Using single-item measures of self-rated physical health (SRPH) and self-rated mental health (SRMH) widely used in research on inequalities in health in European contexts, this study reports changes in these measures by seropositivity before and after the pandemic (24, 25) The study also assesses self-reported changes in behaviours and economic circumstances before and after the pandemic, by known seropositivity after the SARS-CoV-2 outbreak, and by sociodemographic and socioeconomic factors. In this way, it aimed to assess the association of seropositivity with the outcomes studied and any disparities in the impacts of the pandemic.

After the initial outbreak of SARS-CoV-2, we conducted an initial survey (“Ischgl-1”), with a follow-up study in November 2020 (“Ischgl-2”) (23, 26). Ischgl-1 was a cross-sectional epidemiological survey with measurement of seroprevalence targeted all residents of Ischgl/Tyrol regardless of age or gender. It was conducted between April 21st and 27th, 2020. The study was approved by the ethics committee (EC) of the Medical University of Innsbruck with EC number 1100/2020 (Ischgl-1). 1,259 adults, participated in Ischgl-1, corresponding to around 80% of those living in the town at that time (both permanent residents and seasonal workers). By the end of April 42% of the local population (45% of the adult population) were seropositive, one of the highest seroprevalence levels reported in spring 2020 worldwide (26). The follow-up study, Ischgl-2 (EC Number 1330/2020), was conducted between November 2nd and 8th, 2020.

The current sample comprises the 801 adults aged 18 and over who participated in both Ischgl-1 and Ischgl-2. In Ischgl-1, 1,527 adults were invited to participate, of whom 1,259 (82.4%) agreed (26). 813 (64.6%) of them participated in the follow-up study, Ischgl-2, but 12 were excluded from the analysis due to inconsistent recording of age (n = 9), failure to complete a questionnaire (n = 2) and no blood sample (n = 1). The remaining 801 included slightly more (50.2% vs. 45%) who had been seropositive and more women (54% vs. 51.9%). The age distribution was similar in both studies (23).

The materials and methods used to test for SARS-CoV-2 antibodies and generate an indicator of previous infection with SARS-CoV-2 have been described elsewhere (23). In brief, blood samples were analysed for SARS-CoV-2-binding antibodies using four immunoassays. Samples were screened for anti-SARS-CoV-2-S1-protein IgA and IgG positivity by a commercially available anti-SARS-CoV-2-IgA and-IgG ELISA (Euroimmun, Lübeck, Germany), respectively, using the fully automated 4-plate benchtop instrument Immunomat™ (Virion/Serion, Würzburg, Germany). For both assays values >1.1 were considered positive. Borderline values (0.8–1.1) in the Euroimmun IgG ELISA were rated positive; for the Euroimmun IgA ELISA borderline values were rated as negative. Samples were additionally immunoassayed for anti-SARS-CoV-2 N-protein IgG (anti-N IgG) antibodies in a fully automated manner on the ARCHITECT i2000SR system (Abbott, Illinois, USA). Samples were considered positive if the detected relative light unit (RLU) was >1.4. Anti-N IgG was also quantified using the ElecsysAnti-SARS-CoV-2 (Roche Diagnostics, Indianapolis, USA) according to manufacturer’s recommendations. A COI of ≥1.0 was considered positive. Titres of SARS-CoV-2 neutralizing antibodies were determined using a replication defective vesicular stomatitis virus (VSV) pseudotyped with SARS-CoV-2 spike protein Titers of ≤1:4 were considered as negative, titres of ≥1:16 as positive.

Plasma samples from Ischgl-1 were analysed. The serostatus of the samples was defined as p, d, a or n depending on the binding antibody assays: p = positive = anti-S IgG+ AND anti-N IgG+ (either Roche or Abbott assay positive); d = discordant = anti-S IgG+ OR anti-N IgG+ (either Roche or Abbott assay); a = only IgA+ = only anti-S IgA+ but anti-S IgG-AND anti-N IgG-; n = negative = anti-S IgG-AND anti-N IgG-AND anti-S IgA-. To calculate the seroprevalence, all individuals with serostatus p were considered as seropositive. Individuals with ‘d’ and ‘a’ samples were considered as seropositive if they had neutralizing antibodies ≥1:16.

