Background: Loneliness is becoming progressively more frequent despite increasing communication facilities. The COVID lockdown and social interaction restrictions enhanced loneliness complaints in more vulnerable groups while increasing its global prevalence.

Objective: To evaluate the prevalence, characteristics, and predictors of loneliness complaints during COVID19

Methods: The sample includes 5230 participants, 67.7% female, mean age 48.6 years and SD 14.30. To assure complexity/ diversity, an extensive internet survey with 177 questions was applied during the first COVID-19 pandemic wave in Portugal, including data from the Continent and Islands (Madeira and Azores).

Results: The prevalence was higher in females, emerging adults, those living alone, living in a flat, and in a big city. The following variables were higher in LG (Loneliness Group): Stress, depression, anxiety, irritability, worries, Calamity Experience Check List (CECL), economic problems, Sleep latency and Awakenings, Screen time in TV, Mobile, Social networks, negative attitudes and negative behaviors, dependences from TV, Social networks and Games, morbidities, worsening of previous morbidities, and nightmares. The predictors were civil status, living alone, and having negative attitudes during the pandemic.

Conclusions: The study allows us to conclude that loneliness during the COVID-19 pandemic was associated with health, psychological, behavioral, lifestyle, and housing-related factors; it could be predicted by the Calamity Experience Check List (CECL); Frequency of sexual activity; Negative attitudes; Positive attitudes; Negative Behaviors; Civil status; Living alone; Sleep latency weekdays; Sleep latency weekends. There were gender similarities and differences in loneliness predictors.

Keywords: loneliness, covid-19, lockdown, health, risks, predictors

Humans need social connectedness, which, since human existence, assures health, progress, survival, reproduction and consequent genetic transmission; therefore loneliness has potential dramatic effects upon physical and mental human health.^1,2

Before the COVID pandemic social isolation and loneliness were more prevalent in chronic health conditions, mental health conditions, cardiovascular problems, autistic individuals³ and in working adults with disabilities;³ furthermore, they were associated with increased all-cause mortality, and worse cardiovascular and mental health outcomes.³

A German study conducted before the pandemic found that the prevalence of loneliness in adults was 10.5% (felt mild in 4.9%, moderate in 3.9% and severe in 1.7%); it was more frequent in women, in those who lived alone and had no children; it was associated with depression (Odds Ratio (OR) = 1.91), anxiety (OR = 1.21), suicidal ideation (OR = 1.35), higher tobacco consumption and more medical visits.⁴

Loneliness in frontline medical workers figures among the risks of suicidal ideation during COVID, together with low social support; high physical and mental exhaustion and poorer self-reported physical health; sleep disturbances; quarantine; exhaustion; and mental health difficulties.⁵

In a French study loneliness and worries were major contributing factors for mental health and behavioral concerns during the COVID-19 lockdown.⁶

US national surveys made before (2018) and during the pandemic (2020) showed that psychological disturbances increased from 3.9% in 2018 to 13.6%; furthermore, 13.8% (95% Confidence Interval (CI), 11.4%-16.6%) of the adults in 2020 often felt lonely.⁷

Social isolation and loneliness are in older adults a serious problem, often neglected,⁸ sincelow social participation (Relative Risk (RR): 1.41 (95% CI: 1.13-1.75)), less frequent social contact (RR: 1.57 (95% CI: 1.32-1.85)), and more loneliness (RR: 1.58 (95% CI: 1.19-2.09)) are significantly associated with incident dementia;⁹ furthermore, loneliness mediates the effect of social isolation upon subjective cognitive impairment.¹⁰

In emerging adults, from 18 to 25 years of age, higher loneliness was significantly associated with worse self-rated health, higher marijuana consequences, less weekday sleep, and greater odds of feeling bothered by disturbed sleep; the interactions with sexual activity frequency, race/ethnicity, and sexual/gender minority (SGM) status were not significant.¹¹

Australian authors stress the multifold COVID impact upon loneliness. These effects might vary in different groups, such as: singles, physically and mentally disabled, carers and in those with low social capital; 'pruning' of social networks with digital interaction unable to substitute the lost physical contact; younger people isolated due to COVID life disruptions (they are unable to travel, or to attend university and amusements);¹² persistent mourning of those who died during the pandemic; infection worries; persistent isolation of elders living in nursing homes with family forbidden visits, etc.

