 30.3
CaCO3 from 0 to 14 m Mean (SD) 119.0 o173 448 173–209.1 88 4209.1 3
% (mg/L) (51.9) 83.1 16.3 0.6
Bathing at 14 m (min/week) ≦45 245 48.0 46–80 173 33.9 480 92 18.0
24.1 35.2 40.7
142 204 233
24.5 35.2 40.2
500 542 233
32.1 34.8 33.1
44.2 27.8 28.0
154 148 277
26.6 25.6 47.8
608 354 471
42.4 24.7 32.9
Combined score at first months (CaCO3 mg/L*min/week bathing) o3768 386 73.8 92 18.5 56 9.6 534 3768–7512 91 17.4 190 38.2 259 44.7 540 47512 46 8.8 215 43.3 265 45.7 526
33.4 33.7 32.9
Showering/bathing at o61 250 60–104 78 4104 65
113 165 191
4 y (min/week) 63.6 204 19.8 128 16.6 129
Combined score at 14 m (CaCO3 mg/L*min/week bathing) o7140 354 67.8 66 13.3 114 19.7 7140– 124 23.8 166 33.4 246 42.4 15,435 415,435 44 8.4 265 53.3 220 37.9
534 536
33.4 33.5
529
33.1
a
Bathing at first months of life was assessed at 14 m questionnaire. All exposure variables were distributed differently among cohorts (all p-values for Pearson χ2 testo 0.001).
models confirmed the lack of association (p-values for nonlinearity 40.05) (See Supplemental Material, Fig. S1). In the categorical analysis (Table 3), water hardness at home was also not associated with eczema ever at 14 m (OR¼ 0.79; 95%CI ¼0.45–1.39 for the highest vs. lowest tertile) or at 4 y (OR ¼0.94 95%CI ¼0.57– 1.53 in the highest vs. lowest tertile). The exclusion of the cohort with less CaCO3 data (Sabadell) or the stratification using specificcohort tertiles in Gipuzkoa and Valencia also led to null
associations. Excluding children who did not have eczema at 14 m also led to null associations between CaCO3 at home from 0 to 4 y and eczema at 4 y in all the cohorts and combined (OR¼ 1.32; 95% CI ¼0.65–2.69, results not shown). Self-reported bathing during the first months of life was not associated with eczema at 14 m or at 4 y (Table 4). However, when bathing and eczema were analyzed cross-sectionally (at the same age), a significant negative association was detected. OR of eczema at 14 m was 0.49 (95%CI¼ 0.35–0.69) among children in the highest vs. lowest tertile of bathing exposure at 14 m. The score combining bathing and CaCO3 levels at home led null associations with current eczema at 4 y, but some negative associations were detected with eczema ever at 14 m and at 4 y (Table 5). Stratified analysis by maternal allergy or swimming pool attendance did not modify the results (results not shown).
4. Discussion Water hardness at home and bathing duration were not related to higher eczema at 14 m and at 4 y of age in the Spanish INMA birth cohorts evaluated. This is the first longitudinal study assessing this hypothesis and the first focusing on the first years of life. The results are in agreement with a randomized controlled trial (Thomas et al., 2011) and in disagreement with two ecologic (McNally et al., 1998; Miyake et al., 2004) and two cross-sectional (Arnedo-Pena et al., 2007; Chaumont et al., 2012) studies suggesting a positive association between eczema and water hardness during childhood. Early childhood is a critical age period, and this is the first study to evaluate eczema risk and water hardness at this age. Around 60% of eczema cases begin within the first year of life and 85% before 5 y of age (Bieber 2008). In our population, 36% of children had ever had eczema at 4 y, and among them, 71% already had it at 14 m of age, highlighting the importance of early life factors in the development of the disease. Previous studies (McNally et al., 1998; Miyake et al., 2004; Arnedo-Pena et al., 2007; Chaumont et al., 2012) were focused on older children, and the associations detected varied across age groups. The studies in UK and Castelló (Spain) found positive associations among children at primary school but not among those at secondary school (McNally et al., 1998; Arnedo-Pena et al., 2007). The Japanese study was conducted among children aged 6–12 y and found a positive association among the oldest but not among the youngest children (Miyake et al., 2004). Previous literature was inconsistent in arguing whether water hardness would increase the risk of new eczema cases (Thomas et al., 2011) or would exacerbate the disease (McNally et al., 1998). In the present study we could estimate
L. Font-Ribera et al. / Environmental Research 142 (2015) 579–585
583
Table 3 Association between eczema at 14 m and 4 y and CaCO3 at home (mg/L) in the Gipuzkoa (G), Valencia (V) and Sabadell (S) cohorts. Eczema ever at 14 m CaCO3 from 0 to 14 m
Eczema ever at 4 y CaCO3 from 0 to 4 y
Current eczema at 4 y CaCO3 from 0 to 4 y
Tertiles CaCO3 (mg/L)
N
OR (95%CI)
N
OR (95%CI)
N
OR (95%CI)
All cohorts (G, V, S)
o 173 173–209 4 209 N model
553 589 496
1 1.00 (0.66, 1.51) 0.79 (0.45, 1.39) 1581
554 582 502
1 0.94 (0.66, 1.35) 0.94 (0.57, 1.53) 1459
554 582 502
1 0.79 (0.51, 1.22) 1.25 (0.71, 2.20) 1428
Gipuzkoa, Valenciaa
o 151 151–176 4 176 N model
373 379 367
1 0.93 (0.63, 1.38) 1.06 (0.72, 1.57) 1104
373 375 371
1 0.94 (0.59, 1.48) 1.08 (0.67, 1.73) 987
373 375 371
1 1.05 (0.60, 1.84) 1.08 (0.61, 1.93) 967
Gipuzkoaa
o 54 54–143 4 143 N model
188 177 174
1 0.64 (0.36, 1.12) 0.67 (0.38, 1.18) 529
188 176 175
1 0.73 (0.44, 1.22) 0.88 (0.53, 1.46) 409
188 176 175
1 0.67 (0.36, 1.27) 0.86 (0.47, 1.60) 389
Valenciaa
o 175 175–179 4 179 N model
268 199 113
1 1.32 (0.81, 2.15) 1.27 (0.71, 2.27) 575
262 203 115
1 1.26 (0.84, 1.90) 1.57 (0.98, 2.52) 578
262 203 115
1 1.01 (0.63, 1.62) 1.27 (0.73, 2.19) 578
Models adjusted by cohort, sex, and maternal asthma and education. a
Exposure assessment in Sabadell (S) was based on less CaCO3 measurements and this cohort was excluded from the cohort-stratified analyses.
Table 4 Association between duration of bathing and showering and eczema at 14 m and at 4 y in the three INMA birth cohorts. Eczema ever at 14 m Eczema ever at 4 y Current eczema at 4 y OR (95%CI) OR (95%CI) OR (95%CI) Bathing first months (min/week) o31 1 1 31–35 0.98 (0.71, 1.34) 0.95 (0.72, 1.25) 435 0.76 (0.53, 1.09) 0.85 (0.63, 1.15) N model 1558 1416
1 0.85 (0.61, 1.17) 0.96 (0.68, 1.34) 1389
Bathing 14 m (min/week) o46 1 46–80 0.56 (0.41, 0.78) 480 0.49 (0.35, 0.69) N model 1526
1 0.70 (0.53, 0.93) 0.64 (0.48, 0.86) 1388
1 0.80 (0.58, 1.10) 0.77 (0.56, 1.08) 1361
Bathing/Showering 4 y (min/week) o60 – 1 60–104 0.69 (0.52, 0.92) 4104 0.62 (0.47, 0.82) N model 1420
1 0.72 (0.51, 1.00) 0.71 (0.52, 0.97) 1422
Models adjusted by cohort, sex, maternal asthma and maternal education.
