Eczema in early childhood, often the start of the “atopic march”, increases the likelihood of developing subsequent atopic conditions (food allergy, allergic asthma, and allergic rhinitis)1, 2. Modifiable environmental risk factors, important to eczema pathogenesis, may represent potential interventions. Studies demonstrate protective effects of domestic pets on eczema development3, 4. Previous reports focus on eczema in early childhood but do not address potential effects in late childhood3, 4. This study compares the associations of prenatal and first-year dog exposure to eczema in early childhood (age 2 years) and late childhood (age 10 years) in the Wayne County Health, Environment, Allergy and Asthma Longitudinal study (WHEALS), a racially and socioeconomically diverse birth cohort. We also assessed whether prenatal dog exposure is associated with persistence or resolution of eczema from age 2 to 10 years and whether dogs are associated with atopic and non-atopic eczema.
WHEALS enrolled pregnant women (21–49 years) due September 2003-December 2007 residing in the metro-Detroit area. Participants provided informed consent, and study protocols were approved by the Henry Ford Health (HFH) Institutional Review Board5. Our analysis included maternal-child pairs who completed a prenatal interview and had eczema history evaluated by a physician during study clinic visits at age 2 and/or 10 years (N=794; Figure E1). Using patient history and physical examination, physicians assessed each child’s history of eczema over time. “Early eczema only” refers to eczema diagnosis prior to or at age 2, but no eczema at age 10. “Late eczema only” refers to current eczema diagnosis at age 10, but not ≤2 years. “Persistent eczema” refers to eczema diagnosis at age ≤2 years with eczema present at age 10; if absent at age 10, the child is classified as having “Resolved eczema”. Prenatal dog exposure was defined as ≥1 indoor dogs in the home for ≥1 hour per day during pregnancy, for at least one week. Median dog exposure was beyond pregnancy at 24 months (IQR=10 to 69 months). Dogs in the first year of life was defined similarly. Serum allergen-specific IgE (sIgE) was measured for 10 common inhalant/food allergens at age 2 (Dermatophagoides farinae, ragweed, Alternaria alternatum, dog, cat, grass, cockroach, egg, milk, and peanut) and 11 allergens at age 10 (dog, cat, cockroach, Dermatophagoides pteronyssinus, Dermatophagoides farinae, ragweed, grass, mold mixture, egg, peanut, and milk). Atopy was defined as sIgE≥0.35 IU/mL to ≥1 allergens. Eczema was further refined as atopic eczema (early atopic eczema: early eczema and atopic at age 2, late atopic eczema: late eczema and atopic at age 10, persistent atopic eczema: persistent eczema and atopic at 2 or 10). Non-atopic eczema was defined similarly.
Logistic regression was used to determine association between dog-keeping and childhood eczema, with main effect p-values<0.05 considered significant. Interaction terms were included to test a priori hypothesized effect modification (maternal eczema, maternal race, mode of delivery, first born child, and prenatal environmental tobacco smoke [ETS] exposure), with interaction p-values <0.10 considered significant. Models were refit after adjusting for potential confounders (a priori hypothesized as associated with dogs and eczema), with confounding indicated by effect size change >20%.
Rate of indoor dog keeping was 26% prenatally, and rates of eczema at ages 2 and 10 were 22% and 21%, respectively (Table 1). Among children with known eczema status at both time points (N=394), 14% had “Early eczema only”, 11% had “Late eczema only”, and 11% had “Persistent eczema” (Table 1). In multivariable analyses, children with prenatal dog exposure had lower odds of “Early eczema only” (aOR[95% CI] = 0.28[0.09, 0.83], p=0.022, Table 2). This association was weaker and not statistically significant for “Late eczema only” (aOR[95% CI] = 1.36[0.63,2.92], p=0.433, Table 2), and did not reach statistical significance for “Persistent eczema” after adjustment (aOR[95% CI] = 0.32[0.10, 1.02], p=0.053, Table 2). We then examined the impact of effect modifiers. Mode of delivery significantly modified the effect of dogs in the first year of life on early eczema development (interaction p=0.056). Specifically, an effect was observed only among vaginally delivered children (OR[95% CI]=0.07 [0.01, 0.54], p=0.011), but not among children delivered via C-section (OR [95% CI]=0.68 [0.22, 2.12], p=0.51); all other interaction p-values were≥0.10.
