Evidence suggests an increase of depression and anxiety symptoms during the Covid-19 pandemic but most studies relied on cross-sectional designs and/or small samples, and they often overlooked subgroup effects in the impact of the lockdown. We investigated the effect of the pandemic on depression and anxiety symptoms, and whether it differed by employment situation and alcohol consumption.
This longitudinal study used 23 waves of the Covid-Questionnaire (April 2020—July 2021), within the Lifelines cohort from the Netherlands (n = 76,254). Depression and anxiety symptoms were combined in a “mental health score”. Linear fixed-effects models were fitted to analyse trends in mental health throughout the observation period. The moderating role of pre-existing mental health, employment situation, and alcohol consumption was tested.
Depression and anxiety symptoms fluctuated considerably during the observation period, with clear peaks in winter 2021, during the strictest lockdown period. Moreover, temporal patterns differed by employment situation and alcohol consumption patterns, suggesting that various subgroups reacted to the pandemic and the lockdown in different ways.
Lockdowns increased depression and anxiety symptoms in the Netherlands. The effect was particularly strong for unemployed individuals, those with risky alcohol consumption patterns and those with pre-existing mental health disorders.
Two years of the Covid-19 pandemic are likely to have affected mental health. Lockdowns -i.e. preventive measures implemented to control the spread of the virus- led to reduced social interactions and higher levels of social isolation [1, 2]. Moreover, they meant spending long hours at home for many people, which may also have had mental health consequences. Systematic reviews have consistently shown that the prevalence of mental health problems increased during lockdown periods [3, 4].
However, not all results point in the same direction. Some studies showed no evidence of a negative effect of lockdowns on mental health: a longitudinal study covering the first six weeks of the lockdown in the United Kingdom showed that depression and anxiety decreased during that period , and another study showed that mental health in Germany improved during the first eight weeks of lockdown . However, these studies covered only a short period of time at the beginning of the lockdown, and cannot be generalized to the whole pandemic period. A British study explained this pattern by suggesting an initial “shock” caused by the first lockdown, which led to more prevalent mental health symptoms, followed by improvement in subsequent weeks .
Although mental health disorders are complex, multifactorial phenomena, there is agreement that situations, in which individuals do not perceive real control over their lives, can trigger symptoms of depression or anxiety . This is particularly true for the employment domain . Unemployment has been shown to lead to poor mental health , partly because it often entails substantial uncertainty [1, 10, 11]. Therefore, we expect that being laid off in the already uncertain context of a pandemic will lead to a stronger mental health deterioration, as the impact of both unemployment and the pandemic are likely to interact and “accumulate” over time .
The pandemic also led to more frequent working from home. There is mixed evidence on the mental-health consequences of doing so. While some studies reported stronger negative effects, particularly for those living alone , allegedly due to stronger feelings of loneliness and isolation , others reported an overall positive effect of working from home, due to greater perceived control over work and better work-life balance . Although the consequences of working from home are likely to depend on numerous factors -e.g. type of job or household composition , such a sudden adjustment is likely to have a negative effect on mental health. Hence, we expect that those who worked full-time from home during the pandemic will report a stronger deterioration of their mental health.
Another group that has been proven vulnerable to the negative impact of lockdowns are people with pre-existing chronic conditions . However, evidence regarding the consequences for their mental health is less clear: although those with pre-existing mental disorders reported a higher prevalence of depression and anxiety symptoms , whether the effect of lockdowns was stronger for them remains unclear. In fact, one study showed that those with previous mental disorders may have adjusted slightly better to the lockdown, allegedly because their habits may have become “more in sync with a quarantined society” . Yet, the conservation of resources theory suggests that resilience during a disruptive event, such as the Covid-19 pandemic, requires certain resources, among which a fair health status . Thus, we expect those with pre-existing mental health conditions, to suffer a greater deterioration of their mental health.
In addition, the role of lifestyle factors, such as alcohol consumption, in mental health changes during the lockdown remains unclear. While alcohol consumption is considered a coping mechanism for pre-existing depression and anxiety according to self-medication theory , it may also increase the risk of mental disorders [19,20,21], and mood instability, i.e. sudden and frequent fluctuations of mood. These fluctuations are associated with worse treatment outcomes and increased health service use . Since alcohol abuse may interact with the already negative effects of the lockdown, we expect a stronger deterioration of mental health for those reporting risky alcohol consumption during the strictest lockdown periods .
Thus far, most of the evidence about the impact of lockdowns on mental health is based on studies using cross-sectional designs [23, 24] or, at best, short observation periods [2, 5, 14, 25], which results in limited evidence of mental health changes over time . Some studies have analyzed fluctuations in mental health symptoms longitudinally [6, 26] but they have relied on rather small samples and limited observations. Meanwhile, studies with larger samples mainly focused on the prevalence of mental health among different subgroups [1, 27] rather than changes over time, or focused on describing temporal trends but did not delve into potential subgroup effects . This study contributes to the literature by analyzing differences across subgroups in the association between lockdowns and depression and anxiety symptoms. To our knowledge, our study is the first to test whether the employment situation individuals faced during the pandemic, as well as their health-related behaviours -i.e. drinking patterns- and outcomes -pre-existing mental health issues moderated the impact of the lockdowns.
