An analysis has found that prenatal omega-3 long-chain polyunsaturated fatty acid supplementation may be associated with a reduced risk of childhood atopic dermatitis only among offspring of mothers carrying the cyclooxygenase-1 rs1330344 TT genotype.

In the secondary analysis of a randomized clinical trial, published in JAMA Dermatology, investigators used data from the Copenhagen Prospective Studies on Asthma in Childhood 2010 (COPSAC2010) birth cohort to examine 635 mother-child pairs followed-up to age 10 years. Among the children, 57% (n = 363) of them were female and 43% (n = 272) of them were male. The intervention group comprised 321 (51%) pairs, while the control group had 314 (49%) pairs. Pregnant patients were randomly assigned to receive either 2.4 g/d of long-chain polyunsaturated fatty acid (n-3 LCPUFA) (55% eicosapentaenoic acid and 37% docosahexaenoic acid) or placebo from 24 weeks' gestation until 1 week postpartum.

The study was conducted from January 2019 to December 2021, with data analyzed from January to September 2023. It was a prespecified secondary analysis of the n-3 LCPUFA randomized clinical trial (ClinicalTrials.gov identifier NCT00798226) within the COPSAC2010 cohort, which was originally conducted from March 2009 to March 2014.

Genomewide genotyping was conducted using the Illumina HumanOmniExpressExome-8 V1.2 BeadChip. Urinary eicosanoids were quantified using mass spectrometry. Statistical analyses included Cox proportional hazards regression models, linear regressions, and mediation analyses.

The key findings included:

• Among mothers with the TT genotype (390 pairs, 61% of the cohort), n-3 LCPUFA supplementation was associated with a 30% reduced risk of offspring developing atopic dermatitis (AD) by age 10 years (hazard ratio [HR] = 0.70, 95% confidence interval [CI] = 0.50–0.98, P = .04).
• No significant association was found for mothers with the CT genotype (209 pairs, 33% of the cohort; HR = 1.29, 95% CI = 0.79–2.10, P = .31).
• An increased risk was observed for mothers with the CC genotype (37 pairs, 6% of the cohort; HR = 5.77, 95% CI = 1.63–20.47, P = .007).

The study also explored potential mechanisms underlying these associations. Urinary eicosanoid profiles were available for 450 children (64%) at age 1 year. The n-3 LCPUFA intervention was associated with reduced levels of combined thromboxane A2 (TXA2) eicosanoids (β = −0.46, 95% CI = −0.80 to −0.13, P = .006; false discovery rate [FDR]-corrected P = .04). Specifically, levels of 11-dehydro-TXB2 and 11-dehydro-2,3-dinor-TXB2 were significantly reduced.

The cyclooxygenase-1 (COX1) rs1330344 genotype was associated with TXA2 eicosanoid levels, with the C allele linked to higher levels (β per additional C allele = 0.47, 95% CI = 0.20–0.73, P = .001; FDR-corrected P = .02). The n-3 LCPUFA intervention was associated with lower TXA2 eicosanoid levels only among children with the TT genotype (P = .02 for interaction).

Child genotype analysis revealed a significant interaction between prenatal n-3 LCPUFA supplementation, child genotype, and AD risk (P = .002 for interaction). For children with the TT genotype (n = 361), the HR was 0.70 (95% CI = 0.48–1.00, P = .05). For the CT genotype (n = 222), the HR was 1.59 (95% CI = 0.99–2.56, P = .06), and for the CC genotype (n = 34), the HR was 3.16 (95% CI = 0.66–15.28, P = .15).

Additional maternal genotype analyses showed odds ratios (OR) for current AD at age 10 years and generalized estimating equation odds ratios (GEE OR) for yearly prevalence of AT at ages 0 to 10 years.

Current AD at age 10 years:

• TT genotype: OR = 0.49, 95% CI = 0.24–0.98, P = .05
• CT genotype: OR = 1.11, 95% CI = 0.38–3.27, P = .80
• CC genotype: OR = 3.19, 95% CI = 0.36–68.20, P = .30.

Yearly prevalence of AD at age 0 to 10 years:

• TT genotype: GEE OR = 0.62, 95% CI = 0.41–0.94, P = .02
• CT genotype: GEE OR = 1.40, 95% CI = 0.78–2.52, P = .26
• CC genotype: GEE OR = 6.22, 95% CI = 1.17–32.00, P = .03.

A mediation analysis suggested that the association between n-3 LCPUFA intervention and AD risk was mediated through altered levels of combined TXA2 eicosanoids in the TT genotype strata. The results were similar after adjustments for sex, high-dose vitamin D intervention, and preintervention levels of EPA and DHA.

The investigators noted that the C allele of rs1330344 was more prevalent in African and Asian populations (40% to 50%) compared with European populations (21%).

Limitations of the study included the lack of a validation cohort and the low number of mothers with the CC genotype.

Conflict of interest disclosures can be found in the study.