Ischgl-2 participants were interviewed to answer a questionnaire which included questions about changes to self-rated physical and mental health between November 2019 and November 2020 (Ischgl-2), pre-existing medical conditions, selected medications and symptoms (from a pre-specified list but with the ability to specify other symptoms) including their duration since April 2020 (Ischgl-1) (Table 1). The questionnaire was designed to generate data comparable with existing surveys, including the Community Response Survey developed by John Hopkins University (JHU CRS) that was specifically developed for assessing the impacts of COVID-19 (27), as well as other research on COVID-19 symptoms known about at the time, modified for the Austrian context and translated into German.

Data were collected using Askimed, a web-based eCRF system for data collection and management. The system included checks for completeness, internal consistency, and validity. All interviewers attended a pre-fieldwork training session on the questionnaire and the Askimed system [Askimed access date: 20. 07. 2021 (Available online at: https://www.askimed.com/)].

Seropositivity was defined according to status at Ischgl-1 while all outcome measures were defined using data collected at Ischgl-2. These outcome measures were:

For the analyses presented here, seropositivity in Ischgl-1 is used to assess the relationship between prior SARS-CoV-2 infection and the outcome measures.

Summary characteristics are presented as means ± standard deviation (SD) or median and Interquartile Range (IQR), where appropriate, for continuous variables and number (n) and percentage (%) for categorical variables.

The relationships between the dependent variables and seropositivity were analysed using multivariate binary and multinomial logistic regression models, where appropriate. Age group (18 to <35 years, ≥35 to <50 years, ≥50 to <70 years, and ≥ 70 years), sex and morbidity (binary variable of self-reported hypertension, diabetes, chronic kidney disease (CKD), cancer, cardiovascular disease (CVD), neurological diseases and other lung diseases = 1, or none) were controlled for in all models. Other covariates included body mass index [“normal or underweight” (<25 kg/m²), “overweight” (≥25 to <30 kg/m²) or obese (≥30 kg/m²)], education (Compulsory/High School/University/Other) and working in the tourism sector (yes/no) as well as the outcome measures outlined above used as covariates in models, where appropriate.

Adjusted odds ratios (ORs) from binary logistic regression models are reported and adjusted relative risk ratios (RRs) from multinomial logistic regression models are reported with 95% confidence intervals (95% CI). Statistical tests were two-tailed statistical significance was set at an alpha level of 0.05. All analysis was done using Stata version 17.1 (Stata Corp., College Station, TX, USA).

Table 2 presents the characteristics of participants in the follow-up sample (n = 801). Just over 50% were seropositive in Ischgl-1 (50.3%, n = 403), 54.9% were female (n = 440), median age was 45 years (SD 15.8), one quarter reported at least one specified morbidity (25.7%, n = 205) and over 40% were overweight (BMI ≥ 25 kg/m², n = 331). There was no significant difference in seropositivity by sex, age, morbidity or BMI in the sample.

Figure 1 shows the total percentage of participants reporting negative impacts across all outcome measures in November 2020 by sex, age and seropositivity. Over 55% of participants reported the pandemic had a moderate or severe impact on them. An age gradient was apparent; 63.2% of 18 to <35-year-olds reported a moderate/severe effect of the pandemic compared to 42.9% of adults ≥70 years. A greater proportion of participants reported worse economic outcomes than health outcomes (including negative impacts on behaviours). 8.6% of all participants reported a decline in physical health and nearly a quarter reported a decline in their mental health between November 2019 and November 2020 (the last year). Nearly half of all participants were concerned about their future professional situation (48.4%); the highest proportion of concern was reported among adults aged ≥35 to <50 years (63.9) and the lowest among those aged ≥70 years (5.5%). Similarly, 45.4% of participants were concerned about their future financial situation, with 68.7% of those aged ≥35 to <50 years concerned but only 12.5% of those aged 70 years and over. Only 5.3% of all participants increased alcohol consumption, and only 8.6% of all participants reported a decline in their physical health since the first lockdown implemented in Ischgl in mid-March 2020.

