The study is based on data from the population survey “Health on equal terms?” conducted in collaboration with the Public Health Agency of Sweden [31]. The national survey started in 2004 and has been carried out every two years since 2016 to monitor the health of the population in Sweden. The age group addressed is 16–84 years. The sample frame is the total population register at Statistics Sweden, the statistical administrative authority in Sweden, covering all inhabitants in the country. The national simple random sample in 2020 included 40,000 persons.
The present study is based on data from one county (Värmland) where an extended simple random sample was drawn. In total, the questionnaire was sent to 5091 persons in the county and 2273 individuals answered the questionnaire giving an overall response rate of 45%. The questionnaire was postal but could also be answered online. Data collection was discontinued after two postal reminders. In Värmland county, the first COVID-19 cases were reported on 6th March 2020 [31]. To define those who replied before and after the COVID-19 outbreak in Sweden the respondents were divided into early (n = 1711) and late (n = 562) respondents, i.e. those responding between 3th February and 11th March 2020, and those responding between 12th March and 5th May, respectively. The date 11th March coincided with posting the first reminder of the survey.
Värmland county is situated in the west of Mid-Sweden, bordering to Norway, and comprises about 282,000 inhabitants. It includes one bigger city with over 90,000 inhabitants and 15 smaller municipalities. The incidence of COVID-19 was lower in Värmland than in Sweden in general during March-May 2020 and by the last week in May 533 persons had been diagnosed with COVID-19 in Värmland [31].
The measures taken to combat the COVID-19 pandemic in Sweden included e.g. recommendations to keep distance to other people, to wash hands often, to stay at home when having symptoms of flu, to avoid travelling abroad and unnecessary travelling in Sweden, to avoid public places with crowds, and to work from home when possible. Those over 70 years of age were recommended to refrain from seeing people outside their own family. In the end of March, public gatherings of more than 50 persons were forbidden. No total lockdown was, however, instituted in Sweden.
To explore whether the results observed in 2020 are due to the COVID-19 pandemic, the same analyses were run in the corresponding “Health on equal terms?” survey which was carried out between 28th February and 18th June 2018. In total, 2142 persons aged 16-84 years responded to the survey in Värmland county with an overall response rate 42%. Out of these, 1660 individuals responded before (early respondents) and 482 after (late respondents) the first reminder sent on 10th April 2018.
Confounding variables
Information on gender, age, level of education and country of birth are based on register data from Statistics Sweden. Educational level was categorised into three levels: compulsory education, secondary education, and postsecondary education. Country of birth was dichotomized into those born in Sweden and those born outside Sweden.
Outcome variables
Living conditions
Social support was measured with the question “Do you have anyone you can share your innermost feelings with and confide in?” (yes/no).
Economic difficulties were estimated with the question “During the last 12 months, have you ever had difficulty in managing the regular expenses for food, rent, bills etc.?”. The response options were “no”, “yes, once”, “yes, more than once” where the last two categories were combined to yes.
Trust in other people was measured with the question “Do you think that, in general, people can be trusted?” (yes/no). Employed people were defined as being worried about losing their job if they answered “yes” to the question “Are you worried about losing your job in the coming year?”
Lifestyle factors
Two questions for measuring physical activity were used. The first question was: How much time do you spend in a normal week on physical training that leaves you out of breath – for example running, fitness training, or ball sports? The response options were: 0 min/no time; less than 30 min; 30–59 min (0.5–1 h); 60–89 min (1–1.5 h); 90–119 min (1.5–2 h); 2 h or more. The second question was: How much time do you spend in a normal week on daily activities – for example walking, cycling, or gardening? Count all time together (at least 10 min at a time). The response options were: 0 min; less than 30 min; 30–59 min (0.5–1 h); 60–89 min (1–1.5 h); 90–149 min (1.5–2.5 h); 150–299 min (2.5–5 h); 5 h or more. These questions are used to measure whether the respondent reaches 150 activity minutes per week as recommended by the WHO. The number of minutes from the physical training and daily activities were summed together, with the number from the first variable counting double [32].
Sitting duration was asked with the question “How much do you sit during a normal day, not counting sleep?” The answer categories were dichotomised into those who sit less than 10 h and those who sit at least 10 h a day.
Smoking was measured using the question “Do you smoke” (“no”, “yes, sometimes”, “yes, daily”).
Alcohol consumption was measured using Alcohol Use Disorders Identification Test-C (AUDIT-C). AUDIT-C is a widely used and validated screening instrument of alcohol use. It comprises three questions on the frequency and quantity of alcohol consumption. We used the following cut-offs for risk-drinker: 6 or more points in men and 5 or more points in women [32].
Health
The following variables were used to measure the respondents’ health [32]. Self-rated health (SRH) was measured with the question “How would you describe your health in general?”. Response options were very good, good, fair, poor and very poor. In the statistical analysis the options were dichotomised into good (very good or good) and poorer than good (fair, poor or very poor) SRH.
Illnesses were measured with the following question: Do you have any of the following illnesses (with answer options No; Yes, but no discomfort; Yes, minor discomfort; Yes, severe discomfort)? Illnesses included high blood pressure, and the last three categories were combined to Yes.
Symptoms were derived from the question: Do you have any of the following discomforts or symptoms? These included “aches in the shoulders or neck”, “sleeping difficulties” and “anxiety or worry”. The answer categories were No; Yes, minor discomfort and Yes, severe discomfort, where the two latter categories were combined to Yes.
Stress was measured with the question “Do you feel stressed at present? By stressed, we mean a condition where you feel tense, restless, nervous, uneasy or unable to concentrate.” The answer options were Not at all; To some extent; Quite a lot and Very much, where the last three options were defined as having stress.
Ethical considerations
The study followed the Swedish guidelines for studies in social sciences and humanities, in accord with the Declaration of Helsinki and the data are protected by the law of official statistics. The participants were informed that completed questionnaires would be linked to the Swedish official registries through personal identification numbers, to access registry information on gender, age, country of birth and educational level. The respondents thus gave their informed consent to the linking of registry data. The personal identification numbers were deleted before the data was delivered to Region Värmland. Statistics Sweden carried out the sampling, data collection and linkage with registry data and delivered the de-identified data. The study was approved by the Swedish Ethical Review Authority (Dnr 2020–04202).
Statistical analysis
Differences in the distribution of background characteristics and SRH between early and late respondents were tested using chi-square statistics. Difference in mean age was tested using independent samples t-test. P-values < 0.05 were considered as statistically significant. The differences in living conditions, lifestyle habits, and health between early and late respondents were studied using multivariate binary logistic regression, with early/late response as the independent variable (reference category = early response), adjusting for background characteristics gender, age group, educational level, and country of birth. The results are reported as odds ratios (OR) and 95% confidence intervals (95% CI) for each living condition, lifestyle habit and health condition as outcome at a time. All analyses were conducted in IBM SPSS Statistics, version 26.