Table 1:
Descriptive characteristics of children with known AD status at age 2 or age 10 (N=794)
| Variable | Level | N = 794 | % |
|---|---|---|---|
| Mom Eczema (ever in lifetime) | No | 590 | 74.7 |
| Yes | 200 | 25.3 | |
| Maternal Race | White | 199 | 25.1 |
| African American | 471 | 59.3 | |
| Hispanic | 47 | 5.9 | |
| Arabic | 41 | 5.2 | |
| Other/Mixed Race | 36 | 4.5 | |
| Mode of Delivery | Vaginal | 496 | 62.6 |
| C-Section | 296 | 37.4 | |
| First Born Child | No | 495 | 62.3 |
| Yes | 299 | 37.7 | |
| Prenatal Indoor Dog(s) | No | 587 | 73.9 |
| Yes | 207 | 26.1 | |
| Indoor Dog(s) in First Year of Life | No | 444 | 69.6 |
| Yes | 194 | 30.4 | |
| Prenatal Indoor Cat(s) | No | 657 | 82.8 |
| Yes | 137 | 17.2 | |
| Prenatal ETS Exposure | No | 609 | 76.7 |
| Yes | 185 | 23.3 | |
| Eczema Category | Never | 255 | 64.7 |
| Early1 | 54 | 13.7 | |
| Late2 | 43 | 10.9 | |
| Persistent3 | 42 | 10.7 | |
| Atopic at Age 2 | No | 251 | 48.5 |
| Yes | 266 | 51.5 | |
| Total IgE at Age 2 | Geometric Mean (SD) | 21.1 (4.2) | |
| Atopic at Age 10 | No | 226 | 41.9 |
| Yes | 314 | 58.2 | |
| Total IgE at Age 10 | Geometric Mean (SD) | 68.7 (4.9) |
N=21 with early atopic eczema, N=18 with early non-atopic eczema, N=15 with missing atopy status
N=31 late atopic eczema, N=12 late non-atopic eczema
N=34 persistent atopic eczema, N=3 persistent non-atopic eczema, N=5 with missing atopy status
Table 2:
Association between dog exposure and eczema from ages 2 to 10.
| Exposure | Outcome | Unadjusted | Adjusted | ||||
|---|---|---|---|---|---|---|---|
| N | OR (95% CI) | p-value | N | OR (95% CI) | p-value | ||
| Prenatal Dog Exposure | Early Only vs. Never Eczema | 309 | 0.32 (0.14, 0.74) | 0.008 | 240 | 0.28 (0.09, 0.83)1 | 0.022 |
| Early Atopic vs. Never Eczema | 276 | 0.36 (0.1, 1.25) | 0.107 | 220 | 0.48 (0.13, 1.81)1 | 0.277 | |
| Early Non-Atopic vs. Never Eczema | 273 | 0.13 (0.02, 0.97) | 0.046 | 217 | N/A1,2 | N/A1,2 | |
| Late Only vs. Never Eczema | 298 | 0.93 (0.46, 1.88) | 0.842 | 294 | 1.36 (0.63, 2.92)3 | 0.433 | |
| Late Atopic vs. Never Eczema | 286 | 1.02 (0.46, 2.27) | 0.955 | 282 | 2.14 (0.84, 5.46)3 | 0.111 | |
| Late Non-Atopic vs. Never Eczema | 267 | 0.72 (0.19, 2.72) | 0.624 | 263 | 0.47 (0.11, 2.07)3 | 0.321 | |
| Persistent vs. Never Eczema | 297 | 0.36 (0.15, 0.88) | 0.026 | 230 | 0.32 (0.10, 1.02)1 | 0.053 | |
| Persistent Atopic vs. Never Eczema | 289 | 0.46 (0.18, 1.16) | 0.098 | 225 | 0.41 (0.12, 1.37)1 | 0.148 | |
| Persistent Non-Atopic vs. Never Eczema | 258 | N/A2 | N/A2 | 202 | N/A1,2 | N/A1,2 | |
| Resolved Eczema: Yes vs. No | 96 | 0.89 (0.28, 2.89) | 0.851 | 72 | 1.08 (0.21, 5.54)1 | 0.93 | |
| Dog Exposure in the First Year of Life | Early Only vs. Never Eczema | 266 | 0.28 (0.11, 0.69) | 0.006 | 211 | 0.22 (0.06, 0.78)1 | 0.019 |
