Prevalence statistics of hearing loss in adults: harnessing spatial big data to estimate patterns and trends

Background: Hearing loss is estimated to affect over eight million adults aged over 50 in England. These estimates are based on the prevalence data reported in the ‘Hearing in Adults’ study by Davis, who collected audiological data for 1,538 subjects 50 years old and above in the 1980s. The prevalence (%) per age group from this study’s samples is being applied to the most recent English census. We aimed to (a) explore regional patterns and trends of hearing loss in a representative longitudinal prospective cohort study of the English older population, and (b) identify potential regional differences with the current prevalence estimates.Methods: We utilised the full dataset (74,699 person-years) of the English Longitudinal Study of Ageing (ELSA). We used local spatial statistics to analyse spatial distributions, patterns and trends of the geographical data, examining both self-reported and objective hearing data. The objectively identified hearing loss was defined as greater than 35 dB HL at 3.0 kHz, in the better hearing ear, as measured by a handheld audiometric screening device (HearCheck Screener).Results: There was a wide variation in hearing loss prevalence in representative samples from different regions in England with similar age profiles (Fig 1.). In a period of 15 years (2002-2017) the increase rate of hearing loss ranged between regions from 3.2% to 45%. The Getis-Ord Gi* spatial statistic showed marked regional variability, and hearing health inequalities between Northern and Southern England that were previously unknown.Conclusion: The profound multidisciplinary professional and experimental care in the ‘Hearing in Adults’ study by Davis is broadly recognised; however, our study showed that this data does not remain valid and generalisable. The time of small-scale research should be consigned to the past; applying computational approaches in audiology might offer promising solutions to generate large-scale epidemiological inferences to improve population’s hearing health