Data collection

The surveys are conducted by I&O Research and commissioned by the GOR Network. Sample selection and data collection for research round 4 took place between 25 May 2022 and 7 June 2022. In total, 2,227 I&O Research panel members were contacted (aged 16 – 25 years). Another 3,833 panel members with children living at home were asked to participate. Parents who gave permission for their children to participate were able to share the questionnaire with their children. Quality checks were carried out to clean up the dataset, leaving a total of 4,179 questionnaires completed by young people (aged 12-25 years) participating in research round 4.

The results from previous research rounds are also posted on the RIVM website.

Participants in research round 4

The gender distribution of participants was 39% male and 61% female. 30% of the participants are aged 12 to 17 years and 70% are aged 18 to 25 years. The majority (70%) live with their parent(s). The province of South Holland has the highest number of participants (22%). The provinces of Drenthe (2%) and Flevoland (3%) are least represented.

Materials

Data collection took place online using a pre-programmed survey. The survey questionnaire was compiled by the GOR Network and consisted of various questions with a wide range of answer categories. The following topics were covered:

• socio-demographic features
• experiences during the COVID-19 period
• perceived impact of the coronavirus measures
• general health
• mental health
• physical health
• care and support needs
• delayed care
• knowledge integration for policy

Representative sample and weighting

The sample was disproportional with respect to province, meaning that each province was sufficiently represented, regardless of the size of the province. We tried to find at least 200 participants per province. Following data collection, we corrected for deviations in representative sample by assigning weighting factors based on sex, age, region and level of education; we then compared the composition of the sample to that of the population of the Netherlands (derived from the ‘Golden Standard’ used by Statistics Netherlands – CBS).

Data analysis

The website offers weighted cross-sectional figures. Where possible, the text is also accompanied by trend graphs that show the responses to specific topics on previous surveys. This helps us provide insight into the responses of Dutch youth in terms of mental and physical health, support needs and perceived impact of the measures.

In addition, we assessed the factors related to a higher risk of negative impact on the health and well-being of young people. We also checked for factors that had a shielding effect against negative health impacts, by using multivariable regressions. This analysis assessed the potential effect of socio-economic characteristics (sex, age, level of education), whether participants were still suffering from experiences related to COVID-19, and social activities. Such assessment considers multiple factors simultaneously. In other words, we are not only looking at whether loneliness affects mental health, but also whether loneliness is still significant if experienced events and confidence in the future are also taken into account. In previous rounds, we used this same method to look at the increased risk of suicidal thoughts. Since June 2022, we have been doing this for all topics. Considering the sample size, we kept to a significance level of 0.1% (p < 0.001). A wider significance level (p < 0.05) was used for PTSD, since these questions were answered by fewer than 650 participants.

We checked to see if variables were too strongly interconnected to be included simultaneously in a multivariable regression. This applies, for example, in the context of age and living situation: a 13-year-old almost always still lives at home, while young people aged 18 years and older often no longer live with their parents. Therefore, living situation was not included in order to ensure a reliable analysis.  We modified the regression analysis for all topics related to mental health. Age and gender were also left out of that analysis. This was necessary because there was too much overlap between girls of all ages and young people aged 18–25 in terms of more frequent mental health impact. Although it was more common among girls of all ages and young people aged 18–25, there was no strong correlation between these factors and negative mental health impact. We carried out a univariable regression for the need for support.

Trend charts

Starting from research round 4 (June 2022), most results are accompanied by trend charts. The objective of this study is to monitor trends in mental and physical health among the population of the Netherlands during the COVID-19 pandemic. Trend charts provide such insights. The benefit is that it shows how questions about a specific topic were answered differently over time. For example, it is possible to see when people started developing specific symptoms more often, or which support needs people still have.

We are showing trend charts for all topics for which we have collected data using the same method during 3 or more rounds. That means there is no trend graph to show for loneliness, because rounds 1 and 2 used a different tool than rounds 3 and 4. Similarly, trend charts are not available for symptoms of post-traumatic stress disorder (PTSD) or for searching for and finding help, because there have not been enough surveys on these questions (2 rounds instead of the 3 required for a trend chart).

About surveillance for the Nivel Primary Care Database

The surveillance figures from the Nivel Primary Care Database are intended for use in detecting infectious diseases and other health conditions in the general population of the Netherlands, but can also be used to monitor other acute symptoms. This resource provides weekly figures on symptoms and conditions among the Dutch population based on anonymised data from electronic medical records kept by GP practices. The aggregated figures show how many people consulted their GP for specific symptoms or health conditions in that week, divided by the total number of patients registered in that practice (prevalence). The data comes from about 380 GP practices, offering healthcare services to roughly 1.6 million registered patients (9% of the Dutch population). Health data on these symptoms and conditions is routinely recorded, based on the International Classification of Primary Care (ICPC), version 1. This ICPC-1 system is used by all general practitioners in the Netherlands.

Acute health problems/conditions that may be related to the COVID-19 pandemic (whether directly or indirectly), as presented to GPs, have been mapped. In this study, we focus on a number of non-specific symptoms (designated as “not otherwise specified” or NOS) that could have a physical or mental cause. This includes both symptoms that are recorded in the context of episodic health conditions and symptoms that are not associated with a specific health condition. Some of these symptoms are listed by RIVM as possible persistent symptoms after an infection with the coronavirus SARS-CoV-2. However, it is not possible to identify any direct link, since SARS-CoV-2 infection is not recorded in the data for this study. See the table below for the symptoms included in the study and the corresponding ICPC-1 codes.

List of non-specific symptoms and corresponding ICPC-1 codes

* Relevant for consideration of Long COVID, according to the RIVM list

It is relevant to mention here that young people in the Netherlands are allowed to visit the GP alone (unaccompanied by parent/guardian) from the age of 12. However, they do require permission from a parent/guardian until the age of 16. That means that for the younger age groups, the person experiencing the symptoms is not always solely the one who decides whether or not to visit the GP. However, since this has been the case in every year of this study, it does not explain differences over time.