This study uses a large sample and 23 waves of panel data gathered frequently during the Covid-19 pandemic, which allows us to examine the development of depression and anxiety symptoms over a 15-month observation period. We aim to advance the existing longitudinal evidence by testing: (1) whether the lockdown affected depression and anxiety symptoms; (2) whether this effect differed by employment situation, and (3) whether this effect differed by alcohol consumption patterns. Furthermore, this study analyses the main determinants of greater fluctuations in depression and anxiety symptoms during the pandemic.
The Lifelines Covid-19 Questionnaire was launched within the Lifelines Cohort Study, a large prospective population-based cohort study and biobank in the three northern provinces of the Netherlands, examining in a three-generation design the biomedical, socio-demographic, behavioural, physical and psychological factors contributing to the health and disease of 167,729 individuals living in the north of the Netherlands -composition and characteristics of the sample have been discussed elsewhere .
To assess the effects of the pandemic, attitudes towards Covid-19 regulations, and the health behaviours of the study population, the Lifelines Covid-19 cohort was developed. Participants over 18 years with known email addresses (n = 140,145) were asked to complete detailed web-based questionnaires about their physical and mental health, living situation and health behaviours. A total of 76,795 individuals completed at least one wave of the questionnaire (54.8% response rate), with an average of 17 observations per person, from 24 waves in total. The characteristics and representativity of the sample are presented in Mc Intyre et al. . For our study, 23 waves of the Covid-19 questionnaire were used -wave 12 had no information on mental health-, covering the period between April 2020 and July 2021. Our final sample consists of 76,254 individuals (61.3% female; mean age 54.4 years; average follow-up of 146.54 days) and 876,874 observations. Data from the same participants in waves 4 and 5 of the Lifelines Cohort Study were used for comparison.
The Lifelines Covid-19 questionnaire is based on DSM-5 criteria and covers most depression and anxiety symptoms using the Mini International Neuropsychiatric Interview (MINI) . However, due to the frequency of data collection within our longitudinal design, the symptom duration does not follow the gold-standard definition for “major depressive disorder” (“for more than two weeks”) and “generalized anxiety disorder” (“at least 6 months”) . Instead, the questionnaire refers to symptoms since the last observation (“in the last 7 days” or “in the last 14 days”, depending on the observation). Therefore, we label these as “symptom scores” rather than disorders.
Depression symptom score
This variable ranges from 0 to 8 symptoms: (1) Depressed mood; (2) diminished interest or pleasure in most or all activities; (3) significant weight or appetite changes; (4) insomnia or hypersomnia; (5) psychomotor retardation; (6) feelings of worthlessness or excessive or inappropriate guilt; (7) diminished ability to think or concentrate; (8) recurrent thoughts of death or suicidal ideation.
Anxiety symptom score
This variable ranges from 0 to 6 symptoms: (1) Worrying excessively about daily problems; (2) worries present almost every day; (3) worries are hard to set aside and/or prevent concentration; (4) restlessness or feeling “keyed up or on edge”; (5) muscle tension (6) increased irritability.
In addition, both scores were summed to generate a combined mental health score, ranging from 0 to 14.
Main independent variable
Lockdown period or “days since first lockdown (15th of March, 2020)” was converted into a categorical variable with categories coinciding with every month of the observation period (category 1 = April 2021; category 24 = July 2021) to capture non-linear changes throughout the Covid-19 pandemic. Due to the timing of Lifelines assessments, a few months had very few (e.g. August 2020) or no observations (e.g. February 2021). These categories were merged with the previous month (e.g. “July/August 2020”) or dropped (February 2021).
Employment situation is a categorical variable comparing those “working as usual”, i.e. those whose employment situation did not change due to the pandemic (including full- and part-time work), with those “working full-time from home” (those combining working from home with working from their usual location were not included in this category, as we were interested in grasping the effects of fully adjusting to working from home), “retired”, “unemployed”, “occupationally disabled”, and “others” (including individuals on maternity leave, homemakers, students, etc.).
Alcohol consumption was assessed in 13 waves. The questions referred to the amount of alcohol consumed: “How many glasses of alcohol did you drink in the past 7 days?” or “in the past 14 days” from wave 7 onwards. In the Netherlands, standard glass is defined as containing roughly 10 g of alcohol . We assessed the following categories: “Abstinence” (no alcohol consumption since last observation); “moderate drinking” (< 1.5 drinks/day on average); “heavy drinking” (1.5–3 drinks/day) ; and “hardcore drinking” (> 3 drinks/day) .
Gender (male/female); Age at baseline, categorised in age groups (< 40, 41–50, 51–60, 61–70 and > 70); and educational level, categorised based on the Dutch educational system: “low” (up to general secondary education), “middle” (secondary vocational education, or higher general and pre-university education) or “high” (higher professional education or university education).
Pre-pandemic mental disorders
Based on data from wave 4 of the Lifelines Cohort Study (gathered between 2014 and 2017), gold-standard assessments of “major depressive disorder” and “generalized anxiety disorder” were combined in the variable pre-pandemic mental disorder.
Pre-pandemic alcohol consumption
Based on wave 5 of the Lifelines Cohort Study (gathered between 2016 and 2019), the following drinking patterns were assessed: “Abstinence”, “Moderate drinking” and “Heavy or Binge drinking” (including either “ > 1.5 glasses/day on average” or “4 or more drinks/occasion” in the last month ).
