Coronavirus disease 2019 (COVID-19) has given rise to an unprecedented pandemic, as no infectious disease had ever affected these many countries or this large a population. In Spain, as in many other countries, it has motivated the imposition of significant social isolation measures, such as lockdowns and school closures.1

In previous epidemics considered public health emergencies, evidence emerged of their deleterious effect on mental health.2–5 The scientific literature has described anxiety, depression, sleep disorders, post-traumatic stress disorder, addictive behaviours and eating disorders,6 as well as hallucinations and psychotic episodes.7

Several studies have confirmed the negative psychological impact of the COVID-19 pandemic, the lockdown and other measures used to contain it on the mental health of the population.7 The earliest evidence was published by Wang et al.,8 who reported a moderate to severe impact on mental health in 54% of individuals. In the paediatric population, early evidence from China described high levels of anxiety in children and adolescents of both sexes (which we will refer to from hereon as “children”) and depression symptoms in more than 22.9,10 In Germany, Ravens-Siebere et al11 found a higher incidence of mental health disorders in comparison with past records: 17.8% versus 9.9%, as well as higher levels of anxiety. The impact was hardest on children in low-income households, immigrants and individuals living in cramped spaces.

In Spain, several studies in adults reported a moderate to severe psychological impact on 24.7% of individuals, with deterioration of mental health in nearly 50%12,13 and a higher incidence of stress, anxiety and depression in the young.13 Lavigne-Cerván et al14 found a higher incidence of anxiety at ages 9–12 years and more sleep disturbances at ages 13–18 years. Orgiles et al15 described a substantial impact on the lives of Italian and Spanish youth, and many parents reported changes in mood and behaviour in their children. In Catalonia, the authors of the EmCoVID-19 found a significant increase in emotional symptoms and behavioural problems in children and adolescents.16 In vulnerable families, the incidence of mental health symptoms has been even higher; according to Save the Children,17 stress levels and household conflict had increased in 4 out of 10 disadvantaged households in association with cramped and poor living conditions.

Few studies have analysed the impact on mental health in children. The aim of our study was to assess the psychological impact of home confinement and school closures between March and September 2020 in the paediatric population of Catalonia.

We invited 529 children to participate, and the final analysis included 369 of them (Fig. 1).

Figure 1.

Flow chart of sample selection.

(0.12MB).

Results of the strengths and difficulties questionnaire

Fig. 2 presents the results of the SDQ in the first phase of an ongoing study, which corresponded to the beginning of the school year: the questionnaire found that mental health disorders were probable in 9.8% of participants (score ≥ 17), and possible in 11.1% (score, 14–16).

Figure 2.

Results of the Strengths and Difficulties Questionnaire (SDQ).

(0.1MB).

The proportion of participants with SDQ scores in the abnormal range was significantly higher in the group with a previous diagnosis of mental health or learning disorder (30.6%) compared to participants without this history (7.2%).

Fig. 3 illustrates that the scores in the conduct problems and hyperactivity symptoms subscales were higher in children aged 5–8 years.

Figure 3.

Results in the abnormal range of the SDQ by age group and subscale.

(0.24MB).

Compared to the rest of the sample, participants with SDQ scores in the abnormal range had significantly less room of their own to study, were more worried about their health (as were their parents) and perceived their home life more negatively, both during the lockdown and at the beginning of the school year. Also, in both periods, participants with abnormal scores exhibited significantly more emotional symptoms and conduct problems (irritability, sadness, nervousness and aggressiveness) and greater difficulty sleeping. During the lockdown, participants in this group were also more likely to use screens excessively.

Table 2 presents the multivariate analysis of patients with a SDQ score in the abnormal range (≥17 points) compared to the rest of the sample. We fit the model by adding only those variables that could have contributed to the impairment of mental health and with significant differences between 2 groups. The variables that remained significant in the multivariate analysis were current worry of the child regarding their own or their parents’ health, and a positive perception of the home life during the lockdown and at the beginning of the school year.

Table 2.