| Early Atopic vs. Never Eczema | 239 | 0.21 (0.05, 0.94) | 0.042 | 195 | 0.33 (0.07, 1.55)1 | 0.16 | |
| Early Non-Atopic vs. Never Eczema | 234 | 0.14 (0.02, 1.06) | 0.057 | 190 | N/A1,2 | N/A1,2 | |
| Late Only vs. Never Eczema | 257 | 1.00 (0.48, 2.06) | 0.997 | 255 | 1.31 (0.60, 2.87)3 | 0.503 | |
| Late Atopic vs. Never Eczema | 246 | 1.13 (0.49, 2.59) | 0.774 | 244 | 1.95 (0.74, 5.15)3 | 0.177 | |
| Late Non-Atopic vs. Never Eczema | 230 | 0.72 (0.19, 2.79) | 0.635 | 228 | 0.6 (0.15, 2.51)3 | 0.488 | |
| Persistent vs. Never Eczema | 253 | 0.50 (0.21, 1.20) | 0.119 | 200 | 0.45 (0.14, 1.47)1 | 0.187 | |
| Persistent Atopic vs. Never Eczema | 247 | 0.64 (0.26, 1.57) | 0.331 | 196 | 0.61 (0.18, 2.08)1 | 0.428 | |
| Persistent Non-Atopic vs. Never Eczema | 221 | N/A2 | N/A2 | 177 | N/A1,2 | N/A1,2 | |
| Resolved Eczema: Yes vs. No | 81 | 0.56 (0.17, 1.86) | 0.348 | 61 | 0.64 (0.10, 3.98)1 | 0.628 | |
Adjusted for maternal race, mode of delivery, prenatal indoor cats, and log(total IgE at age 2).
Model did not converge.
Adjusted for maternal race, mode of delivery, prenatal indoor cats, and log(total IgE at age 10).
Previous studies reveal associations between prenatal and early life dog exposures with prevention of early childhood eczema3, 4, 6. Our study confirms this, with protective effects on early, but not late childhood, and strongest effect among children delivered vaginally. Though the effect size appeared larger for early non-atopic eczema than early atopic eczema, we were limited in sample size to fully elucidate these differences. Prenatal dog effect on persistent eczema failed to reach significance after covariate adjustment. However, the effect size was strong (OR=0.32) and warrants further exploration in larger studies.
Dogs may provide exposure to microbial diversity beneficial to immune development7, and alterations may influence the immunological mediators leading to atopic conditions6, 7. The first year of life is potentially the critical window6. Our results suggest that in utero exposure may prevent early eczema. Most households with dogs during pregnancy keep their pets during infancy, making effects difficult to disentangle. Study strengths include its prospective nature, large sample size, and representing a racially and socioeconomically diverse population. We accounted for potentially modifiable factors which did not modify the overall effect except for delivery mode. The lack of significant interactions could be due to insufficient sample size. Some biases may include loss to follow-up, potentially biasing effect estimates. Parents with allergic diseases may avoid domestic animal exposure, which could result in reverse causation. Due to limited data, we considered maternal rather than paternal eczema. Our eczema prevalence is higher than previous reports8. However, eczema is reportedly more common in African Americans9.