In a loneliness study and its relations with stringency measures conducted in 5 European countries it was found that age (younger groups), gender (female), education, and cohabitation status play important roles in loneliness severity; furthermore they observed changes in social interactions across the variations in lockdown policies; they also found a marked heterogeneity in loneliness across individuals, and stressed the need to take it into account in order to propose adequate public policies.¹³

The possible mechanisms underlying the health consequences of loneliness would result from the activation of the sympathetic and hypothalamic-pituitary-adrenocortical (HPA) nervous systems.¹⁴ However, a recent experimental study did not find increased cortisol levels associated with loneliness and subjective health in both genders of middle age and older individuals, therefore rendering questionable the HPA role in this relation.¹⁵

Portugal had some specificities concerning the COVID pandemic: together with Slovenia and Spain, it is among the three countries with highest variability in COVID incidence;¹⁶ the stringency measures were quite high, being top when compared with EU countries in several pandemic periods, and had, among EU members, the most significant drop in real GDP.¹⁷ Furthermore, Portugal ranks 4th in depression risk.¹⁸ Since the risk of depression in EU, besides increasing from 20 to 21, is higher in females, young adults, unemployed, and in those complaining of financial difficulties.¹⁹

Since these factors are similar to the loneliness risks, we hypothesized that a higher loneliness prevalence during COVID should be expected in our country So far, the loneliness risks currently considered are mainly related to physical and mental health. We hypothesized a broader level of risks, which, in an ecologic perspective, includes lifestyles, attitudes and housing.

Among young people, loneliness is assuming a growing relevance, namely with the increasing use of information and communication technologies (ICTs), which, while facilitating communication between distant or isolated people, have been shown to isolate each on their own screen, reducing face-to-face interpersonal contact and sometimes constituting a psychopathology of addiction without substance.²⁰ The relations between screen time and depression have been often evaluated and discussed; a significant number of papers demonstrates increased depression levels in high media users, but recently a birectional relation between depression and social media use has been proposed, verifying that higher depression is associated with increased use.²¹ COVID-19 pandemic has not only heightened this use and abuse of technologies for the purposes of socialization, as it came to bring a new way of interacting with others. Confinements for the purpose of controlling the epidemic and fear of the other as a contact agent increased loneliness dimension, in spite of connecting the individuals in the virtual space.^20,22

In Portuguese adolescents social distancing from peers has ambivalent effects: it reduces health risks such as alcohol, tobacco, drug use, violence involvement (fights, bullying victimization and injuries), while decreasing the perception of wellbeing and life satisfaction and increasing the psychopathological symptoms.²²

Considering these facts and their known impact upon mental health and specifically upon loneliness a study focused on the associated factors and risks is required to evaluate regional data and further possible sources of loneliness heterogeneity. The study's main objective is to understand and characterize, from an ecological perspective, the factors linked to risk and protection in relation to loneliness in a pandemic context.

Participants

The sample includes 5230 participants, 67.7% female, mean age 48.6 years and SD 14.30, range from 18 to 90 years. The exclusion criteria adopted included being under 18 years of age, reporting incomplete questionnaires, and providing erroneous information.

Instruments/measures

The survey had 177 questions: demographics, health status; confinement characteristics; mood, attitudes, behaviors; calamity scales; sleep; physical activity; multimedia use; nutrition; toxics and addictions.

Health status included yes/no questions to: being healthy or suffering from the following: insomnia, sleep apnea, depression, anxiety/panic, other psychiatric disorder, sleep-related movement disorders, narcolepsy or hypersomnia, delayed sleep-wake phase disorder, shift work disorder, burnout or stress, parasomnias, cognitive deterioration/dementia, epilepsy, parkinsonism, headache/migraine, fibromyalgia, diabetes, hypertension, cardiovascular, cerebrovascular, respiratory, allergies, gastrointestinal, rheumatologic, endocrinologic/metabolic, autoimmune, orthopedic, cancer, renal, dermatologic, neurologic, hematologic, gynecologic, urologic, ear-nose-throat (ENT), ophthalmologic, chronic pain, fatigue, dizziness.