Odds ratio for patients with SDQ scores in the probable mental health problem range (≥ 17) compared to the rest of the sample.

Variables		Odds ratio (95% confidence interval)
Own space to study		0.677 (0.232−1.976)
Current worry of child regarding child’s own or family’s health	None/some	Ref
	A little	2.092 (0.760−5.759)
	A fair amount/a lot	4.916 (1.704−14.179)
Home life during lockdown	Very bad/bad	Ref
	Normal	0.279 (0.107−0.723)
	good/very bad	0.090 (0.025−0.320)
Current home life	Very bad/bad	Ref
	Normal	–
	good/very bad	0.571 (0.222−1.463)
Economic problems since pandemic started		1.294 (0.534−3.134)
Perceived falling behind in school work		1.694 (0.710−4.045)
Pre-existent mental health diagnosis		5.227 (1.826−14.968)

Odds ratio obtained by logistic regression analysis for each individual variable adjusted for all other variables featured in the table.

Statistically significant results are presented in boldface.

Symptoms during the lockdown and at the beginning of the school year

As can be seen in Fig. 4, the overall incidence of emotional symptoms and conduct problems in the sample was significantly higher during the lockdown than at the beginning of the school year, in particular when it came to sadness, nervousness and irritability. We found differences in sleep patterns and screen time, with 62% of participants reporting more than 6 h of screen time during the lockdown compared to 2% at the beginning of the school year. When we compared emotional symptoms and conduct problems by age group (Fig. 5), we found that children aged 5–8 years exhibited irritability, nervousness, euphoria and aggressiveness more frequently during the lockdown. Apathy, sleep disturbances and excessive screen use were more frequent in adolescents aged 12–14 years. During the lockdown, children in all age groups were less physically active and engaged in virtual conferences with family, and 17% gained weight. The difference in the manifestations observed during the lockdown compared to the beginning of the school year were more marked in the group aged 5–8 years (Figs. 5 and 6). The level of worry of children during the lockdown and the beginning of the school year ranged from low to moderate and remained stable, whereas parents were clearly more anxious during the lockdown. The perception of the home life remained stable, ranging from good to very good, in most cases.

Figure 4.

Comparison of emotional and behavioural symptoms in the lockdown period and at the beginning of the school year.

(0.22MB).

Figure 5.

Comparison of emotional and behavioural symptoms by age group.

(0.28MB).

Figure 6.

Comparison of emotional and behavioural symptoms at the beginning of the school year by age group.

(0.23MB).

The scores obtained at the beginning of the school year (following 6 months of school closure) revealed that mental health problems were probable in 9.8% of patients. Children aged 5–8 years seemed to be most affected. The reported level of worry of children regarding their own health and the health of their families was a risk factor for mental health problems, while a positive perception of the home life had a protective effect. During the lockdown, there was a higher frequency of emotional symptoms and behaviour problems (sadness, irritability and nervousness), sleep disturbances and excessive screen use compared to the beginning of the school year.

When it came to the patients in whom mental health problems were probable, we found a proportion of 9.8% at the start of the school year, which would correspond to an increase of 2 percent points compared to the 7.5% prevalence reported for the 2018–2019 period in Catalonia.18 If we were to add the percentage of patients in whom mental health problems were possible (11.1%) the estimated prevalence would increase to 20.9%, substantially greater compared to the 13.2% reported by the Ministry of Health, also based on SDQ scores, in 2017.25 A higher proportion of patients with a previous diagnosis of mental health or learning had SDQ scores in the abnormal range (30.6%) compared to children without a pre-existing diagnosis (7.2%) (P = 0.000). When we compared our findings to the previous literature, we found that studies that used the SDQ as an assessment tool found similar results in this population (Liu and Ezpeleta).26,27 On the other hand, the data published by Via et al16 show an increase ranging from 13% and 34.7% in the risk of mental health problems, chiefly symptoms of anxiety and depression. The higher frequency of mental health problems found by the latter authors could be due to the use of different instruments for detection and to motivational bias related to recruitment through social media, as opposed to random selection in our study.

