In addition to age-related changes in individual sensory systems, older adults also experience changes in how information from multiple sensory modalities is combined. In particular, older adults exhibit increased multisensory integration, showing larger gains than younger adults when information in one sensory modality is accompanied by matching information in a different modality. Although these enhanced integration capabilities can be beneficial in certain situations, older adults are also more distracted by irrelevant or incongruent sensory stimuli than younger adults. The overarching purpose of this thesis was to investigate the source of age-related increases in multisensory integration and to evaluate the effects of a cognitive training intervention designed to reduce distractibility in older adults.

Based on evidence that older adults show impairments on some measures of attention, we hypothesized that age-related enhancements in multisensory integration were due to deficits in older adults' ability to focus attention on a single sensory modality. In contrast to our expectations, older adults were able to effectively engage modality specific selective attention to limit multisensory integration on a cued audiovisual discrimination paradigm. These results suggested that top-down mechanisms of modality-specific attention are intact in older adults, and that increased distractibility may be the result of increases in baseline sensory processing.

To counteract age-related increases in distractibility we developed a novel cognitive training program aimed at reducing older-adults' susceptibility to irrelevant sensory information. A randomized controlled trial demonstrated that this intervention successfully reduced cross-modal distraction and multisensory integration in healthy older adults. Modest training-related improvements were also noted in dual-task performance, processing speed, and working memory, functions not explicitly targeted by the intervention program.

Additionally, the intervention also resulted in increased resting cerebral blood flow to the prefrontal cortex, a region known to demonstrate age-related hypoperfusion and to participate in attention, inhibition, and other executive functions. Although these perfusion changes were not accompanied by clear alterations in brain structure or functional activity, they were modestly correlated with behavioral improvements, suggesting that resting blood flow may be a sensitive marker for evaluating the effects of cognitive training interventions.