Morbidities index (MI) equals the sum of all referred symptoms/morbidities at baseline, COVID-19 Worsening (Morbidities worsening index (MWI))

Confinement attitudes and behaviors were evaluated by yes/no answers. The average and number of both positive and negative attitudes and behaviors were computed per subject.

Stress at pre-pandemic work was measured using a 1 to 10 scale; idem for depression, anxiety, and irritability.

CECL used VAS (Visual Analog Scales) of 1 to 10 and was validated.²³

Moral or sexual harassment pre-COVID, loneliness, conflicts and traumas during COVID were evaluated by yes/no answers.

Sleep included data relative to weekdays and weekends: Subjective sleep duration (hours), latency (min), night awakenings (number) and sleep and awakening quality in a 1-10 scales.

The Frequency of Sexual Activity was evaluated in a VAS, ranging from 1 (absent) to 10 (very frequent).

Physical activity (PA): hours/week. TV, social networks, mobile phone, gaming were quantified in hours/day.

Nutrition included: meals/day. Calculation of recommended food considered their frequency and type according to Portuguese Health authorities’ recommendations.

Smoking: yes/no; cigarettes/day. Alcoholic intake: Yes/No; glasses/day of beer, wine, aperitive wine, brandy. Drugs: no; occasionally; sometimes; regularly.

Lockdown housing included yes/no questions concerning: Housing (same home, family home, country house); Type of house (flat, villa, condominium, farm house, free comments); House location (City, Small town, Village, Isolated zone, Degraded zone); Number of people living with (alone, +1,+2,+3,+4, >5). Free comments in all subitems and in description of those living together were possible.

Lockdown attitudes and circumstances were evaluated by questions of yes/no answer of several categories:

• Negative attitudes: fed up or tired, cannot stand it, loneliness, missing family or friends, felt in imprisonment /claustrophobia, had worries and fears, had unexpected conflicts, cannot stand companion, cannot stand children, cannot stand elderly, fed up with the children tele-school
• Trauma/Violence: had traumatic events, had domestic violence, there was violence against children
• Positive attitudes: felt OK, had less stress, made important discoveries. Free comments. The average and number of positive, negative and violent attitudes were computed for each subject.

Lockdown behaviors were evaluated by the following yes/no answers: 1) Positive behaviors: tidying up, new type of work, phone friends, decided life changes, write a book/ articles/ memories, learned new abilities, gardening/ agriculture, invented funny or spiritual things, worked, walking/ gym / sports, Reading/Music/Studying, domestic work 2) 1) Negative behaviors: developed new addictions, get bored, mourned all time, slept as much as possible. Free comments. The average and number of both positive and negative behaviors were computed for each subject.

Procedures

The overall project was approved by CENC´s Ethical Committee 1/2020.

Online surveys were disseminated on the iSleep website. The study includes data from all country: Continent and Islands (Madeira and Azores). Support and dissemination of several entities was an important help. Among them figure the Portuguese professional Associations of Medical Doctors, Nurses, Psychologists, Pharmacists; Scientific Societies of Neurology, Sleep and Sleep Dentistry; a Network of Portuguese Sleep Labs and CENC.

Survey Legend® platform was used. Surveys were anonymous, for adults (>18y) allowing data analysis and statistical use. The first page included: purpose, authors, Ethical reference, contact person and supporting entities, and a question of whether data analysis was permitted. It was online during the 1st COVID-19 wave, from April to August 2020.