Our analytical strategy was based on the following steps: first, linear regression models (OLS) accounting for fixed-effects (FE) were fitted , with depression and anxiety symptoms as separate outcome variables, and the lockdown period as the main independent variable. Based on these FE models, the predicted symptoms at every time point were estimated.
Second, depression and anxiety symptoms were combined in an overall “mental health score” and the potential roles of the employment situation and alcohol consumption as moderators were tested by adding two-way interactions with the main independent variable. Again, the predicted symptoms at each time point for these subgroups of individuals were estimated.
Last, fluctuations in mental health  were studied as sensitivity analyses. To do so, the intra-individual standard deviation of mental health scores across waves was calculated. Next, this variable was regressed on a series of covariates that included mental health scores at baseline, as well as pre-pandemic mental disorders and drinking patterns. Missing values in pre-pandemic variables (25% and 40%, respectively) were imputed using multiple imputations by chained equations (MICE) with the rest of the covariates, including the outcome to improve the model’s predictive power. However, the imputed model was estimated using only observed values of the outcome , hence the somewhat lower sample size (n = 65,503). All analyses were performed with Stata 13.
The mean baseline depression and anxiety scores of the study population, broken down by the main variables of interest, are shown in Table 1. Overall, baseline levels are relatively low, since most individuals did not report any depression or anxiety symptoms at all (67 and 70%, respectively). Yet, the dispersion is pretty large, as shown by the standard deviations, because some groups -e.g. those with occupational disabilities or pre-existing mental health issues- did report high symptom scores. Thus, meaningful differences between groups are observed: unemployed individuals reported significantly higher levels than those working as usual (whereas retired individuals reported significantly lower symptoms), and younger individuals’ depression and anxiety levels doubled, and tripled (respectively) that of the oldest group. Differences by gender and living arrangement are also relatively large, with women and those living alone reporting significantly higher depression and anxiety symptoms than their counterparts, whereas differences by educational level or pre-pandemic drinking patterns are rather small.
Depression and anxiety symptoms during lockdown
An overview of the most relevant preventive measures implemented in the Netherlands during the observation period is shown in the appendix (Table A1). Table 2 shows the shifts during the lockdown period in depression symptoms (Model 1), anxiety symptoms (Model 2), and combined mental health symptoms (Model 3). Although the three outcome variables have different ranges, hence the different coefficient sizes, their temporal patterns are quite similar, with only slight differences between depression and anxiety in certain periods. For example, between September and November 2020 depression symptoms remained unchanged (β = 0.00; 95% CI 0.00; 0.01), whereas anxiety symptoms slightly decreased (β = − 0.03; 95% CI − 0.04; − 0.02).
As shown in Fig. 1, depression and particularly anxiety symptoms were relatively high at the beginning of the observation period and decreased steadily in the subsequent months, consistent with the negative coefficients shown in Table 2, with a dip in July/August 2020 (β = − 0.06; 95% CI − 0.06; − 0.05, and β = − 0.16; 95% CI − 0.17; − 0.15, respectively). However, both increased after the summer -anxiety more steeply- coinciding with a tightening of preventive measures and followed by a plateau during autumn. A third phase started in the winter of 2021, coinciding with the strictest period of the lockdown, during which a curfew was imposed and all non-essential businesses were closed: both anxiety and depression escalated and peaked by March 2021, when they surpassed the levels reported in baseline. Thus, depression symptoms increased by 30% and anxiety by 7% compared with baseline levels.
Finally, both symptom values sharply decreased afterwards, with lower anxiety and depression levels by April 2021 (anxiety even returning to baseline levels), and a subsequent steady decrease of symptoms -particularly anxiety- until summer. Notably, although the “reopening plan” or progressive relaxation of restrictions, including opening bars, allowing more guests at home, etc., did not start until the 28th of April 2021, schools had already reopened at the beginning of March. This, and the end of the winter season, coincided with a reduction in anxiety and depression levels. However, by the end of the observation period, the average depression levels remained 11% higher than at baseline (β = 0.05; 95% CI 0.04; 0.06 in June/July), whereas anxiety symptoms were slightly lower (β = − 0.02; 95% CI − 0.03; − 0.01).
The moderating role of the employment situation
Due to the similarity in the patterns of depression and anxiety symptoms, interaction terms were only added to Model 3 using a combined mental health symptom score as the outcome. As shown in Table A2 in the appendix, the large number of significant interaction coefficients show that employment situation moderated the association between the lockdown and mental health.
First, unemployed individuals reported a delayed improvement in their mental health after the first lockdown, in the spring of 2020: as shown in Fig. 2, their symptoms did not decrease until July/August 2020, when they abruptly did so, whereas their employed peers had started improving that May. Moreover, they showed a delayed improvement in their symptoms after the second lockdown: in May 2021, while most individuals reported decreased symptoms, this group experienced increased symptoms (β = 0.11; 95% CI 0.01; 0.21), and their symptoms did not decrease drastically until early summer 2021 (β = − 0.12; 95% CI − 0.22; − 0.01).
