Introduction
Mental health, a state of wellbeing that allows the individual to cope with stressors of everyday life and function productively [
1], is characterized to a large part by the absence of mental distress and mental disorders. Both mental distress and mental disorders have become major public health concerns affecting the quality of life, work productivity, physical illnesses, and life expectancy of a large proportion of the general population [
2]. While mental distress refers to distress in response to an external stressor and can be characterized by e.g., (symptoms of) anxiety or depression [
3], psychological disorders consist of a pattern of persistent behavioral or psychological symptoms that influence several areas of life.
Mental health is subject to different temporal trends: it may vary depending on age, time period, and birth cohort. Moreover, mental health differences are frequently found between women and men, and between regions. When analyzing temporal trends, it is important to note that age, period, and cohort effects are highly related to one another. Age effects refer to developmental or age-specific transformation, the general pattern of individual transformation from childhood through adulthood and old age that are consistently noted in all birth cohorts and across all time periods [
4]. Period effects, on the other hand, describe differences among individuals due to historical events that leave unique imprints, no matter the age. Finally, cohort effects refer to differences among individuals categorized by their time of birth, they share critical formative moments and similar socialization experiences with their respective birth cohort [
4]. Ideally, age, period, and cohort effects are examined simultaneously. However, due to the exact multicollinearity and conceptual relationship between age, period, and cohort, it is difficult to correctly estimate these effects. Studies apply different methodical frameworks based on specific theory-based assumptions to disentangle these effects [
5]. Thus, comparisons between studies are feasible only to a limited extent.
With regard to age effects, a Dutch study revealed self-reported prevalence of mental illnesses to be lower and general mental health to be better for the elderly [
6]. Similar results are found in a study in the USA; using the Composite International Diagnostic Interview (CIDI), lifetime and recent major depressive episodes are less prevalent among respondents of 65 years and older [
7]. However, a longitudinal survey study covering a 15-year period found a U shape for depressive symptoms with the highest symptoms burden for the age group 25–35 years and from 75 years onwards after controlling for cohort effects [
8]. A study including 27 European countries assessed self-reported depressive symptoms and also found the highest prevalence of the current depressive disorder among persons 75 years and older [
9]. A hierarchical age-period-cohort analysis (HAPC) on the life course trajectory of mental health from the UK partly confirmed this finding, as it revealed mental health increases throughout the life-course, but slows during middle-age and worsening again slightly in older age [
10]. In Germany, the prevalence of current depressive symptoms (self-reported) was highest among 18–29-year-olds and decreased with age, whereas the lifetime prevalence of diagnosed depression was highest among 60–69-year-olds [
11]. Among the German elderly (53–80 years), a U shape for the prevalence of depressive symptoms was found [
12].
Period and cohort effects are also found in mental health studies. In a study applying HAPC models, recent birth cohorts in the UK generally reported worse mental health [
10]. A study from 1993 examining age and cohort effects for the occurrence of depression in a US sample reported that the birth cohort 1950–1959 had the lowest age of a first episode of depression [
13]. An age-period-cohort (APC) analysis in Canada and USA revealed the highest levels of psychological distress in the oldest (born before 1939) and more recent (born 1989–1992) birth cohorts [
14]. With regard to period effects, a systematic review and meta-analysis addressing the period 1980–2013 reported the highest prevalence estimates of common mental disorders worldwide in studies undertaken in the 1990s [
15]. In Canada and the USA, levels of mental distress were highest around 2000 [
14]. In the USA, reported depressive symptoms were highest between 2000 and 2010 [
16]. A German study applying HAPC models to examine age-period-cohort trends in depressive symptoms found a U-shaped cohort effect where cohorts born around 1930 until 1950 exhibited less depressive symptoms compared to earlier and later-born cohorts [
17].
Within Germany, differences in mental health are found between the former eastern and western federal states. Founded after World War II (WWII), the two German States existed from 1949 to 1990. They evolved with contrary and antagonistic political and economic systems. The federal republic of Germany followed the (capitalist) system of the Western European countries, whereas the German Democratic Republic (GDR) followed the example of the Russian-Soviet (socialist) system. The socialization processes and living conditions were therefore extremely different between the former East and West Germany, leading to different risk and protective factors regarding mental health. While many people suffered from political persecution and repression, leading to increased somatic symptoms, anxiety and depression later in life [
18], other system-related factors like a low official unemployment rate and increased social mobility could be regarded as protective factors for mental health. The re-unification was accompanied by drastic changes in almost all aspects of life; while average income has been increasing in East Germany since 1990, they remain lower compared to West Germany [
19,
20], and the unemployment rate 20 years after re-unification was still twice as high in East compared to West Germany [
21]. Regarding demographic characteristics, the East German population was reduced by 16% since re-unification [
19], especially young people and women left East Germany. Even though the inner German migration has aligned now, the East still consists of an older population strata [
19] and has a lower life expectancy than West Germany. These sociodemographic and socioeconomic differences are reflected in birth cohorts.
