Background

Observational studies suggest an increased risk of eczema in children living in hard versus soft water areas, and there is therefore an interest in knowing whether softening water may prevent eczema. We evaluated the feasibility of a parallel-group assessor-blinded pilot randomised controlled trial to test whether installing a domestic ion-exchange water softener before birth in hard water areas reduces the risk of eczema in infants with a family history of atopy.

Methods

Pregnant women living in hard water areas (>250 mg/L calcium carbonate) in and around London UK, were randomised 1:1 antenatally to either have an ion-exchange water softener installed in their home or not (i.e. to continue to receive usual domestic hard water). Infants were assessed at birth and followed-up for 6 months. The main endpoints were around feasibility, the primary endpoint being the proportion of eligible families screened who were willing and able to be randomised. Clinical endpoints were evaluated including frequency of parent-reported doctor-diagnosed eczema and visible eczema on skin examination. Descriptive analyses were conducted, and no statistical testing was performed as this was a pilot study.

Results

One hundred and forty-nine families screened were eligible antenatally and 28% (41/149) could not have a water softener installed due to technical reasons or lack of landlord approval. Eighty of 149 (54%) were randomized, the primary endpoint. Two participants withdrew immediately after randomisation, leaving 39 participants in each arm (78 total). Attrition was 15% (12/78) by 6 months postpartum. All respondents (n=69) to the study acceptability questionnaire reported that the study was acceptable. Fifty-six of 708 (7.9%) water samples in the water softener arm were above the hard water threshold of 20 mg/L CaCO3. By 6 months of age 27/67 infants (40%) developed visible eczema, 12/36 (33%) vs 15/31 (48%) in the water softener and control groups respectively, difference -15% (95% CI -38, 8.3%), with most assessments (≥96%) remaining blinded. Similarly, a lower proportion of infants in the water softener arm had parent-reported, doctor-diagnosed eczema by 6 months compared to the control arm, 6/17 (35%) vs 9/19 (47%), difference -12% (95% CI -44, 20%).

Conclusion

A randomised controlled trial of water softeners for the prevention of atopic eczema in high-risk infants is feasible and acceptable.