Overview of main findings

This review and meta-analysis aimed to provide an overview of the secular trends in mental health problems among young people in Norway, by synthesizing data from all large-scale, repeated cross-sectional surveys. Seven surveys were included, covering 35 measurement points between 1992 and 2019, with a total sample of 776,606 young people. Our study demonstrates an increase in self-reported mental health problems among young females in Norway over the past few decades, while trends are less marked for males. In the individual surveys, the difference in the proportion of individuals scoring above the problematic score threshold from the first to the last survey year was, on average, 11.2% for females and 5.2% for males. Pooled meta-regression estimates showed that from 1992 to 2019, the mean symptom scores increased by approximately 17% for females and 5% for males.

A comparative perspective on the trends in mental health problems among young people

Our findings align with prior studies from other high-income countries in Europe [3, 37,38,39,40,41,42,43] and North America [44, 45]. Combined, these studies show that self-reported mental health problems among young people have increased over the past 3 decades, particularly among females. The findings for males are more mixed, however, with certain studies identifying increases over time [39, 42, 44] and others finding more stable or decreasing trends [37, 38, 41]. This was also the case in our study, as we pinpointed shorter periods indicating flat or even declining trends in some surveys (e.g., see [13]). However, over a longer time period, the studies all show a general increase over time and secular trends in self-reported mental health problems coincide with increasing rates of treatment for mental disorders among young people [5, 46,47,48,49].

Other studies from other country contexts have noted trends of mental health problems among youth may depend on survey characteristics like sex, age, study period, country, and outcome measures [1,2,3, 43]. For example, a review of 36 UK surveys of children and youth found large variations both within and between surveys based on survey characteristics, such as country, time, age, and outcome measure [43]. The researchers observed an increase in long-standing mental health conditions but a stable trend for measures of psychological distress and emotional well-being. In line with Pitchforth and colleagues [43], our findings also indicated substantial variation between comparable national surveys, which could significantly affect precision if not appropriately addressed. However, even after accounting for between-survey characteristics and comparing our findings with other mental health outcomes, such as reports of increase of diagnoses of mental health disorders, increased use of healthcare services, increased use of antidepressants, and increasing rates of self-harm among Norwegian youth [15,16,17], the evidence consistently shows deteriorating mental health among young people in Norway, particularly among young females, between the 1990s and the present.

Other extensive comparative studies have also emphasized the need for caution when comparing cross-national trends, given the substantial variation within and between nations, even when employing similar mental health outcomes [3, 6]. Country-specific societal factors, such as policy or economy difference, have the potential to strongly influence or mediate the trajectories of mental health problems among youth populations, thus leading to genuine cross-national differences [50]. However, as our findings show, in line with Pitchforth and colleagues [43], survey characteristics substantially influence outcome variation even within nations. This underscores the need for consistent methodology, identifying the best methodological practices, and the use of a common instruments to enhance precision in country-specific estimates of mental health problems. Additionally, improving methodological consistency within countries can feasibly enhance validity when comparing mental health problems cross-nationally in the future.

Has there been a general increase in mental health problems among young people?

The secular trends identified in the previous studies and in our own are exclusively related to internalizing problems as opposed to externalizing problems. Studies that have assessed both externalizing and internalizing problems suggest that only the latter have increased over time, among young people [38, 40, 51]. Therefore, the observed increase in mental health problems appears to be specific to symptoms of anxiety and depression (e.g., see [37]). This is also mirrored in the rates of treatment for mental disorders, where the proportions treated for externalizing problems have decreased over time [17, 47]. Unfortunately, the surveys included in our review did not include standardized and comparable measures of externalizing problems. However, there are no indications otherwise that externalizing problems have increased in Norway the past decades—if anything available evidence (statistics produced publicly by Norwegian health registries) suggests that such problems have decreased in the youth population [17]. Furthermore, the increase in mental health problems is related to age and more pronounced among older youth, compared to child and adult populations [10,11,12, 39, 44]. Studies of children under the age of 11 generally show no increasing mental health problems over time [1, 2, 11, 52]. None of the surveys included in our review included children under the age of 13 years, so we were unable to investigate trends in this age group further. A recent Norwegian population study found that mental health problems increase over the last decades among young people, but not for adults [10]. In fact, declining rates of mental health problems were evident among those aged 60 or older. This may suggest that increasing mental health problems are specific to internalizing problems among cohorts of female youth, as opposed to constituting a broader phenomenon.

Potential causes of the trends in mental health problems

Several explanations for the increasing trends in mental health problems both in Norway and internationally have been proposed; changes in health-related behaviours, frequent social media use, increasing school-related stress, and greater willingness to report symptoms of ill-health.

To our knowledge, only two publications, which utilize the included survey data, have empirically investigated a range of potential causes for the highlighted Norwegian trends in our review. One of these [13] suggests that the increase in self-reported depressive symptoms among boys and girls from 1992 to 2002 could be partially attributed to increases in eating problems and cannabis use. Reduced satisfaction with own appearance also appeared to contribute, particularly among girls. The second study [14] found that increase of self-reported eating problems, such as bulimia nervosa symptoms and food preoccupation, over time could be partially linked to appearance satisfaction, alcohol intoxication and global self-worth. However, these studies generally found that these proposed mechanisms only account for a small part of the increasing trajectories of mental health problems between 1992 and 2010. Additionally, other mechanisms might underlie the continued increase in mental health problems after 2010.

