The complex and highly heterogeneous course of MS is underpinned by multiple pathogenetic mechanisms including inflammation, neurodegeneration, and patient-specific factors [4] which track different disability trajectories across individuals and over time.

Age is one of the major patient-specific factors contributing to making the disability trajectories different throughout time. A wide range of processes in peripheral immune cells and CNS cells, like astrocytes and microglia, change with aging and may affect pathophysiology, disability level, and treatment response in MS, highlighting the link between the clinical course of the disease and chronological age [21]. Accordingly, in this real-world study, applying LMMRM, which enabled us to compare the rate of the disability accrual among patients with different age at clinical onset, we provided evidence that the evolution of disability accrual is entirely different in POMS, AOMS, and LOMS regardless of treatment. No differences, indeed, were found among the groups in terms of total exposure time, proportion of follow-up time spent on pDMT and proportion of patients starting with a moderate efficacy or with a high efficacy DMT.

POMS, despite a higher frequency of relapses in the period between disease onset and the first visit, exhibited a slower disability accumulation in comparison to AOMS and LOMS.

Patients with PIRA showed a significantly steeper increase in EDSS scores than those without PIRA, in all age at clinical onset groups, as already reported in a previous paper [14]. Most importantly, we demonstrated that, although the median time to the first PIRA event from disease onset did not differ in POMS, AOMS and LOMS (2.41; 2.08 and 2.17 years, respectively), in POMS, the difference in yearly rates of delta3-EDSS increase over time between individuals with and without a PIRA is delayed and less pronounced than in AOMS and LOMS.

This finding is consistent with previous research that demonstrates children with MS not only recover from relapses considerably better than AOMS, but also considerably improve their functional system and EDSS scores three to five times more frequently than in adults [22]. The mechanisms of this improved recovery may be related to a greater capacity of remyelination and neuroplasticity in younger individuals, subsequently decreasing with aging [23, 24]. Disability trajectory in AOMS patients showed, after an initial reduction in the mean EDSS, a slight but constant increase in neurological disability. Delta-EDSS trajectory in LOMS showed a faster trend of progression compared to POMS and AOMS patients, and subjects with PIRA have a higher slope of disability compared to those not presenting it.

Several previous studies have highlighted the influence of age on disease course [25,26,27,28,29]. A recent study used group-based trajectory models to define four MS severity profiles among RRMS patients in the I-MS&RD, resulting in LOMS being associated with a rapid worsening of EDSS scores [30]. In line with these results, MS literature has shown that adults with disease onset at older age reach ambulatory disability milestones faster than younger adults [31].

Disability trajectories of MS patients have been already assessed in previous studies using data from the Big Multiple Sclerosis Data (BMSD) network [32, 33]. Three distinct disability trajectories have been observed in the more recent research on the long-term disability of people with secondary progressive MS, which was supposed to reflect different pathogenic processes of progression [33].

The observation of a slower accumulation of disability in treated POMS patients vs treated AOMS and LOMS patients during a 5-year follow-up period could support also the idea that treatment benefits on disability progression are highest in younger individuals and decrease with age. A meta-analysis of the main RCTs showed that the efficacy of DMTs decreases with increasing age [34]. Furthermore, a recent observational study by the I-MS&RD has shown that the efficacy of DMTs in delaying the achievement of EDSS 4.0 is greater in POMS and AOMS than in LOMS [35]. To reinforce the concept, another real-world study using data from the I-MS&RD revealed that in POMS the risk of persistent disability has decreased by 50–70% in recent diagnosis epochs, probably due to the improvement and timing in therapeutic and diagnosis [36].

A recent study from our group pointed out that PIRA can occur at any age and also in POMS, which is not protective against progression phenomena, rising dramatically its frequency with increasing age, in parallel with a worsening trend of disability [13]. Therefore, although PIRA resulted rarely detectable in children, POMS is currently not considered a protective factor against PIRA, which occurs throughout the disease course and gradually becomes more frequent with aging.

Some limitations of our study deserve discussion. Our analysis of disability accumulation relies only on the EDSS score. Although the baseline MRI features are a crucial prognostic factor, we could not include MRI data because of the lack of a systematic MRI acquisition and protocol analysis. Despite these considerations, our study used the large real-world database of the I-MS&RD, constantly improving quality of data and able to delineate disease evolution over time [37]. In conclusion, our results further support the new view of MS as a single, continuous process over time. We confirm that age at clinical onset remains highly relevant in determining the rate of disability accrual in contemporary cohorts of MS patients treated with DMTs. Moreover, the results showed PIRA occurrence in POMS is not uncommon, but its effect on the yearly rates of EDSS increase over time is delayed and less pronounced than in AOMS and LOMS.