A European cross-sectional study was conducted with the aims of: i) examining inter-country differences in prevalence of tinnitus by comparing prevalence estimates in countries with relatively low versus high gross domestic product (GDP) per capita and comparing prevalence estimates in countries within different European regions (i.e., Northern, Southern, Eastern and Western Europe); ii) understanding the effect of sex, age, socio-economic variables, body mass index (BMI), and marital status on overall prevalence of tinnitus; iii) assessing the effect of hearing difficulty on tinnitus prevalence; and iv) exploring the effect of tinnitus severity on use of healthcare resource for tinnitus Overall, approximately 12,000 individuals aged 18 years or over are enrolled in 12 different European countries (BG, DE, ES, FR, GR, IE, IT, LV, PL, PT, RO, UK). Survey sample from each country were representative of the country-specific population in terms of sex, age, and socio-economic characteristics.

BACKGROUND AND RATIONALE Tinnitus is a common symptom of clinically heterogeneous pathologies and is defined as the conscious perception of an auditory sensation in the absence of a corresponding external stimulus. Tinnitus, in its debilitating form, negatively affects emotional health and social well-being, and can precipitate psychological distress severe enough to cause self-harm and suicidal ideations, while exerting substantial individual and societal financial burden. Therefore, it becomes crucial to implement adequate preventive measures and health interventions. According to population-based studies from Western Europe and US, tinnitus prevalence lies between 10% and 15%. Nonetheless, there is a lack of clear understanding of global prevalence of tinnitus. Inherent methodological drawbacks and between study differences hinder deriving a global estimate by pooling data from studies conducted in different countries and world regions. Therefore, uncertainties with regards to the prevalence of tinnitus remain.

OBJECTIVE The primary objective of this population-based cross-sectional study is to estimate the overall and country specific prevalence of any tinnitus, bothersome tinnitus, and severe tinnitus in 12 European Union (EU) countries.

The study data have been analyzed to fulfil the following secondary objectives:

• To examine inter-country differences in prevalence of tinnitus by making the following comparisons:

  • Comparing prevalence estimates in countries with relatively low versus high gross domestic product (GDP) per capita
  • Comparing prevalence estimates in countries within different European regions (i.e., Northern, Southern, Eastern and Western Europe)

• To understand the effect of sex, age, socio-economic variables, body mass index (BMI), and marital status on overall prevalence of tinnitus

• To assess the effect of hearing difficulty on tinnitus prevalence

• To explore the effect of tinnitus severity on use of healthcare resource for tinnitus

METHODS Study design To achieve the aims of this study, the investigators conducted face-to-face population-based cross-sectional survey in 12 strategically selected EU28 countries (Bulgaria, England, France, Germany, Greece, Ireland, Italy, Latvia, Poland, Portugal, Romania, and Spain).

For each country, the study sample is representative of the general population aged 18 years and above in terms of age, sex, geographic area, and socio-economic characteristics. Multi-stage methodology was used as the preferred sampling methodology. In the first stage, the primary unit of selection was geographic area or voting center. In the second stage, households or municipalities were selected. In the last stage, respondents were chosen randomly, in order to be representative of the population. Other methods, including stratified or simple sampling or quota methods, were also accepted wherever it is not possible to conduct a multistage random sampling.

Selection of countries For the present study, an appropriate sample of 12 European countries-all EU28 Member States (MS) -representing geographical, legislative and cultural variations across the EU, have been selected. The selected European countries were as follows: Bulgaria (BG), England (UK), France(FR), Germany(DE), Greece (GR), Ireland (IE), Italy (IT), Latvia (LV), Poland (PL), Portugal (PT), Romania (RO), and Spain (ES). These countries were selected after taking into account several aspects, including: i) different geographical distribution; ii) different level of per capita income. According to geographic area, the investigators considered countries from northern (IE, UK), central/western (DE, FR), central/eastern (BG, LV, PL, RO), and southern Europe (ES, IT, GR, PT). Moreover, the selection included the 7 most populated EU28 MS (DE, ES, FR, IT, UK, PL, RO), and, among less populated countries, the investigators prioritized middle-income countries (BG, RO). Overall, the 12 selected EU MS represent 78.8% of the EU28 overall population.

Study population In each of the 12 selected European countries, approximately 1000 individuals, representative of the general population aged 18 years and above in terms of age, sex, geographic area, and socio-economic characteristics, were enrolled in this cross-sectional study. The final sample therefore includes a total of 11,427 subjects.

Sample size computation The sample size for each country allowed us to obtain prevalence estimates with a maximum standard error (SE) lower than ± 1.6%. Therefore, with such a sample size the investigators have been able to provide stable prevalence estimates (with a relatively small 95% confidence interval), both overall and country-specific.

Data collection In each country, individuals were surveyed by trained interviewers through face-to-face interviews using a structured standardized questionnaire. Data collection was coordinated by DOXA, a leading market research organization and Italian branch of the Worldwide Independent Network/Gallup International Association (WIN/GIA), and performed in collaboration with its European partners. Data was collected within computer assisted personal interviews (CAPI). The fieldworks lasted two to three months in each country. The English version of the questionnaire has been developed by collaborators from University of Nottingham (UK) University of Regensburg (Germany), and Mario Negri Institute (Italy). Before fieldwork, the tinnitus-related questions were translated from English into various country-specific languages using the good practice guidelines for questionnaire translation. The details of the translation process have been published as a technical note.

