Background. A pro-inflammatory transcriptional setpoint measured before vaccination predicts a stronger antibody response in younger adults but loses that value, or reverses, in older adults. This reversal has been characterised almost entirely for influenza, and it is unknown whether it is specific to influenza or a more general feature of immune ageing. Results. Using the harmonised HIPC-II ImmuneSignatures2 compendium and two independent cohorts processed from raw data with the same pipeline, we built a baseline inflammation score from inflammatory and NF-kB blood transcription modules and tested its interaction with age on the antibody response. The inflammation-by-age interaction was negative in a pooled mixed model (t between -2.1 and -3.0 across three score definitions) and for a binarised high-responder outcome (p = 0.015). For influenza it survived leave-one-study-out, the non-normalised data, continuous-age and study-fixed-effect specifications. A random-effects meta-analysis of the old-minus-young inflammation-to-response slope difference across five cohorts, two of them independent, gave -0.15 (95% CI -0.27 to -0.03; p = 0.015; I-squared = 0%). An independent out-of-compendium replication across five Stanford influenza seasons (115 unique donors after deduplicating repeated enrolments) reproduced the negative interaction in direction in every season (pooled -0.23; I-squared = 0%), though its deduplicated interval still included zero. Within the compendium the reversal was present among viral vaccines (extending from influenza to a single small zoster cohort, 35 adults) and absent for recombinant hepatitis B; a formal vaccine-platform moderator was not significant (p = 0.47) and could not be separated from whether the response is primary or recall. Conclusions. This confirmatory reanalysis replicates and formally quantifies a previously reported, age-dependent reversal of the baseline-inflammation benefit for the vaccine antibody response (Avey et al., Science Immunology 2017; Fourati et al., Nature Communications 2016). Its contribution is not the effect's existence but its formal estimation as an inflammation-by-age interaction across the ImmuneSignatures2 compendium, together with an independent, out-of-compendium replication across five Stanford influenza seasons (consistent in direction in every season, I-squared = 0%, though the deduplicated interval still spanned zero). Whether the reversal generalises beyond influenza to other vaccine types could not be established: a systematic search of the public systems-vaccinology record (ImmuneSpace) found no adequately powered non-influenza cohort, so the cross-type question remains open, neither confirmed nor refuted.
Maire, K.
Advertisement
Stats
- Recommendations n/a n/a positive of 0 vote(s)
- Views 2
- Comments 0