In contemporary discussion around deteriorating mental health among young people, the “social media hypothesis” is garnering the most attention. The social media hypothesis posits that excessive social media use might be the major cause of the surge of mental health problems among youth, after around year 2007 [53]. Recent reviews do indeed suggest that there might be a weak association between social media and mental health problems [54,55,56]. However, there is a lack of evidence to establish whether this association is causal [57]. An important prior study leveraged a natural experimental design and reported that the introduction and expansion of social media, specifically Facebook, in American student communities in 2004–2006 had an adverse impact on their mental health [58]. However, this study used data collected during the emergence of social media and there is a lack of similar studies that capture the past decade’s developments in social media platforms. Additionally, there is some evidence that the negative effects of social media use are more pronounced for females than males [59, 60]. This seemingly fits the pattern that mental health problems have increased more among females than males. Nevertheless, the effect of social media use on mental health remains a subject of the ongoing debate and there is a need for future research that can establish the degree to which social media use can account for the increasing trend of mental health problems observed in Norway and other high-income countries.

In light of the current study, social media also cannot account for the observed increase in mental health problems that occurred before the huge expansion of social media platforms. The decades prior to 2007 were also characterized by increasing screen time in relation to e-mailing, Internet browsing, and computer gaming. Such activities have been associated with poor perceived health (mediated through negatively affecting sleeping habits) [61] and negative physical complaints [62]. It is important to account for the increases in mental health problems prior to the emergence of social media as well, as several studies including our own, indicate that the deterioration of mental health among young people was evident even before the emergence of social media (e.g., [63]).

Others have suggested that school-related stress and pressure have contributed to the mental health trajectories [64]. This has been spurred by observations that stress and pressure related to school have increased over the past decades [6, 42, 65]. Large comparative studies of 43 countries in Europe and North America demonstrated an increase in schoolwork pressure and its association with increased mental health problems over time [6, 65]. However, these associations were generally modest. Another study [66] found that the effects of school stress on psychosomatic symptoms became stronger over time in the period between 1993 and 2017, but that school stress only partly explained the increase in such symptoms. Interestingly, one study showed that the association between school stress and mental health problems was stronger for countries that were richer and more educated, suggesting that societal factors may influence trends of mental health problems [50]. The aforementioned studies focus on a relatively recent time period (i.e., from 2000 and onwards). A less recent study conducted in Scotland [67] showed that school disengagement and worries about school were among the explanatory factors most strongly associated with the increase in mental health problems in the time period from 1987 to 2006. This suggests that the association between school-related variables and mental health trajectories may have persisted over the last 3 decades. Despite the evidence that support the notion that school-related stress and pressure may have contributed to the rising trajectories of mental health problems, such contributors only account for a relatively small portion of the trends.

It has also been suggested that the increasing trend of mental health problems could be “inflated”, due to reduced societal stigma and a subsequent increased willingness to report symptoms [2]. If willingness to report mental health symptoms had changed over time, one would expect to find signs of factorial invariance across time, when assessing the psychometric properties of the instruments used, which at present, does not seem to be the case [13, 68]. Moreover, a study conducted in the UK found that improvements in attitudes toward mental illness did not mirror changes in self-reported mental health problems across English regions over the past decade [69]. In addition, evidence from experimental studies suggests that training youth to recognize symptoms of mental health problems does not influence mental health problem outcomes, at least in a controlled setting [70].

While conclusive causal analyses regarding deteriorating mental health in Norway are lacking, several concurrent societal trends might have contributed to the observed increase in mental health problems. Recent national studies indicate that young people are dedicating more time to digital screens and social media, and subsequently spend less time with friends [9]. Additionally, negative attitudes toward school have increased during the past decade [9]. Moreover, there has been a notable reduction in the stigma surrounding mental health problems in recent decades, encouraging today’s youth to be more open about such issues compared to older generations.

Despite increased efforts to understand the determinants of increasing mental health problems, there is still a need for future research to extensively examine the potential causes. Considerable challenges persist in establishing causality between explanatory factors and secular trends in mental health problems. Where associations do exist, they are generally small and feasibly account for only a small portion of the total increase in mental health problems over time. It seems more and more unlikely that there is a single catalyst for increasing mental health problems among young people, but rather that several determinants working together to drive the negative trend in Norway and several other high-income countries.

Limitations

A major strength of this study is that it includes a sample of over 770,000 to examine the trends in youth mental health problems in Norway, which provides a solid evidence base for public health decision-makers. However, several limitations should be noted. First, based on the protocol for this study, we planned to include socioeconomic status and minority background in our analyses. Unfortunately, we were not able to include comparable measures of socioeconomic status and minority background as explanatory variables in our data extraction and analysis, due to the data being missing or not comparable across surveys.

Another limitation is that despite the individual surveys included in this study having both comparable designs, and being drawn from the same population, there were still substantial variations between them. Other factors not directly examined in this study, that might have contributed to between-survey differences could be (a) the difference in the response rate between surveys and the change in rates over time, (b) the sampling procedure and efforts to control for low initial response rates, (c) and the length and content of the different survey questionnaires (e.g., various versions of the main outcome measure). This does, however, emphasize the need for a common instrument and consistent methodology when operationalizing mental health problems. Further research into psychometric properties, item functioning, and other validation work on youth health measures can provide further insights into the current youth trends.