both cumulative incidence and current prevalence at 14 m and at 4 y and distinguish between earlier and later-onset cases, but we did not find an association between any of these eczema indicators and water hardness. The Belgian study among 5–6 y old children found a positive association between water hardness and eczema especially among atopic children (Chaumont et al., 2012). According to the natural history of the disease, eczema usually initiates in early infancy in a nonatopic form due to the impairment of the epidermal barrier (Bieber, 2008). In 60–80% of patients, sensitization to allergens will occur later on leading to a true atopic dermatitis (Bieber, 2008) and facilitating the development of other atopic conditions (Bieber, 2008; Spergel and Paller, 2003). In this scheme, we hypothesize that if water hardness would disrupt the skin barrier
and increase eczema risk during the first year of life it would do so before (or independently) of an allergic sensitization status. We only had IgE measured at 4 y in one of the cohorts, with a prevalence of sensitization of 7% (Gascon et al., 2014). Unfortunately, this small sample size did not allow us to evaluate sensitized and non-sensitized eczema. We stratified by parental atopy as a surrogate and did not find a different association between water hardness and eczema among the two groups. Outcome misclassification is a potential concern, given that eczema was self reported. A clinical visit is not always feasible in large epidemiological studies, and we used a question from the validated ISAAC questionnaire (Ellwood, 2005). This is the most used world-wide for epidemiological studies on atopic diseases in childhood, with similar questions as the previous studies on the topic (McNally et al., 1998; Miyake et al., 2004; Arnedo-Pena et al., 2007; Chaumont et al., 2012), which allows to compare with previous studies. Misclassification can also be a concern for other reported variables, such as bath duration, as is common in epidemiological studies. The inclusion of three birth cohorts in different areas in Spain allowed us to have a wide range of water hardness (from 47.5 mg/L to 272.8 mg/L of CaCO3). Sabadell and Valencia are in the Mediterranean coast where water is hard to very hard, while Gipuzkoa is in the Northern Atlantic coast with softer water. The source of water at home was municipal in 95% of children and only 0.7% reported having decalcifier filters at home. The level of water hardness is comparable to that reported in the study in the UK (118 to 314 mg/L of CaCO3) (McNally et al., 1998), much higher than the study in Japan ( o48 to 76 mg/L) (Miyake et al., 2004), but lower than in the Belgian (o150 to 4350 mg/L) (Chaumont et al., 2012) and Spanish (o200 to 4300 mg/L) (Arnedo-Pena et al., 2007) study. The birth cohort design allowed a prospective measurement of exposures, outcomes and several covariables, having a lower risk of measurement error, bias and reverse causation. Changes over time, for example in home address that affected 17.4% of children, could be considered. The exposure assessment is another improvement of the current study. CaCO3 data was available for 55 different water zones during the study period (2006–2012). It
584
L. Font-Ribera et al. / Environmental Research 142 (2015) 579–585
Table 5 Association betwCen eczema at 14 m and 4 y in the Sabadell (S), Gipuzkoa (G), and Valencia (V) cohorts and a score combining water hardness at home (CaCO3 mg/L) and bath exposure (min/week). G, Va OR (95%CI)
Ga OR (95%CI)
Va OR (95%CI)
Eczema ever at 14 m – Exposure at first months (CaCO3 mg/L*min/week bathing) o3768 1 3768–7512 0.75 (0.51, 1.11) 47512 0.59 (0.39, 0.88) N model 1565
1 0.79 (0.50, 1.26) 0.63 (0.38, 1.06) 1094
1 1.08 (0.59, 1.98) 0.62 (0.23, 1.64) 519
1 0.53 (0.26, 1.08) 0.47 (0.24, 0.95) 575
Current eczema at 4 y – Exposure at 14 m (CaCO3 mg/L*min/week bathing) o7140 1 7140–15,435 1.07 (0.74, 1.54) 415,435 0.89 (0.61, 1.31) N model 1394
1 1.07 (0.70, 1.63) 1.08 (0.68, 1.71) 952
1 1.06 (0.57, 1.95) 1.23 (0.53, 2.90) 374
1 1.06 (0.58, 1.93) 1.04 (0.57, 1.88) 578
Eczema ever at 4 y – Exposure at 14 m (CaCO3 mg/L*min/week bathing) o7140 1 7140–15,435 0.84 (0.62, 1.14) 415,435 0.72 (0.52, 0.99) N model 1421
1 0.81 (0.57, 1.14) 0.82 (0.56, 1.20) 971
1 0.77 (0.46, 1.29) 0.94 (0.45, 1.94) 393
1 0.81 (0.49, 1.33) 0.79 (0.48, 1.29) 578
All cohorts (G, V, S) OR (95%CI)
Models adjusted by cohort, sex, maternal asthma and maternal education. a
Exposure assessment in Sabadell (S) was based on less CaCO3 measurements and this cohort was excluded from the cohort-stratified analyses.