With regard to mental health, evidence shows that results for East and West Germany strongly differ between survey year and mental health outcome. A study examining psychological distress and mental disorders in former East and West Berlin one year after the fall of the Berlin wall did not find differences in ICD-10 diagnoses [
22]. However, a large nationally representative survey reported higher prevalences of mental disorders (assessed with CIDI) in the Western compared to the Eastern states in 1998/1999 [
23] concerning depression, somatoform disorder, substance abuse, eating disorders, and social anxiety. Studies comparing mental health between East and West Germany 10 years after reunification found no differences in mental health between participants residing in East and West Germany [
24], whereas life satisfaction was higher among West Germans compared to East Germans [
25]. Perceived stress did not differ between formerly East and West Germany 20 years after reunification [
26], whereas the prevalence of depression diagnosis was found to be lower in East compared to West Germany [
27]. Another study, however, did not confirm these differences for adolescents [
28]. Evidence for mental health differences between East and West Germany is thus inconclusive. An analysis covering a longer time frame and using representative data from Germany is thus needed to shed further light on the temporal trends in these two regions.
Differences in mental health exist between women and men. For most internalizing disorders (e.g., major depression [
29] and anxiety disorders), women are more frequently affected than men [
30,
31], whereas for externalizing disorders (e.g., substance abuse) men are more frequently affected [
32]. Representative German studies have shown that approximately one in three women and one in four or five men had a diagnosis of a mental disorder in the previous 12 months [
33]. Sex differences in mental health can be explained by hormones [
34,
35] and dysregulations in the hypothalamic–pituitary–adrenal (HPA) axis [
34,
36], especially for stress-related mental disorders. Gender differences in mental health can be explained by e.g., gender-based violence [
37], low self-esteem [
34], and differences in risk behavior and identification of disease symptoms [
38]. Hence, sex and gender interact in the development of diseases [
39,
40]. Myocardial infarction and depression are gender-stereotypical diseases [
41]. Myocardial infarction is known as a typical disease for men and is therefore often overlooked for women [
40], while depressive disorders are considered a typical disease for women and underdiagnosed for men [
41].
Age effects in mental distress differ for women and men. Results from an Australian study reveal a consistent decrease in mental distress for women, whereas for men, the decrease only starts in late adulthood [
42]. In the USA, women in all age groups report depression more often than men and this gender gap increases in adulthood [
43]. Regarding period effects, a Swedish study showed that the prevalence of self-reported anxiety increased between 1980/81 and 2004/5 for women and men in most age groups, except for men aged 64–71 and women aged 56–63 [
44]. Additionally, cohort effects were found; for men, anxiety increased from birth cohort 1942–40 onwards, while for women, this increase was already observed from birth cohort 1926–33 and stagnated with birth cohort 1974–81 [
44]. With regard to period trends, a British study shows increased mental distress especially for women between the years of 1991 and 2008 [
45]. Moreover, an Australian study revealed increased mental distress between 2001 and 2017 for both women and men [
46]. The prevalence of diagnosed depression increased in Germany between 2009 and 2017, especially in young men [
47]. Once again, none of these studies estimated age, period, and cohort effects simultaneously with regard to differences between men and women.
Discussion
Findings on the mental health of residents from former Eastern and Western Germany have been contradictory. Analyzing repeated cross-sectional data from representative German surveys spanning 15 years, this study used HAPC models to disentangle age, time period and birth cohort effects on mental distress while testing for gender and regional effects (i.e. former Eastern and Western German states).
Findings revealed significant period effects. Research has shown that global public health is closely linked to political, economic, and social determinants [
65]. In this study, peak values for mental distress were found in the years 2017 and 2020 for both women and men and in East and West Germany. Important political events can affect mental health in general and serve as an explanation for these peaks.
For the increased level of mental distress in 2017, one explanation can be found in several political and demographic upheavals taking place in the previous years. In a referendum in 2016, the United Kingdom decided to leave European Union. Furthermore, in November 2016 Donald Trump was elected president of the United States of America. These two events exemplify a wave of right-wing anti-globalization politics, which has risen in much of the Western world [
65]. Shortly thereafter, in 2017, elections in three west European countries (France, the Netherlands, and Germany) took place and one of the main campaign issues was the alliance with the European Union. The discussions about this topic destabilized democratic cohesion, with antagonistic groups in society forming around this issue, further paving the way for right-wing parties such as the German
Alternative für Deutschland [
66]. These political movements were partly set in motion through the so-called “refugee crisis” in 2015 und 2016. In those years, large numbers of refugees from countries such as Syria, Iraq, Afghanistan, and Somalia applied for asylum in European countries which was shown to destabilize democracies in Europe [
67] and which strongly contributed to the success of the radical right and right-wing populist parties [
68]. This could have caused an increase in mental distress, since countries with a liberal democratic political system report on average more positive results on the population’s physical and mental health indicators [
69]. Furthermore, due to the perceived threat associated with the “refugee crisis” and democratic instability, quarrels and protests increased. A systematic review from 2020 revealed compelling evidence that protests, also nonviolent, can be associated with adverse mental health outcomes, especially major depression [
70].