In summary, our data suggest that prenatal and early life dog exposure has a significant protective effect on eczema development at or before age 2. Since pet-keeping influences infant gut microbial composition7, the lower rate of eczema in dog-exposed children may be linked to altered early-life immune development triggered by microbial exposures. Clinically, our findings suggest that prenatal dog exposure could protect against early eczema.
Supplementary Material
Clinical Implications:
Dog exposure is protective in early eczema development, but effects in late childhood are unknown. We find prenatal and first year dog exposure has a significant protective effect on early eczema development by age 2.
Funding:
National Institutes of Health (R01AI051598) and the Fund for Henry Ford Hospital
Footnotes
Conflicts of Interest: The authors have nothing to disclose.
REFERENCES
- 1.Hill DA, Spergel JM. The atopic march: Critical evidence and clinical relevance. Ann Allergy Asthma Immunol. 2018;120(2):131–7. doi: 10.1016/j.anai.2017.10.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Johnson CC, Ownby DR. Allergies and Asthma: Do Atopic Disorders Result from Inadequate Immune Homeostasis arising from Infant Gut Dysbiosis? Expert Rev Clin Immunol. 2016;12(4):379–88. Epub 20160207. doi: 10.1586/1744666X.2016.1139452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Pelucchi C, Galeone C, Bach JF, La Vecchia C, Chatenoud L. Pet exposure and risk of atopic dermatitis at the pediatric age: a meta-analysis of birth cohort studies. J Allergy Clin Immunol. 2013;132(3):616–22 e7. Epub 20130524. doi: 10.1016/j.jaci.2013.04.009. [DOI] [PubMed] [Google Scholar]
- 4.Thorsteinsdottir S, Thyssen JP, Stokholm J, Vissing NH, Waage J, Bisgaard H. Domestic dog exposure at birth reduces the incidence of atopic dermatitis. Allergy. 2016;71(12):1736–44. Epub 20160809. doi: 10.1111/all.12980. [DOI] [PubMed] [Google Scholar]
- 5.Aichbhaumik N, Zoratti EM, Strickler R, Wegienka G, Ownby DR, Havstad S, Johnson CC. Prenatal exposure to household pets influences fetal immunoglobulin E production. Clin Exp Allergy. 2008;38(11):1787–94. Epub 20080811. doi: 10.1111/j.1365-2222.2008.03079.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Wegienka G, Havstad S, Kim H, Zoratti E, Ownby D, Woodcroft KJ, Johnson CC. Subgroup differences in the associations between dog exposure during the first year of life and early life allergic outcomes. Clin Exp Allergy. 2017;47(1):97–105. Epub 20161010. doi: 10.1111/cea.12804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Levin AM, Sitarik AR, Havstad SL, Fujimura KE, Wegienka G, Cassidy-Bushrow AE, Kim H, Zoratti EM, Lukacs NW, Boushey HA, Ownby DR, Lynch SV, Johnson CC. Joint effects of pregnancy, sociocultural, and environmental factors on early life gut microbiome structure and diversity. Sci Rep. 2016;6:31775. Epub 2016/08/26. doi: 10.1038/srep31775. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kurukulaaratchy R, Fenn M, Matthews S, Hasan Arshad S. The prevalence, characteristics of and early life risk factors for eczema in 10-year-old children. Pediatr Allergy Immunol. 2003;14(3):178–83. doi: 10.1034/j.1399-3038.2003.00036.x. [DOI] [PubMed] [Google Scholar]
- 9.Wegienka G, Havstad S, Joseph CL, Zoratti E, Ownby D, Woodcroft K, Johnson CC. Racial disparities in allergic outcomes in African Americans emerge as early as age 2 years. Clin Exp Allergy. 2012;42(6):909–17. doi: 10.1111/j.1365-2222.2011.03946.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.