The study was initiated in early 2021 at University Medicine Halle (Saale), a mid-sized city of roughly 230,000 people. The immediate trigger was the pandemic: researchers needed rapid, repeated data on infections, vaccinations and testing behaviour. But from the outset, the team had a longer ambition in view. DigiHero was conceived as a permanent research platform for studying chronic diseases, mental health, health behaviour and healthy ageing. It has since become part of the German Center for Mental Health, contributing directly to that institution’s research agenda on the mental wellbeing of the population.

Recruitment relied on two complementary strategies. The first was classical, population-based sampling: invitation letters sent by postal mail in five waves, with addresses obtained from local registration offices, service providers or a federal data portal in line with German data protection rules. The campaign began in Halle, where all 120,000 households received an invitation, then spread across Saxony-Anhalt, Saxony and Thuringia before expanding to randomly selected NUTS-2 regions in the remaining federal states. Large cities with more than 500,000 inhabitants were deliberately included to represent Germany’s urban centres, with 90,000 addresses selected in each of the city-states Berlin and Hamburg and 60,000 in Magdeburg. In total, about four million households received a postal invitation between January 2021 and February 2025. The second strategy was continuous open registration through the study website, supported by media campaigns, radio spots and press releases. Around 14 percent of participants registered without a personal invitation letter, and the researchers can compare these self-recruited volunteers with the invited sample in their analyses.

Eligibility was defined broadly: anyone born between 1936 and 2003, corresponding to ages 18 to 85 in 2021, with a main residence in an invited municipality, an internet connection, a personal email address and sufficient German language skills. To enable household-level studies, such as investigations of how respiratory viruses spread within families, the team maximised the number of participating households by inviting only one person, an anchor, per household sharing an identical address and surname. The result is a cohort that covers both dense urban districts and remote rural regions, something the designers say distinguishes DigiHero from study-centre-based cohorts tied to single locations.

Once registered, participants complete a baseline questionnaire covering sociodemographic characteristics, and then enter a rhythm of four regular contacts per year. All surveys run on LimeSurvey, an open-source platform hosted on an internal web server in a demilitarised zone within the IT infrastructure of University Medicine Halle. Each participant is linked to a 15-digit alphanumeric pseudonym that connects responses across questionnaires, and the software automatically saves progress, so a questionnaire can be paused and resumed on any device, from desktop to smartphone. Up to two reminders are sent per survey, ten to fourteen days apart, and participants can skip individual questionnaires without leaving the cohort entirely. Beyond the regular cycle, the design allows rapid additional invitations: when Russia invaded Ukraine in early 2022, the team fielded a mental health survey just one week after the war began, and a follow-up six months later. A similar rapid-response survey captured heat-related stress during the summer 2024 heatwave.

What has the cohort actually delivered? Its early findings have already shaped the scientific understanding of post-COVID condition. In an analysis of roughly 17,000 participants with at least one SARS-CoV-2 infection, the researchers found a substantially lower risk of ongoing symptoms after infections with newer virus variants, and evidence that people infected previously without developing post-COVID had a reduced risk after a subsequent infection. Vaccination, they concluded, protects indirectly by lowering the risk and severity of infection rather than reducing post-COVID risk after a breakthrough infection. A dedicated post-COVID registry now follows more than 4,000 participants at six-month intervals, over 1,000 of whom still report ongoing symptoms. In a separate analysis, fewer than 40 percent of respondents reported being symptom-free one year after infection, underscoring how persistent the condition can be.