All results reported are from adjusted analyses and control for seropositivity.

We found that seropositivity was associated with indicators of worsened physical health (although absolute levels were low) and certain symptoms in the subsequent months, but not changes in behaviours.

Combining the ≥50 to <70 and ≥70 age groups, 11.0% reported worsened SRPH, similar to the Austrian SHARE study (10.9% of those over 50), which compared pre-SARS-CoV-2 data to June 2020. Though using a different time frame and a three-point response scale (28), both studies align. Despite the low absolute prevalence of poor physical health in our study, it was linked to seropositivity, suggesting a limited but lasting impact of SARS-CoV-2 on self-reported health.

Behavioural changes were even more limited than those in self-reported physical health and were unrelated to seropositivity. Few changes were noted; smoking increased more than it decreased, while alcohol consumption declined and physical activity rose. The drop in alcohol use aligns with NPI-related bar closures, but increased physical activity contrasts with most studies showing declines (29). his may reflect altered work patterns and unique access to the Alps in Ischgl, allowing for socially distanced exercise. Differences in pre-pandemic activity norms, like Iceland’s reliance on team sports, and study design variations may also explain these trends (30).

Despite stable behaviours and modest health declines, 1 in 2 participants reported a moderate or severe pandemic impact, mainly on mental health and finances—both unrelated to seropositivity. This underscores the strong, lasting indirect effects of the pandemic.

By November 2020, 8 months into the pandemic, 15.7% of participants reported poor physical or mental health. Over 20% saw a mental health decline since November 2019, with few improving, while 10% reported worse physical health. In August 2019, the OECD found 7.8% of Austrians aged 15+ had poor SRH. Though that study did not separate physical and mental health, our findings suggest mental health decline was a major pandemic impact (31).

A March–April 2020 study in Austria found 43% of those over 16 reported moderate or severe psychological impact, with higher rates in women (6). Our study, conducted in November 2020, found lower rates of poor SRMH (13.4%) and SRMH decline (23.3%). Differences may stem from study design, bias, and timing. Early 2020 saw strict NPIs, while by November, the population had adapted, possibly easing mental health effects despite Ischgl’s second wave. The NAKO study linked lower psychological distress to lower incidence rates, and Ischgl’s relatively few new infections may have mitigated mental health declines (23).

The most prevalent impact of the pandemic in our study population was economic; 50% of people reported a worsening of their professional and/or financial situation. This did not differ by seropositivity. However, certain segments of the population were disproportionately affected by particular impacts of the pandemic.

Tourism workers faced the greatest professional and financial impact due to NPIs, reflecting the sector’s local importance. Early in the pandemic, the economic effects of NPIs were unclear. By October 2020, a Tyrolean government report estimated the early ski season closure cut Ischgl’s seasonal revenue by 20–30% (17), despite mitigation efforts. Our study supports the well-known link between economic strain and mental health. Tourism workers had higher SRMH decline rates from November 2019–2020, explained by financial and job pressures. Similar findings were reported in Germany’s NAKO study (32).

Other population groups were also disproportionately affected by the pandemic—with differential effects by age and sex, in line with other studies in other European populations, and in Austria (6, 33). Overall, those aged 18 to <35 years who were most likely to report a moderate or severe overall impact effect of the pandemic, despite older age groups being at greatest risk for serious illness and mortality caused by COVID-19. For example, at Ischgl-1, only 2% of all participants found to be seropositive had been hospitalised because of COVID-19, yet 14.8% of seropositive participants aged over 70 were hospitalised. In November 2020, the oldest age group was more likely to report declines in SRPH since November 2019 than other age groups, but these declines did not appear to be directly exacerbated by SARS-CoV-2 infection. In a stratified analysis, no relationship was found between seropositivity and a decline in SRPH among this age group. However, those over 70 years of age were far less likely to be concerned about their economic future than the other age groups, and less likely to report any declines in mental health. Declines in mental health and current poor MH were most acute in the middle age groups (≥35 to <50 and ≥ to <70 years) which is in line with the SHARE study when comparing to older age groups (34).