Second, retired individuals showed significant interaction coefficients in the summer of 2020 (July/August: β = 0.13; 95% CI 0.10; 0.16), when overall mental health improved substantially, suggesting that they experienced less improvement during the first period of relaxation of the measures. However, as Fig. 2 shows, after that period, the retired showed the most stable temporal pattern, with relatively low anxiety and depression symptoms, similar to those who were working as usual.
Third, for those working from home, mental health symptoms were particularly high in the autumn of 2020 and early spring of 2021. Moreover, an increasing gap with those working as usual was observed, combined with periods of convergence that coincide with the holiday periods (summer and Christmas). Thus, compared with the scores in May 2020 (β = 0.05; 95% CI 0.03:0.08), after the first lockdown, this group’s coefficients had increased by November 2020 (β = 0.10; 95% CI 0.06:0.14) and were even greater in April 2021 (β = 0.14; 95% CI 0.09:0.20), at the end of the strictest lockdown period, indicating an accumulation of mental health symptoms over time. Sensitivity analyses (not shown) showed that the association between lockdowns and mental health did not depend on whether these individuals lived with family or on their own.
Last, individuals with occupationally disability not only showed much higher levels of depression and anxiety symptoms than the others but also showed a significantly stronger deterioration of their mental health after the summer of 2020: in September/October 2020, coinciding with the tightening of preventive measures, their mental health symptoms increased by 11% compared to baseline (β = 0.15; 95% CI 0.07; 0.23) measures (as shown in Figure. A1). By the Spring of 2021, their depression and anxiety symptoms had increased by 28% (β = 0.35; 95% CI 0.25; 0.46), whereas other groups—e.g. those working as usual—showed a slight improvement (β = − 0.02; 95% CI − 0.05; 0.00). Interestingly, those with pre-existing mental disorders followed a similar pattern, although with somewhat higher symptom scores, suggesting that pre-existing mental disorders and occupational disability are strongly associated.
Overall, results suggest that unemployed individuals and those with occupational disabilities required some time after the relaxation of the measures in late spring 2020 to experience mental health improvements. Moreover, the latter were particularly vulnerable to the uncertainty that the pandemic sparked, judging by their strong fluctuations in mental health symptoms. Conversely, retired individuals seemed to overcome the initial shock quite fast and became less affected by the lockdown measures and the evolution of the pandemic over time. Finally, for those working from home mental health symptoms seemed to accumulate: while their mental health improved during holiday periods, their symptoms worsened coinciding with every new set of restrictions.
The moderating role of alcohol consumption
As shown in Table A3, while the pattern did not differ significantly between those reporting moderate and heavy drinking, those drinking more than 3 drinks/day reported significantly higher instability in their mental health throughout the observation period. Figure 3 shows that individuals reporting “hardcore drinking” moved from having average symptoms during periods with more relaxed measures, such as July/August 2020 and May 2021, to report the highest anxiety and depression levels during the strictest lockdown periods (by March 2021, mental health symptoms had increased by 44% compared to baseline levels). In some cases, these changes were very drastic, e.g. between July and September 2020, they increased steeply, but they decreased rapidly between March and May 2021. In contrast, abstainers reported relatively high levels of depression and anxiety but a more stable pattern overall, including during periods of relaxed measures (e.g. the summer of 2020 and spring of 2021).
Altogether, our moderation analyses show clear differences by employment situation and, to a lesser degree, by drinking pattern in the association between the lockdowns and mental health. Furthermore, symptom fluctuation throughout the observation period was relevant, as this indicates higher mental instability, and, therefore, higher vulnerability, for some groups. Hence, sensitivity analyses were conducted to identify the main determinants of mental health fluctuation during lockdowns.
Fluctuations in mental health during lockdown periods
As shown in Table A4 in the appendix, pre-existing mental disorders and mental health at baseline were the strongest determinant of mental health fluctuations, i.e. stronger fluctuations coincided with higher depression and anxiety symptoms before and at the beginning of the pandemic. Furthermore, fluctuations were particularly large among unemployed individuals and, to a lesser degree, those working from home, while retired individuals reported lower levels of fluctuation. Interestingly, the association between disability and mental health fluctuations weakened after pre-existing mental disorders were accounted for, suggesting that occupational disability was partly due to mental health conditions.
Meanwhile, alcohol consumption showed a gradient in the association with fluctuations in mental health: while abstainers did not differ substantially from those drinking moderately, the results show that those who reported heavy, and, particularly, hardcore drinking experienced significantly stronger mental health fluctuations, even after accounting for pre-pandemic drinking patterns and other covariates, suggesting that alcohol consumption independently contributed to mental health instability.
Our study, which relies on a very large sample and a robust longitudinal design provides detailed analyses of the trends in depression and anxiety symptoms that people experienced during lockdown periods. Moreover, by focusing on individuals’ working situations during lockdowns, as well as their mental health and drinking patterns before and during the pandemic, this study provides a thorough analysis of the main determinants of depression and anxiety in the context of the pandemic. Our results show that depression and anxiety symptoms fluctuated throughout lockdown periods, with relatively low values during summer periods, when restrictions eased, and a peak during the winter of 2021, coinciding with the strictest lockdown period in the Netherlands. Although these fluctuations were relatively small for the population as a whole, they particularly differed by employment situation and alcohol consumption, suggesting that various subgroups reacted to the pandemic and lockdowns differently.