Statistics

Loneliness was defined as the subjective feeling of being lonely during lockdown. Data was divided in 2 groups: Loneliness YES or NO. For each subgroups qualitative variables were described by absolute frequencies, while quantitative variables were calculated by the mean or median depending on data distribution. The effect of variables of interest was evaluated using Chi-square tests (qualitative answers/frequency tables) or ANOVA (unidirectional analysis of variance). For Some variables resulting from computation such as number of attitudes and Behaviors the Percentile 75 was computed, and variables transformed into categorical: higher versus lower than P75. Categorical variables with low prevalence (less than 50 cases) were not used.

Binary Logistic Regression was calculated using the variables significantly different in One-way- ANOVA as covariates (those highly correlated were excluded), and loneliness YES/NO as dependent variable; Omnibus Tests of Model Coefficients was used to check whether the new model is an improvement over the baseline model and goodness of fit was evaluated by 2log likelihood, R2 Cox and Snell, R2 Nagelkerke; the final classification was obtained comparing group membership of predicted and observed values. Regression was computed for the global population and for each gender, using the same covariates. Significance was set at 0.05; SPSS® v25 was used.

During the first COVID wave the prevalence of loneliness was 12,3%, ie, 647 subjects in a population of 5251 individuals. The demographic data are in Table 1. Average values of age and BMI were significantly lower in the loneliness group (LG); Loneliness was more frequent in females and in emerging adults in those without a couple (single, widows and divorced), in those with a master’s degree, and in those living alone, in a flat, in a big city.

Table 2 presents data concerning Stress and subjective evaluation of Mental health. Stress, depression, anxiety, irritability, worries, CECL, economic problems were significantly higher in LG, while the way people were living during the pandemic was worse and the frequency of sexual activity lower.

Table 3 includes Sleep data. Sleep duration in weekdays and weekends, Sleep quality and Sleep awakening quality were significantly lower in LG; while Sleep latency and Awakenings were increased both during weekdays and weekends together with nightmares.

Table 4 shows the conventional habits. There were no differences in nutrition habits (number of meals and type of food recommended yes or no) and hours of practicing physical activity; LG group drank lower amounts of alcohol. Screen time in TV, Mobile, Social networks was higher in LG group, no difference in games.

Table 5 presents attitudes and behavior during pandemic restriction; LG had more negative attitudes and negative behaviors and less positive attitudes, with no differences in the remaining.

Table 6 shows that LG had higher levels of dependences from TV, Social networks and Games, more morbidities, and more worsening of previous morbidities, but no differences in alcohol dependence.

Among the survey responders only 29.1% considered themselves healthy; the percentage of being subjectively healthy was lower in the LG. Loneliness group suffered more frequently of insomnia, narcolepsy/hypersomnia, shiftwork, depression, anxiety/panic, burnout, headaches, fatigue, respiratory diseases, allergies, endocrine, autoimmune, and dermatologic disorders, tinnitus and dizziness. There were no differences in Movement sleep disorders, Chronic Pain, Rheumatologic disorders, Diabetes, Hypertension, Heart diseases, Gastrointestinal, Rheumatologic, Orthopedic and Ophthalmologic Disorders, Cancer, ENT. Loneliness was less frequent in Sleep apnea patients. Table 7 presents data from health and diseases in the LG.

Data from Logistic Binary Regression for the all sample and for the genders male and female show that only a small set of variables predicts loneliness for all of them, namely Civil Status: OR 2.0 (95% CI 1.3 – 3.08), Living alone OR 0.17 (95% CI 0.10 – 0.26), CECL OR 1.23 (95% CI 1.1 – 1.37), Negative attitudes OR 5.5 (95% CI 3.58– 8.41) (Table 8).

Table 8 Loneliness - logistic binary regression

Males have a high probability of negative attitudes OR 13.9 (95% CI 5.1 – 38.1) while in females OR 4.6 (95%CI 2.8 –7,5); for positive attitudes the values are not significant for males but are significant for females with OR <1 (OR=0.16 (95% CI 0.04 – 0.7). Negative behaviors are significant in males OR=3.52 (95% CI 1.5 – 8.5) and not in females. The Civil status (living together legally or not) association with loneliness is significant in both genders but the OR are much higher in males. Living alone is significant with similar OR for loneliness in both genders; idem for CECL. Females have significant differences in sleep latency both during weekdays and weekends, but the ORs are close to 1.