The country-specific information provided by DOXA and its European partners for each European country include information on sampling methodology (e.g., sampling method used, response rate, fieldwork dates, age range). The survey questionnaire is designed to collect individual-level information for each survey participant, and is divided in two sections. The first section provided information on: socio-economic and demographic characteristics (e.g., sex, age, level of education, and profession). The second section provided information on tinnitus prevalence, tinnitus severity, healthcare resource use for tinnitus, and hearing difficulty prevalence.

Statistical analysis plan Creation of the survey dataset Once the fieldwork was completed, DOXA collected data from its European partners and after internal checks provided to Mario Negri Institute the 12 anonymous data files in text format, with the corresponding codebook in English language. The 12 data files were managed by the team at the Laboratory of Lifestyle Epidemiology, Department of Environmental Health Sciences of Mario Negri Institute. Using the statistical package SAS, version 9.4 (SAS Institute, Cary, NC, USA), they performed the input of the 12 textual data file into 12 corresponding SAS datasets, and then created a unique final dataset in SAS. This dataset was checked for coherence and for the presence of possible errors, cleaned, and finalized through the generation of queries which were addressed by DOXA and its European partners.

Statistical analyses Statistical analyses required to achieve objectives has been conducted by the teams from the Laboratory of Lifestyle Epidemiology, Department of Environmental Health Sciences of Mario Negri Institute and the University of Nottingham, using SAS statistical package. Prevalence of three working definitions of tinnitus (any, bothersome and severe tinnitus) have been estimated through univariate statistical analyses (i.e., descriptive statistics for categorical data, including absolute or relative frequencies), overall, by country, and in strata of selected demographic and socio-economic (individual and country-specific) characteristics. Continuous variables have been summarized through the use of means and standard deviations. Alternatively, they have been categorized, and thus described through absolute or relative frequencies.

Multivariate analyses have been conducted to identify sub-groups of the population (e.g., based age-groups, low vs. high educated subjects) or clusters of countries (e.g., based on income levels and regional groups) where tinnitus is more frequent. Thus, odds ratios (OR), and the corresponding 95% confidence intervals (CI), were estimated through logistic regression models after adjustment for selected individual-level characteristics.

ETHICAL AND ADMINISTRATIVE ASPECTS Informed consent At recruitment, information on survey characteristics was given to all participants by suitably qualified professionals, who were able to provide answers to any possible questions and to eliminate any concern respondents may have. Information explaining the condition (i.e., tinnitus) was provided to survey participants in the country-specific language in a clear, jargon free manner, using fully understandable terms and language.

Although in EU countries, according to their current legislation, written consent is not mandatory for this type of study (observational study collecting anonymized information), all participants were asked to sign the informed consent form for participation in the study. Since data was collected with CAPI method (using computer or tablet), participants had the possibility to tick an electronic document in order to express their willingness to participate in the study. No written signature was required.

The Ethics Committee of the study coordinator (Ethics Committee of Fondazione IRCCS Istituto Neurologico Carlo Besta, Milan, Italy) notified that no preventative evaluation was required for the present study, since anonymous data was collected (File number 37/2017).

Access to personal data and withdrawal from the study In this survey no sensitive personal data (e.g., health, sexual lifestyle, ethnicity, political opinion, religious or philosophical conviction), genetic information, or data involving tracking or observation of participants was collected. According to other personal data, it was made clear that subjects are completely free to withdraw their records from participation in the survey, and that the consent given by participants in the study can be withdrawn, without any explanation or justification. In this case, all data pertaining to the subjects in question was destroyed.

Ownership of the data Individual-level anonymous data sent from DOXA and its European partners, and stored at the Mario Negri Institute, are the property of the study coordinator.

RELEVANCE AND IMPLICATIONS One of the expected outcomes of this study is the first estimate of tinnitus prevalence from a Pan-European study. Given the lack of standardized tinnitus prevalence questions and a global prevalence estimate, this study provides both standard assessment measures and an estimate of how prevalent tinnitus symptoms are. In this study, the investigators report the prevalence of any tinnitus using the same set of research questions in various languages. This measure identifies the commonality of the condition. Furthermore, the investigators assess the proportion of individuals having bothersome and severe symptoms, who exert the true burden on financial and personnel resources in the healthcare system. These estimates measured in a Pan-European study demonstrate the size of the problem. This can guide resource allocation for implementing effective measures and attract various stakeholders, including research funders, industrial and healthcare organizations, to engage in needful endeavors for tackling tinnitus.

Inclusion Criteria:

• individuals aged 18 years and above
• individuals who are residents of the 12 selected countries
• individuals able to understand and answer the questions of the study questionnaire in the country-specific language
• individuals who formally accept to participate in the study

Exclusion Criteria:

• No specific exclusion criteria are considered

In each of the 12 selected European countries, approximately 1000 individuals, representative of the general population aged 18 years and above in terms of age, sex, geographic area, and socio-economic characteristics, were enrolled in this cross-sectional study. The final sample therefore includes a total of 11,427 subjects.

The requests of IPD will be evaluated and data will be provided upon reasonable request.