First, our results showed increases in overall levels of depression and anxiety symptoms during the first lockdown in the Netherlands, and an overall improvement in mental health over the subsequent few months. This aligns with studies showing an initial shock, allegedly due to the fear and uncertainty that the outbreak of Covid-19 entailed [1, 2, 11], followed by an improvement in mental health [5, 6]. However, these early studies failed to capture the subsequent rapid increase in symptoms after the first pandemic summer, and the peak in depression and anxiety symptoms between January and March 2021, during the strictest lockdown period, as more recent studies showed [28, 38, 39]. Our findings align with the latter in suggesting that the measures implemented during that period -a curfew from 9 PM to 5 AM, only one visitor allowed at home, and the closing of all public buildings and non-essential shops- were strongly associated with the deterioration of mental health of the observed population . In addition, we observed an overall rapid improvement during the spring of 2021. Although restrictions were not entirely lifted until the end of May, a “cautious easing” of the lockdown started at the beginning of March, with schools reopening and shops and contact-based professionals permitted to receive customers by appointment. This, alongside the first vaccinations, may have sparked hope for a return to normal life that permitted an improvement of mental health symptoms. Alternatively, it could be argued that mental health symptoms tend to decrease after every winter [40, 41]. However, our results show a clear deterioration of mental health, particularly regarding the prevalence of depression symptoms, in the spring of 2021 compared with that of 2020, suggesting that the length of the lockdown was relevant .
Second, how strongly mental health fluctuated during the pandemic varied by individuals’ employment situation. In line with theories on perceived control , the fluctuations were particularly strong among unemployed individuals, who experienced a stronger deterioration in mental health during lockdowns as well as a slower improvement during periods of relaxed restrictions. Similarly, individuals who were working from home not only experienced a stronger deterioration of their mental health, but their mental health symptoms accumulated with every new set of restrictions. In turn, those with occupational disabilities had high baseline levels of depression and anxiety that worsened during lockdowns. Interestingly, the differences in reaction to lockdowns between unemployed and retired individuals—who, after an initial shock, showed a more resilient pattern during the winter of 2021—support the hypothesis that the deterioration of mental health during the pandemic was related to uncertainty and lack of control [7, 8], which often characterizes unemployment . In contrast, the relatively generous retirement pension system in the Netherlands  may have buffered the impact for retired individuals.
Third, results show that those reporting “hardcore drinking” suffered higher mental health instability, and that it was strongly related to the lockdowns, supporting the hypothesis that alcohol abuse increases the risk for depression [19, 21], and mood instability . Abstainers were consistently characterized by “relatively stable distress”, i.e. more symptoms but smaller fluctuations than those drinking moderately. We did not observe different patterns for those with moderate and heavy drinking patterns, suggesting that the threshold of 1.5 drinks per day was insufficient to identify a hazardous drinking pattern, at least in terms of mental health in our study population. However, “hardcore drinking” offered a better indicator, probably because it is a proxy for alcohol intoxication, which has been shown to entail higher risks (and present a steeper socioeconomic gradient) than mild, frequent abuse [44, 45].
Finally, sensitivity analyses showed that individuals with pre-existing mental disorders experienced more fluctuations in their mental health. Although fluctuation could also indicate recovery, these individuals reported more mental health symptoms during the observation period, which seems to contradict the idea of better adjustment among those with pre-existing mental disorders . Instead, our findings suggest that mental health symptoms tended to accumulate with every new lockdown, as previous studies reported , and in line with the concept of “accumulation of risks” .
This study has several limitations. First, although the depression and anxiety symptoms were based on DSM-V criteria, the duration of the symptoms assessed in the Lifelines Covid-19 Questionnaire did not match the standard definition of major depressive or generalised anxiety disorders, as we were interested in short-term responses to the pandemic. This disallowed straight comparisons of pre- and post-pandemic levels and therefore hampered the interpretability of mental health during the pandemic. Second, our alcohol assessment relied on a self-reported average of glasses consumed since the last observation but did not assess the intensity of drinking, i.e. acute intoxication. However, by establishing different cut-off points, we were able to identify a subgroup with a particularly hazardous drinking pattern. Third, despite our longitudinal design, a direct causal effect of the lockdowns on mental health cannot be inferred, as the associations could be confounded by unobserved variables, such as personality traits or genetic predispositions (e.g. individuals with different chronotypes could adjust differently to work-related transformations, such as working from home ). Last, due to data availability, we could not assess some potentially relevant variables, such as financial stress or decreased social life, which may have been strongly associated with mental health during the pandemic. Further research should focus on longitudinal designs to disentangle the numerous factors associated with mental health in the context of the Covid-19 pandemic.
This longitudinal study shows that depression and anxiety symptoms strongly increased in the Netherlands during the COVID-19 pandemic, particularly during the strictest lockdown period, in the winter of 2021. However, the deterioration of mental health during the pandemic was not homogeneous across the population, but particularly strong among unemployed individuals, as well as among those with risky alcohol consumption patterns and pre-existing mental disorders. An accurate profiling of subgroups at higher risk for higher depression and anxiety, as well as stronger mood fluctuations, is key to designing tailored preventive policies in exceptional contexts like the Covid-19 lockdown to prevent further deterioration of mental health in similar scenarios in the future.