allowed having variability between and within cohorts and checking for temporal trends. Since water hardness is not regulated and therefore not compulsory to register, the available CaCO3 data varied between cohorts. Valencia and Gipuzkoa had the richest data and stratified analysis with cohort-specific tertiles could be performed. The new CaCO3 measurements conducted in Sabadell, the city with more participants, found no seasonal variability in water hardness but also showed that there was geographical variation in water hardness within the water areas, indicating some exposure misclassification in this cohort. However, the exclusion of Sabadell from the analysis did not modify the association between water hardness and eczema. As previous studies, water hardness was measured with CaCO3, the most standard parameter, but levels of other potential irritant indicators such as chlorine or pH in water were not measured. This is the first study to consider individual water contact through showering and bathing. Bath duration alone or in combination with the level of water hardness at home was not related to eczema in the next follow-up, but a negative association was detected in cross-sectional analyses (i.e. water hardness and eczema measured at the same follow-up). This probably indicates reverse causation, suggesting that parents bath their children less often or shorter if children have eczema. It has been suggested that is not water hardess per se but an increased soap use with harder water what could be the responsible for increasing the pH, irritating the skin, and leading to an impairment of the epidermal barrier, inflammation and eczema (Cork et al., 2006; Thomas et al., 2011). However, hard water may have a higher pH itself what could affect the skin pH. The skin has an acidic pH that contributes to the optimal barrier function and some skin diseases such as eczema have been found to have a higher pH (Cork et al., 2006). We did not have information on soap use (nor had the previous studies), and this is something that would be useful to include in future studies. We have shown that there is considerable variability in bath exposure even among preschoolers and that this is socioeconomically patterned. It is worth mentioning that, besides indicating longer water contact, bathing exposure may also reflect higher hygenic practices that in turn could increase the risk of eczema, following the “hygene hypothesis”. This has been suggested in the ALSPAC birth cohort, where high levels of hygiene at 15 m of age were positively associated with atopic eczema between 30 and 42 m (Sheriff et al., 2002). Swimming in pools may
be another route of skin exposure to water that was associated with childhood eczema in a study (Font-Ribera et al., 2009) but not in others (Font-Ribera et al., 2011, 2013, 2014 Jacobs et al., 2012). Fifty-three percent of this population had attended a swimming pool at 14 m and 95% did it at 4 y of age, but controlling or stratifying for this exposure did not affect the results.
5. Conclusions This longitudinal study with improved exposure assessment did not detect an association between water hardness at home and bathing exposure with eczema during the first 4 y of life in the Spanish INMA birth cohort. The results from a randomized control trial (Thomas et al., 2011) also did not support this hypothesis previously suggested by two ecologic and two cross-sectional studies. Therefore, it appears to be a controversial topic that should be assessed in future longitudinal studies that measure not only water hardness at home, but also water pH, and individual variables such as bathing, soap use and other hygienic practices. This is a relevant topic for public health since the evaluated exposure is modifiable and eczema is very prevalent and is often the first step to other atopic conditions (Bieber, 2008).
Acknowledgments The authors would particularly like to thank all the participants for their generous collaboration. We thank the City Councils of the study areas for facilitating the data on water hardness, José Antonio Bataller from Health Department in Valencia for providing information on water characteristics and Anna Navarro, Lourdes Arjona and Megan O’Driscoll for participating in the sampling and measurement of water hardness in Sabadell. A full roster of the INMA Project Investigators can be found at http://www.proyectoinma.org/presentacion-inma/listado-in vestigadores/en_listado-investigadores.html. This study was funded by Grants from UE (FP7-ENV.2012.6.4-3 no 308610, FP7-ENV2011 cod 282957 and HEALTH.2010.2.4.5-1), Instituto de Salud Carlos III (Red INMA G03/176, CB06/02/0041, FIS-PI06/0867, FISPS09/00090, FIS-FEDER 03/1615, 04/1509, 04/1112, 04/1931, 05/ 1079, 05/1052, 06/1213, 07/0314, 09/02647, 11/0178, 11/01007, 11/
L. Font-Ribera et al. / Environmental Research 142 (2015) 579–585
02591, 11/02038, 13/1944, 13/2032, 14/00891, and 14/01687), Spanish Ministry of Health (FIS-PI041436, FIS-PI081151), Generalitat de Catalunya-CIRIT 1999SGR 00241, Department of Health of the Basque Government (2005111093 and 2009111069), the Provincial Government of Gipuzkoa (DFG06/004 and DFG08/001) and the Conselleria de Sanitat, Generalitat Valenciana. The study was approved by the ethical committees of the participating centers, and all subjects gave written consent at enrollment and delivery.
Appendix A. Supplementary materials Supplementary data associated with this article can be found in the online version at doi:10.1016/j.envres.2015.07.013.