Some authors have hypothesised that the magnitude of the impact of home confinement and the pandemic will depend, among other factors, on the age of the child.28 To date, the results in the PEDSAME cohort show a higher frequency of symptoms of irritability, sadness and nervousness during the lockdown in children aged 5–8 years, consistent with the findings of Via et al.16 Other studies, however, have found a higher impact on adolescents.29,30 The higher impact on younger children observed in our study could be explained at least in part by the difficulty at these ages to replace face-to-face interactions with peers with virtual relationships, as restrictions on outdoor activities affect this age group more strongly.

The general belief in the scientific community is that home confinement has a deleterious impact on mental health.31 There is evidence that confinement measures and changes in daily routines has a negative impact on psychological dimensions in the parents, exposing children to significant risks to their wellbeing.32 The results of the PEDSAME study evinced an association between the level of worry children experienced about their own and their families’ health and SDQ scores in the abnormal range. Similarly, Li Duan9 reported an association between psychological symptoms and the strategies used to cope with the situation.

When we analysed the symptoms associated with the lockdown, the PEDSAME study found an increase in sadness, irritability, nervousness, sleep disturbances and excessive screen time compared to the beginning of the school year. Jiao et al33 found similar emotional and behavioural symptoms, with nearly one third of children in their sample reporting irritability, nervousness and sleep disturbances. When it came to the use of electronic devices, the age group that made excessive use at all times (screen time > 6 h/day) was adolescents, and we also found excessive screen time in children with SDQ scores in the abnormal range. This trend had been described previously in the literature.9,34

As for physical activity, the percentage of children that did not engage in any form of physical activity during the lockdown was 9%. This contrasts with the data on vulnerable families published by Save the Children, with percentages above 80%.35 Other authors26 consider that physical activity could be an effective measure for alleviating behavioural problems in confined school-age children. A previous study found an association between engaging in physical activity for less than 1 h a day and mental health problems.34 Orgiles et al15 found that physical activity had a protective effect and that the wellbeing of parents was directly correlated to the wellbeing of children, a finding corroborated by the PEDSAME study, as a positive perception of home life was a protective factor for mental health disorders.

The retrospective collection of data for the lockdown period may be a limitation in our study, as it carries a risk of recall bias. But the time elapsed from the end of the lockdown to data collection was not excessively long and its influence was probably small. Another limitation is that there was little representation of the adolescent population, as paediatric services in the public health system only cover the population aged 14 years and younger and therefore our study did not include any adolescent past that age. In addition, the SDQ was completed by the main caregivers or teachers of participants and not by the children themselves (although participants could complete the questionnaire themselves from age 11–12 years).

The random selection of patients through a network of paediatricians representative of the territory of Catalonia is one of the strengths of the PEDSAME study, as most studies on this subject have used convenience or nonprobability sampling methods.

We must also underscore the usefulness of the SDQ for screening of mental health problems in children and adolescents. The translation to Spanish of this instrument has been validated and exhibited a high internal consistency. Recently, several studies have been published that used this same questionnaire to assess the impact of the COVID-19 pandemic in child and adolescent mental health, reporting outcomes consistent with those obtained in the PEDSAME study.11,26,27,32,36

The PEDSAME cohort study is still ongoing, and the mental health of participants will be assessed through the end of the 2020/2021 academic years at 3 additional time points: December 2020, March 2021 and June 2021. Its longitudinal and prospective design will allow detection of changes in mental health based on the evolution of the pandemic.

Home confinement and the closure of schools for 6 months have had a negative impact on the mental health of the child and adolescent population. The most frequent manifestations involved increases in emotional symptoms and conduct problems, sleep disturbances and unhealthy use of screens during the lockdown, and the incidence of psychiatric disorders increased compared to the pre-pandemic period. A positive perception of the home life had a protective effect on mental health, while excessive worry in children was a risk factor for mental health problems.

The authors have no conflicts of interest to declare.