The datasets generated and analysed during the current study were provided by Lifelines under licence. Access to the data can be granted under licence by Lifelines and the authors will share the codes used to produce the results presented in this paper upon request.
Chandola T, Kumari M, Booker CL, Benzeval M (2020) The mental health impact of COVID-19 and lockdown-related stressors among adults in the UK. Psychol Med. https://doi.org/10.1017/S0033291720005048
Fancourt D, Steptoe A, Bu F (2021) Trajectories of anxiety and depressive symptoms during enforced isolation due to COVID-19 in England: a longitudinal observational study. The Lancet Psychiatry 8(2):141–149. https://doi.org/10.1016/s2215-0366(20)30482-x
Wu T, Jia X, Shi H, Niu J, Yin X, Xie J, Wang X (2021) Prevalence of mental health problems during the COVID-19 pandemic: a systematic review and meta-analysis. J Affect Disord 281:91–98. https://doi.org/10.1016/j.jad.2020.11.117
Prati G, Mancini AD (2021) The psychological impact of COVID-19 pandemic lockdowns: a review and meta-analysis of longitudinal studies and natural experiments. Psychol Med 51(2):201–211. https://doi.org/10.1017/S0033291721000015
O’Connor RC, Wetherall K, Cleare S, McClelland H, Melson AJ, Niedzwiedz CL, O’Carroll RE, O’Connor DB, Platt S, Scowcroft E, Watson B, Zortea T, Ferguson E, Robb KA (2020) Mental health and well-being during the COVID-19 pandemic: longitudinal analyses of adults in the UK COVID-19 Mental Health & Wellbeing study. Br J Psychiatry. https://doi.org/10.1192/bjp.2020.212
Ahrens KF, Neumann RJ, Kollmann B, Brokelmann J, von Werthern NM, Malyshau A, Weichert D, Lutz B, Fiebach CJ, Wessa M, Kalisch R, Plichta MM, Lieb K, Tuscher O, Reif A (2021) Impact of COVID-19 lockdown on mental health in Germany: longitudinal observation of different mental health trajectories and protective factors. Transl Psychiatry 11(1):392. https://doi.org/10.1038/s41398-021-01508-2
Ross CE, Mirowsky J (2013) The Sense of Personal Control: Social Structural Causes and Emotional Consequences. Handbook of the Sociology of Mental Health. Handbooks of Sociology and Social Research, Springer, Netherlands, Dordrecht, pp 379–402
Marmot M (2004) The Status Syndrome. How social standing affects our health and longevity, Bloomsbury, London
Paul KI, Moser K (2009) Unemployment impairs mental health: meta-analyses. J Vocat Behav 74(3):264–282. https://doi.org/10.1016/j.jvb.2009.01.001
Miettinen A, Jalovaara M (2020) Unemployment delays first birth but not for all. Life stage and educational differences in the effects of employment uncertainty on first births. Adv Life Course Res. https://doi.org/10.1016/j.alcr.2019.100320
Curran E, Rosato M, Ferry F, Leavey G (2022) Prevalence and risk factors of psychiatric symptoms among older people in England during the COVID-19 pandemic: a latent class analysis. Int J Ment Health Addict. https://doi.org/10.1007/s11469-022-00820-2
Mandemakers J, Kalmijn M (2018) From bad to worse? Effects of multiple adverse life course transitions on mental health. Longitud Life Course Stud 9(3):299–311. https://doi.org/10.14301/llcs.v9i3.484
Niedzwiedz CL, Green MJ, Benzeval M, Campbell D, Craig P, Demou E, Leyland A, Pearce A, Thomson R, Whitley E, Katikireddi SV (2021) Mental health and health behaviours before and during the initial phase of the COVID-19 lockdown: longitudinal analyses of the UK Household Longitudinal Study. J Epidemiol Community Health 75(3):224–231. https://doi.org/10.1136/jech-2020-215060
Ipsen C, van Veldhoven M, Kirchner K, Hansen JP (2021) Six key advantages and disadvantages of working from home in europe during COVID-19. Int J Environ Res Public Health. https://doi.org/10.3390/ijerph18041826
Amerio A, Brambilla A, Morganti A, Aguglia A, Bianchi D, Santi F, Costantini L, Odone A, Costanza A, Signorelli C, Serafini G, Amore M, Capolongo S (2020) COVID-19 Lockdown: housing built environment’s effects on mental health. Int J Environ Res Public Health. https://doi.org/10.3390/ijerph17165973
Pan K-Y, Kok AAL, Eikelenboom M, Horsfall M, Jörg F, Luteijn RA, Rhebergen D, Pv O, Giltay EJ, Penninx BWJH (2021) The mental health impact of the COVID-19 pandemic on people with and without depressive, anxiety, or obsessive-compulsive disorders: a longitudinal study of three Dutch case-control cohorts. The Lancet Psychiatry 8(2):121–129. https://doi.org/10.1016/s2215-0366(20)30491-0
Hobfoll SE (2012) Conservation of Resources Theory: Its Implication for Stress, Health, and Resilience. The Oxford Handbook of Stress, Health, and Coping. Oxford University Press.