The cohort has also quantified the pandemic’s continuing burden on respiratory health. In the winter season 2022/23, approximately 38,000 participants reported their acute respiratory infections and test results, allowing the team to calculate that SARS-CoV-2 contributed an additional 21 percent of acute respiratory infections, equivalent to roughly 17 million extra infections across the German population. Household transmission subprojects in the winters of 2022 and 2023 combined self-collected dry blood samples and nasal swabs with symptom diaries, linking self-reported data to immunological measurements. On the mental health side, the rapid Ukraine-war surveys showed that media exposure to the conflict drove a substantial rise in anxiety symptoms, exceeding levels observed during the pandemic, with people from lower-income households especially affected. Another study comparing post-COVID and depressive disorders across 45 mental health variables found markedly different symptom structures, with the largest gaps in exhaustion, fatigue, impaired task completion and reduced resilience.

Methodologically, DigiHero occupies a distinctive position in the German research landscape. Established cohorts such as KORA, SHIP, CARLA, Recall and the national NAKO study rely on physical study centres, which constrains how often participants can be recontacted and how quickly results can be produced. DigiHero’s fully virtual design integrates research into participants’ daily lives, reduces paper waste and transport resources, and permits much more frequent measurement of change over time. A two-stage design allows broad data collection across the whole cohort followed by in-depth assessment of defined subgroups, an approach particularly useful for studying rare exposures and rare diseases. Future plans include linking self-reported data with health insurance records for validation, connecting postal addresses to geospatial environmental data through open-source sources such as OpenStreetMap, exporting data from wearables and GPS movement tracking, conducting app-based cognitive testing, and even building an embedded multigenerational cohort by including family members.

The researchers are candid about the limitations. Response rates ranged from 1.2 to 8 percent depending on the recruitment area, and the cohort overrepresents people with higher education while underrepresenting those born outside Germany, a well-documented pattern of selection bias in modern cohort studies. Participation also requires internet access and an email address, though in 2023, 99 percent of German households had active internet access and about 83 percent of internet users sent and received emails, so the excluded share of the population is small. Previous work by the group showed that online-only surveys yield answers similar to mixed-mode designs across a wide range of health questions. Compared with German census data, DigiHero participants are better educated than the general population, but the gap is substantially smaller than for other established population-based studies such as ALLBUS, SOEP, SHARE or NAKO.

Data protection is built into the architecture. Names and postal addresses sit on an internet-disconnected computer in a separate database, follow-up responses are stored on the LimeSurvey server, and strict role-based access rights ensure that no individual can link identities to responses, a linkage performed only when a participant withdraws consent and requests deletion. The Ethics Committee of Martin Luther University Halle-Wittenberg approved the study, and new questionnaires must pass ethics review before deployment. There is no defined end point, and the team intends to follow participants indefinitely, building a comprehensive health database spanning diseases, behaviour, mental health and contextual factors. Data are available to researchers through a Use and Access Committee, and the consortium welcomes national and international collaborations with institutions in countries offering data protection comparable to Germany’s. The next milestone is citizen science: a digital dialogue in which participants themselves propose health topics and help shape the research agenda. In a field where the gap between a health crisis and the evidence needed to respond can stretch into years, DigiHero’s bet is that a cohort built for speed, flexibility and scale can close it.

Subject of Research: A nationwide fully digital population-based prospective cohort study in Germany for research on diseases, mental health and healthy ageing

Article Title: Prospective cohort study for digital health research in Germany (DigiHero): objectives, design and baseline characteristics

Article References: Klee, B., Diexer, S., Glaser, N., Gottschick, C., Frost, J., Massag, J., Pfrommer, L. R., Herrmann, T., Purschke, O., Schmidt, A., Dierks, C., Frese, T., Gekle, M., Girndt, M., Hettmer, S., Heuser, M., Höll, J. I., Moor, I., Rosendahl, J., … Mikolajczyk, R. (2026). Prospective cohort study for digital health research in Germany (DigiHero): objectives, design and baseline characteristics. European Journal of Epidemiology. https://doi.org/10.1007/s10654-026-01454-7

Image Credits: AI Generated

DOI: 10.1007/s10654-026-01454-7

Keywords: DigiHero, digital health, cohort study, epidemiology, COVID-19, post-COVID condition, mental health, Germany, public health, online surveys, wearables, population-based research