The second peak in mental distress was found in 2020, which is likely to be caused by the COVID-19 pandemic and its first social lockdown restrictions. Literature has shown that this pandemic increased psychological health problems in Germany as in other countries. A systematic review and meta-analysis revealed that especially the prevalence of depression, anxiety, and distress increased during the pandemic [
71]. A meta-analysis examining longitudinal cohort studies showed a small but significant increase in mental health symptoms early in the pandemic [
72]. The survey included in this study took place at a similar time. However, effects for e.g. anxiety (but not depression) disappeared by mid-2020 and were comparable to pre-pandemic levels [
72]. Furthermore, a large British study examining anxiety and depression symptoms found a decrease in symptoms during the first 20 weeks following the initial lockdown [
73] Findings of this study came from an online panel and therefore could be biased [
74,
75]. In German population surveys an increase in scores for depression and loneliness during the COVID-19 pandemic compared to scores in 2018 was found [
76,
77], even though certain parts of the population (e.g. women, young people, high-risk of poverty) were affected most strongly [
77,
78].
In addition to a period effect, a significant cohort effect was found. Mental distress was highest in the oldest birth cohort, born before 1946. This cohort experienced WW II and the hardships of the post-war era. The elderly have experienced higher lifetime trauma exposure and PTSD prevalence than younger persons [
79‐
81]. Hence, this group may still suffer from mental health problems related to the traumatic WWII experience 50 years after the end of the war [
82]. Traumatic events, especially war related, are highly connected to depressive symptoms [
83,
84]. Furthermore, people from the birth cohorts 1946–1959 and 1960–1969 showed lower mental distress compared to people from the oldest birth cohort. However, they reported higher mental distress compared to people from the youngest two cohorts. A possible explanation for this could be experiencing the negative consequences of the transformation of the system after Germany was re-united, i.e., unemployment [
85], or other economic and social differences [
19], this applies above all to the former East German population.
Age in itself did not affect mental distress when cohort- and period effects were considered. Age effects found in other studies could therefore be merely a result of cohort or period effects. Another explanation could be found in the measure of mental distress in this study. PHQ-4 measures mental distress based on core depressive and anxiety symptoms. Previous research revealed depression to be less prevalent among older adults [
7]. However, generalized anxiety disorder was shown to be higher among older age groups compared to younger age groups [
86]. Therefore, the insignificant effect of age in this study could be caused by the different directions of the effects within our outcome variable.
Women reported significantly higher mental distress than men, which is in line with previous studies reporting more internalizing disorders (e.g., depression and anxiety) for women [
87,
88]. This was consistent over the survey years, also in times of crisis. Other studies also confirmed this, e.g., women reporting higher mental distress during the COVID-19 pandemic than men [
89,
90]. Unlike previous studies [
27,
91], no significant difference between East- and West Germany in mental distress was found. Therefore, the period and cohort effect seems to play a role in the East-West differences. Interestingly, when combining gender and the German region, a significant interaction effect was present, revealing women in West Germany reported more mental distress compared to women in East Germany, whereas it was the other way around for men.
Strength and limitations
This study is the first in Germany to examine age, period, and cohort effects in mental distress for a time period of 15 years including gender and German region. By applying HAPC models, we identified and separated age, period, and cohort effects. Furthermore, it is the first study to combine HAPC models with the generally untested assumption of measurement invariance in age, period, and cohort studies.
Yet, several limitations need to be considered when interpreting these results. Although the HAPC model is currently often applied as an approach to examine age, period, and cohort effects simultaneously, the discussion regarding the appropriate way to analyze such effects remains vivid, as strong assumptions have to be made about the nature of the effects [
5]. Moreover, simulation studies revealed an underestimation of cohort effects when using the HAPC method [
92]. Robustness checks are one way of addressing these issues, but future studies should consider other APC techniques to validate the results, e.g., the newly developed age-period-cohort interaction model [
93]. Furthermore, the total time range covered in this study is still on the low end. Also, the number of years between the individual time points differ. In contrary to most APC studies, we used repeated cross-sectional sample survey data, which only simulates actual longitudinal data. Therefore, our findings do not provide insights into the possible causal effects for observed time trends. With regard to measurements, the classification of birth cohorts is partly theoretically and partly methodologically based, since a substantial amount of respondents is required in each cohort. We performed robustness tests using other cohort groups, which did not change our results. Lastly, we measured self-reported mental distress using the PHQ-4 scale. The PHQ-4 scale includes measured for symptoms of depression and anxiety, but results could differ from studies using solely depression or anxiety symptoms as an outcome as well as from studies that are based on diagnosed psychological disorders.