References Arnedo-Pena, A., Bellido-Blasco, J., Puig-Barbera, J., Artero-Civera, A., CamposCruanes, J.B., Pac-Sa, M.R., et al., 2007. Domestic water hardness and prevalence of atopic eczema in Castellon (Spain) school children. Salud Publica Mex 49, 295–301. Bieber, T., 2008. Atopic dermatitis. N. Engl. J. Med. 358, 1483–1494. Chaumont, A., Voisin, C., Sardella, A., Bernard, A., 2012. Interactions between domestic water hardness, infant swimming and atopy in the development of childhood eczema. Environ. Res. 116, 52–57. Cork, M.J., Robinson, D.A., Vasilopoulos, Y., Ferguson, A., Moustafa, M., MacGowan, A., Duff, G.W., Ward, S.J., Tazi-Ahnini, R., 2006. New perspectives on epidermal barrier dysfunction in atopic dermatitis: Gene–environment interactions. J. Allergy Clin. Immunol. 118, 3–21. Eder, W., Ege, M.J., von Mutius, E., 2006. The asthma epidemic. N. Engl. J. Med. 355, 2226–2235. Ellwood, P., Asher, M.I., Beasley, R., Clayton, T.O., Stewart, A.W., ISAAC Steering Committee, 2005. The international study of asthma and allergies in childhood (ISAAC): phase three rationale and methods. Int. J. Tuberc. Lung Dis. 9, 10–16. Flohr, C., Mann, J., 2014. New insights into the epidemiology of childhood atopic dermatitis. Allergy 69, 3–16. Font-Ribera, L., Kogevinas, M., Zock, J.P., Nieuwenhuijsen, M.J., Heederik, D.,
585
Villanueva, C.M., 2009. Swimming pool attendance and risk of asthma and allergic symptoms in children. Eur. Respir. J. 34, 1304–1310. Font-Ribera, L., Villanueva, C.M., Ballester, F., Santa, M.L., Tardon, A., Espejo-Herrera, N., et al., 2013. Swimming pool attendance, respiratory symptoms and infections in the first year of life. Eur. J. Pediatr. 172, 977–985. Font-Ribera, L., Villanueva, C.M., Gracia-Lavedan, E., Borras-Santos, A., Kogevinas, M., Zock, J.P., 2014. Indoor swimming pool attendance and respiratory and dermal health in schoolchildren – HITEA Catalonia. Respir. Med. 108, 1056–1059. Font-Ribera, L., Villanueva, C.M., Nieuwenhuijsen, M.J., Zock, J.P., Kogevinas, M., Henderson, J., 2011. Swimming pool attendance, asthma, allergies, and lung function in the Avon Longitudinal Study of Parents and Children cohort. Am. J. Respir. Crit. Care Med. 183, 582–588. Garg, N., Silverberg, J.I., 2015. Epidemiology of childhood atopic dermatitis. Clin. Dermatol. 33, 281–288. Gascon, M., Casas, M., Morales, E., Valvi, D., Ballesteros-Gomez, A., Luque, N., et al., 2015. Prenatal exposure to bisphenol A and phthalates and childhood respiratory tract infections and allergy. J. Allergy Clin. Immunol. 135, 370–378. Guxens, M., Ballester, F., Espada, M., Fernandez, M.F., Grimalt, J.O., Ibarluzea, J., et al., 2012. Cohort profile: the INMA – INfancia y Medio Ambiente – (Environment and Childhood) project. Int. J. Epidemiol. 41, 930–940. Jacobs, J.H., Fuertes, E., Krop, E.J., Spithoven, J., Tromp, P., Heederik, D.J., 2012. Swimming pool attendance and respiratory symptoms and allergies among Dutch children. Occup. Environ. Med. 69, 823–830. McNally, N.J., Williams, H.C., Phillips, D.R., Smallman-Raynor, M., Lewis, S., Venn, A., et al., 1998. Atopic eczema and domestic water hardness. Lancet 352, 527–531. Miyake, Y., Yokoyama, T., Yura, A., Iki, M., Shimizu, T., 2004. Ecological association of water hardness with prevalence of childhood atopic dermatitis in a Japanese urban area. Environ. Res. 94, 33–37. Sheriff, A., Goldin, J., 2002. Alspac study team. Hygiene levels in a contemporary population cohort are associated with wheezing and atopic eczema in preschool infants. Arch. Dis. Child 87, 26–29. Spergel, J.M., Paller, A.S., 2003. Atopic dermatitis and the atopic march. J. Allergy Clin. Immunol. 112, S118–S127. Thomas, K.S., Dean, T., O’Leary, C., Sach, T.H., Koller, K., Frost, A., et al., 2011. A randomised controlled trial of ion-exchange water softeners for the treatment of eczema in children. PLoS Med. 8, e1000395. World Health Organization, 2009. Calcium and Magnesium in Drinking-water. Public health significance. Available at: 〈http://whqlibdoc.who.int/publications/ 2009/9789241563550_eng.pdf〉 (last accessed 18.06.15). World Health Organization. 2011. Hardness in Drinking-water. Background document for development of WHO Guidelines for Drinking-water Quality. Available at: 〈http://www.who.int/water_sanitation_health/dwq/chemicals/hardness. pdf〉 (last accessed 1815).