Numeracy, the capacity to comprehend and apply numbers and basic quantitative concepts, plays a crucial role in everyday functioning. Like other cognitive functions, it has been reported to decline with advancing age, potentially increasing older adults' vulnerability in managing daily tasks. Numerical cognition relies on a highly integrated neural network, with key contributions from the intraparietal sulcus and fusiform gyrus, alongside broader frontal and parietal involvement. Age-related functional changes in these regions may therefore contribute to reductions in numeracy skills. Clarifying these processes could help distinguish normal from pathological cognitive aging and inform the design of environments that better support older populations. In this study, we adopted a multidisciplinary approach integrating cognitive tasks including the Number Line Task (NLT-100), visuospatial working memory (VSWM), executive functions, and arithmetic paradigms, with high-density EEG recorded during numerical calculations to examine age-related differences in numeracy. Forty healthy subjects participated: 20 young (55% female, mean age: 23.25±2.76) and 20 older adults (55% female, mean age: 69.75±5.48; mean MoCA: 25.45±3.20). EEG data were recorded using an actiCHamp Plus amplifier (Brain Products GmbH) with a 128-electrode cap. Stimuli were presented via E-Prime 3.0 and responses recorded using a Chronos device (PST, USA). Neuropsychological measures of executive functions (Trail Making Test–B, Category Fluency) and VSWM (Corsi Block-Tapping Test, forward and backward) revealed significantly poorer performance in older adults, indicating reduced global cognitive efficiency. In contrast, no age-related differences emerged in analog magnitude representation (NLT-100) or in the non-symbolic comparison task (Panamath), suggesting preservation of basic numerical abilities (Number Sense) in aging. Similarly, no differences were observed in the automatic retrieval of arithmetic facts or in mental calculation across complexity levels; however, older adults showed significantly longer response times. EEG data revealed age-related differences in brain activation patterns during numerical calculation. Older adults exhibited reduced ERP amplitudes in cortical regions typically involved in numerical processing, along with prolonged latencies, consistent with the behavioral response time findings. In particular, modulation of the P100 and P300 was observed, with the latter showing increased latency in older adults. Furthermore, the late positive potential appeared more prolonged and showed a greater frontal distribution in older participants, indicating broader engagement of attentional and executive processes. We conclude that reduced activity in primary cortical regions may contribute to age-related decline in numeracy, while recruitment of additional cortical areas may serve as a compensatory mechanism.
Numeracy across the lifespan: evidence for stability in older adults from an interdisciplinary study integrating cognitive and high-density EEG measures
Beniamina Mercante
;Federico Zorzi;Valentina Cabriolu;Marta Lodi;Carla Meloni;Rachele Fanari;Paolo Enrico
2026-01-01
Abstract
Numeracy, the capacity to comprehend and apply numbers and basic quantitative concepts, plays a crucial role in everyday functioning. Like other cognitive functions, it has been reported to decline with advancing age, potentially increasing older adults' vulnerability in managing daily tasks. Numerical cognition relies on a highly integrated neural network, with key contributions from the intraparietal sulcus and fusiform gyrus, alongside broader frontal and parietal involvement. Age-related functional changes in these regions may therefore contribute to reductions in numeracy skills. Clarifying these processes could help distinguish normal from pathological cognitive aging and inform the design of environments that better support older populations. In this study, we adopted a multidisciplinary approach integrating cognitive tasks including the Number Line Task (NLT-100), visuospatial working memory (VSWM), executive functions, and arithmetic paradigms, with high-density EEG recorded during numerical calculations to examine age-related differences in numeracy. Forty healthy subjects participated: 20 young (55% female, mean age: 23.25±2.76) and 20 older adults (55% female, mean age: 69.75±5.48; mean MoCA: 25.45±3.20). EEG data were recorded using an actiCHamp Plus amplifier (Brain Products GmbH) with a 128-electrode cap. Stimuli were presented via E-Prime 3.0 and responses recorded using a Chronos device (PST, USA). Neuropsychological measures of executive functions (Trail Making Test–B, Category Fluency) and VSWM (Corsi Block-Tapping Test, forward and backward) revealed significantly poorer performance in older adults, indicating reduced global cognitive efficiency. In contrast, no age-related differences emerged in analog magnitude representation (NLT-100) or in the non-symbolic comparison task (Panamath), suggesting preservation of basic numerical abilities (Number Sense) in aging. Similarly, no differences were observed in the automatic retrieval of arithmetic facts or in mental calculation across complexity levels; however, older adults showed significantly longer response times. EEG data revealed age-related differences in brain activation patterns during numerical calculation. Older adults exhibited reduced ERP amplitudes in cortical regions typically involved in numerical processing, along with prolonged latencies, consistent with the behavioral response time findings. In particular, modulation of the P100 and P300 was observed, with the latter showing increased latency in older adults. Furthermore, the late positive potential appeared more prolonged and showed a greater frontal distribution in older participants, indicating broader engagement of attentional and executive processes. We conclude that reduced activity in primary cortical regions may contribute to age-related decline in numeracy, while recruitment of additional cortical areas may serve as a compensatory mechanism.
I metadati presenti in IRIS UNICA sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono protetti da diritto d'autore, salvo diversa indicazione.