Rehm J, Kilian C, Ferreira-Borges C, Jernigan D, Monteiro M, Parry CDH, Sanchez ZM, Manthey J (2020) Alcohol use in times of the COVID 19: Implications for monitoring and policy. Drug Alcohol Rev 39(4):301–304. https://doi.org/10.1111/dar.13074
Boden JM, Fergusson DM (2011) Alcohol and depression. Addiction 106(5):906–914. https://doi.org/10.1111/j.1360-0443.2010.03351.x
Hillemacher T, Bachmann O, Kahl KG, Frieling H (2018) Alcohol, microbiome, and their effect on psychiatric disorders. Prog Neuropsychopharmacol Biol Psychiatry 85:105–115. https://doi.org/10.1016/j.pnpbp.2018.04.015
McHugh R, Weiss R (2019) Alcohol use disorder and depressive disorders. Alcohol Res. https://doi.org/10.35946/arcr.v40.1.01
Broome MR, Saunders KE, Harrison PJ, Marwaha S (2015) Mood instability: significance, definition and measurement. Br J Psychiatry 207(4):283–285. https://doi.org/10.1192/bjp.bp.114.158543
Alonso J, Vilagut G, Mortier P, Ferrer M, Alayo I, Aragon-Pena A, Aragones E, Campos M, Cura-Gonzalez ID, Emparanza JI, Espuga M, Forjaz MJ, Gonzalez-Pinto A, Haro JM, Lopez-Fresnena N, Salazar ADM, Molina JD, Orti-Lucas RM, Parellada M, Pelayo-Teran JM, Perez-Zapata A, Pijoan JI, Plana N, Puig MT, Rius C, Rodriguez-Blazquez C, Sanz F, Serra C, Kessler RC, Bruffaerts R, Vieta E, Perez-Sola V, group MW, (2021) Mental health impact of the first wave of COVID-19 pandemic on Spanish healthcare workers: a large cross-sectional survey. Rev Psiquiatr Salud Ment (Engl Ed) 14(2):90–105. https://doi.org/10.1016/j.rpsm.2020.12.001
de Vargas D, Ramirez EGL, Pereira CF, Jacinto Volpato R, de Oliveira SR (2022) Alcohol use and anxiety in primary health care patients during the Covid-19 pandemic: a phone-based study. Int J Ment Health Addict. https://doi.org/10.1007/s11469-022-00785-2
Vloo A, Alessie RJM, Mierau JO (2021) Gender differences in the mental health impact of the COVID-19 lockdown: Longitudinal evidence from the Netherlands. SSM Popul Health. 15:100878. https://doi.org/10.1016/j.ssmph.2021.100878
Stolz E, Mayerl H, Freidl W (2021) The impact of COVID-19 restriction measures on loneliness among older adults in Austria. Europ J Public Health. 31(1):44–49
Varga TV, Bu F, Dissing AS, Elsenburg LK, Bustamante JJH, Matta J, van Zon SKR, Brouwer S, Bultmann U, Fancourt D, Hoeyer K, Goldberg M, Melchior M, Strandberg-Larsen K, Zins M, Clotworthy A, Rod NH (2021) Loneliness, worries, anxiety, and precautionary behaviours in response to the COVID-19 pandemic: A longitudinal analysis of 200,000 Western and Northern Europeans. Lancet Reg Health Eur. 2:100020. https://doi.org/10.1016/j.lanepe.2020.100020
Ori APS, Wieling M, Lifelines Corona Research I, van Loo HM (2023) Longitudinal analyses of depression, anxiety, and suicidal ideation highlight greater prevalence in the northern Dutch population during the COVID-19 lockdowns. J Affect Disord 323:62–70. https://doi.org/10.1016/j.jad.2022.11.040
Scholtens S, Smidt N, Swertz MA, Bakker SJ, Dotinga A, Vonk JM, van Dijk F, van Zon SK, Wijmenga C, Wolffenbuttel BH, Stolk RP (2015) Cohort profile: lifelines, a three-generation cohort study and biobank. Int J Epidemiol 44(4):1172–1180. https://doi.org/10.1093/ije/dyu229
Mc Intyre K, Lanting P, Deelen P, Wiersma HH, Vonk JM, Ori APS, Jankipersadsing SA, Warmerdam R, van Blokland I, Boulogne F, Dijkema MXL, Herkert JC, Claringbould A, Bakker O, Lopera Maya EA, Bultmann U, Zhernakova A, Reijneveld SA, Zijlstra E, Swertz MA, Brouwer S, van Ooijen R, Angelini V, Dekker LH, Sijtsma A, Scherjon SA, Wijmenga C, Dekens JAM, Mierau J, Boezen HM, Franke L (2021) Lifelines COVID-19 cohort: investigating COVID-19 infection and its health and societal impacts in a Dutch population-based cohort. BMJ Open 11(3):e044474. https://doi.org/10.1136/bmjopen-2020-044474
Substance Abuse and Mental Health Services Administration - CBHSQ (2016) Impact of the DSM-IV to DSM-5 Changes on the National Survey on Drug Use and Health. Rockville, MD. Available from: https://www.ncbi.nlm.nih.gov/books/NBK519697
Gezondheidsraad, (2015) Dutch dietary guidelines 2015 (Richtlijnen goede voeding 2015). Health Council of The Netherlands, The Hague
Rosenberg G, Bauld L, Hooper L, Buykx P, Holmes J, Vohra J (2018) New national alcohol guidelines in the UK: public awareness, understanding and behavioural intentions. J Public Health (Oxf) 40(3):549–556. https://doi.org/10.1093/pubmed/fdx126