Women were more likely to report a moderate or severe effect of the pandemic than men. In line with the Traunmüller study in Austria, we also found women reported more worsening of their mental health (6). Of the 13% of participants reporting poor SRMH, the majority were women, in line with pre-pandemic studies which have consistently shown women tend to report higher rates of psychological distress than men, reasons for which remain unclear conclusive but have been linked to differences in exposure to risks for developing, vulnerabilities to, experiences of, and responses to mental health between genders (6, 35). Women in our study were also more likely to report declines in their mental health than men. Other studies have linked the disproportionate effect of the pandemic on women’s mental health to be associated with changes in social networks, social activities and contact and social isolation, alongside changes in care-giving responsibilities including a greater responsibility for homeschooling during lockdowns, despite potential other benefits from working from home (10, 36). Further research would be needed to understand these factors in the Austrian context, and it is important to highlight our study lacks data on wider social networks that are clearly integral to both the wider impacts of NPIs and understanding how NPIs impact people. Indeed, the fact that our study population experienced the same NPIs and there were no gender differences in laboratory-confirmed seropositivity suggests it is not NPIs or SARS-CoV-2 infection, per se, that has led to this disparity.

The strength of this study is that it is based on a population with laboratory-confirmed SARS-CoV-2 status, all infected in the same outbreak and experiencing the same waves of the pandemic and NPIs. Nonetheless, it has some limitations. It relies on self-reported data, which introduces the possibility of recall bias concerning their health, behaviours, or economic status. Also, while the situation in Ischgl provides an unusual opportunity to undertake such a study, as a community heavily reliant on tourism the generalisability of the findings is limited. Another limitation stems from the timing of data collection. The follow-up survey occurred in November 2020, during the second wave of the COVID-19 pandemic, and will not capture the long-term effects of the virus or the non-pharmaceutical interventions used to control its spread. Also, while we examined some behaviours directly related to health, there may be others, such as gambling, that could have been impacted by the pandemic. Our health assessments relied on single-item measures for physical and mental health, which, while widely used, may not fully capture the complexity of health changes experienced by participants. Additionally, variations in the intensity and timing of infection waves and NPIs within Austria could introduce variability not fully addressed in the analysis. Finally, we lacked data on social networks, which may be relevant given the impact of social connectivity on mental health during periods of restrictive measures.

This study reinforces recommendations that have been made by others in the light of experience during the pandemic. Mental health emerged as a critical concern, particularly for women and middle-aged adults, many of whom reported declines in well-being. Pandemic preparedness should include the ability to scale up online counselling and other forms of remote support, some of which could be delivered by volunteers with basic training and appropriate safeguards. Economic strain, especially for workers in tourism-reliant regions like Ischgl, highlights the need for financial aid and job retraining programmes that can be deployed during and after a pandemic. Looking ahead, especially in the light of climate change and its impact on snow cover, diversifying local economies in ski resorts will offer longer-term stability. For younger adults who expressed concerns about job security, initiatives like internships, apprenticeships, and guaranteed job opportunities can help stabilise their professional trajectories. There is a clear need to monitor individuals with persistent physical symptoms.

Preparedness for future pandemics must include tailoring non-pharmaceutical interventions to specific community contexts, aiming to minimise mental health and economic impacts. Collecting appropriate data on social, psychological, and economic outcomes during health crises is vital for holistic response plans. Gender-sensitive policies should also be prioritized to address the disproportionate burdens on women, such as expanded childcare support, flexible work arrangements, and mental health initiatives tailored to their needs.