Berg N, Kiviruusu O, Huurre T, Lintonen T, Virtanen P, Hammarstrom A (2017) Associations between unemployment and heavy episodic drinking from adolescence to midlife in Sweden and Finland. Eur J Public Health 28(2):258–263. https://doi.org/10.1093/eurpub/ckx207
Brüderl J, Ludwig V (2015) Fixed-Effects Panel Regression. In: Wolf C (ed) The Sage handbook of regression analysis and causal inference. SAGE, Los Angeles, pp 327–357
Schoevers RA, van Borkulo CD, Lamers F, Servaas MN, Bastiaansen JA, Beekman ATF, van Hemert AM, Smit JH, Penninx B, Riese H (2021) Affect fluctuations examined with ecological momentary assessment in patients with current or remitted depression and anxiety disorders. Psychol Med 51(11):1906–1915. https://doi.org/10.1017/S0033291720000689
White IR, Royston P, Wood AM (2011) Multiple imputation using chained equations: Issues and guidance for practice. Stat Med 30(4):377–399. https://doi.org/10.1002/sim.4067
Wirkner J, Christiansen H, Knaevelsrud C, Lüken U, Wurm S, Schneider S, Brakemeier E-L (2021) Mental health in times of the COVID-19 Pandemic. Eur Psychol 26(4):310–322. https://doi.org/10.1027/1016-9040/a000465
Salfi F, Amicucci G, Corigliano D, Viselli L, D’Atri A, Tempesta D, Gorgoni M, Scarpelli S, Alfonsi V, Ferrara M (2022) Two years after lockdown: Longitudinal trajectories of sleep disturbances and mental health over the COVID-19 pandemic, and the effects of age, gender and chronotype. J Sleep Res. https://doi.org/10.1111/jsr.13767
Ayers JW, Althouse BM, Allem JP, Rosenquist JN, Ford DE (2013) Seasonality in seeking mental health information on Google. Am J Prev Med 44(5):520–525. https://doi.org/10.1016/j.amepre.2013.01.012
Fellinger M, Waldhor T, Serretti A, Hinterbuchinger B, Pruckner N, Konig D, Gmeiner A, Vyssoki S, Vyssoki B, Fugger G (2022) Seasonality in Major depressive disorder: effect of sex and age. J Affect Disord 296:111–116. https://doi.org/10.1016/j.jad.2021.09.051
Brydsten A, Hammarstrom A, San Sebastian M (2018) Health inequalities between employed and unemployed in northern Sweden: a decomposition analysis of social determinants for mental health. Int J Equity Health 17(1):59. https://doi.org/10.1186/s12939-018-0773-5
CFA Institute (2020). Mercer CFA Institute Global Pension Index 2020. D. Knox, (Ed.). Mercer, Melbourne.
Mackenbach JP, Kulhanova I, Bopp M, Borrell C, Deboosere P, Kovacs K, Looman CW, Leinsalu M, Makela P, Martikainen P, Menvielle G, Rodriguez-Sanz M, Rychtarikova J, de Gelder R (2015) Inequalities in alcohol-related mortality in 17 European countries: a retrospective analysis of mortality registers. PLoS Med 12(12):e1001909. https://doi.org/10.1371/journal.pmed.1001909
Mangot-Sala L, Smidt N, Liefbroer AC (2021) The association between unemployment trajectories and alcohol consumption patterns. evidence from a large prospective cohort in The Netherlands. Adv Life Course Res. https://doi.org/10.1016/j.alcr.2021.100434
Yan L, Gan Y, Ding X, Wu J, Duan H (2021) The relationship between perceived stress and emotional distress during the COVID-19 outbreak: Effects of boredom proneness and coping style. J Anxiety Disord 77:102328. https://doi.org/10.1016/j.janxdis.2020.102328
Kuh D, Ben-Schlomo Y, Lynch J, Hallqvist J, Power C (2003) Life course epidemiology. J Epidemiol Commun Health 57:778–783
This work was supported by the Research Fund of the Royal Netherlands Academy of Arts and Sciences (KNAW-Institutes).
Conflict of interest
The authors declare that they have no conflict of interest.
The LifeLines Cohort Study is conducted according to the principles of the Declaration of Helsinki and is approved by the medical ethical committee of the University Medical Center Groningen (UMCG), the Netherlands.
Consent to participate
Informed consent was obtained from all individual participants included in the study.
Below is the link to the electronic supplementary material.
About this article
Cite this article
Mangot-Sala, L., Smidt, N. & Liefbroer, A.C. Changes in anxiety and depression symptoms during the Covid-19 lockdown in the Netherlands. The moderating role of pre-existing mental health, employment situation and alcohol consumption. Soc Psychiatry Psychiatr Epidemiol 58, 1561–1571 (2023). https://doi.org/10.1007/s00127-023-02480-6
- Mental health
- Heavy drinking
- Alcohol abuse