The classification achieved for group membership of Loneliness YES /NO was 90.2% correct for the all group, 93.6% for males and 89.0 for females.

The loneliness prevalence data obtained in our study (12.6%) are in line with data of the pre COVID era in Germany – 10.5%⁴ and with those from USA during COVID 13.8%.⁷ Our first hypothesis that the prevalence would increase in our country due the high variability of COVID stringent measures was not verified.

Loneliness was more frequent in females and in emerging adults, in those without a couple (single, widows and divorced), in those with a master’s degree, and in those living alone, in a flat, in a big city.

In synthesis the following variables were higher in LG: Stress, depression, anxiety, irritability, worries, CECL, economic problems, Sleep latency and Awakenings, Screen time in TV, Mobile, Social networks, negative attitudes and negative behaviors, dependences from TV, Social networks and Games, morbidities, and worsening of previous morbidities, nightmares. Sleep duration in weekdays and weekends, Sleep quality and Sleep awakening quality, positive attitudes were significantly lower.

LG complain or suffer more frequently of insomnia, narcolepsy/hypersomnia, shiftwork, depression, anxiety/panic, burnout, headaches, fatigue, respiratory diseases, allergies, endocrine, autoimmune, and dermatologic disorders, tinnitus and dizziness and less frequently from Sleep apnea. No differences in Movement Sleep disorders, Chronic Pain, Rheumatologic disorders, Diabetes, Hypertension, Heart diseases, Gastrointestinal, Rheumatologic, Orthopedic and Ophthalmologic Disorders, Cancer, ENT.

The higher female prevalence in consensual in many studies. Females clearly have a different risk profile of loneliness when compared to males.

Some studies found, as observed in this study, a higher prevalence in emerging adults,^11–13 while other studies point to the elderly (being than a risk for cognitive impairment and dementia.¹⁰ The differences concerning the elder in our study are likely related to two factors: the elders answering the survey are dwelling elders and do not live in nursing homes; furthermore, Portugal is a southern country and therefore keeps still often the family home with 2 or 3 generations living either together or close by Caro JC.¹³

Concerning education only those with a master’s degree had a higher loneliness prevalence; the PhDs and those with lower education had similar prevalence values. We have no clearcut explanation for that; but a possible explanation is that their jobs predominate in Science, Education and Health, all three sectors with lot of job difficulties due either to hard work, job insecurity or both. Masters have in fact less economic problems when compared with the other education groups, except graduates and PhDs, but had the highest CECL value of all education groups, but with no significant differences, exception made for the PhD group (lowest CECL and different from the others).

Living alone is associated with increased loneliness frequency but is not a loneliness risk factor, since the OR is smaller than 1: OR=0.17 (95%CI 0.1 – 0.3). Living with more people, although preventing loneliness, increases, however, the contamination risk¹³ and the risk of conflicts. Living in a flat or in a big city, represents the modern paradigm of being alone surrounded by many people. Nowadays many modern buildings have no balconies, terraces or gardens. The COVID pandemic and the associated social isolation showed how important it is the open-air exposure. Those living in a villa or in a farm had a lower loneliness prevalence, among other factors, walking outside was easier for them.

Work stress before COVID, depression, anxiety, worries and CECL are all higher in LG, but only CECL is a significant predictor of loneliness: OR= 1.23 (95% CI 1.1 – 1.4).

Sleep disturbances (latency, awakenings, sleep and awakening quality, sleep<5h and nightmares) were more prevalent in LG but only sleep latency had predictive value with low OR.

Idem for the frequency of sexual activity, lower in LG, with small OR. The same interaction with sexual activity was described by others.¹¹

Loneliness was more prevalent in sleep disorders (insomnia, narcolepsy/hypersomnia, shiftwork). There is a consensus that Sleep was particularly affected during COVID.^240–26 Insomniacs worries, narcolepsy vulnerabilities and shift work disorders. In a period with marked and unexpected modifications in daily life and work habits had reasons for feeling loneliness more acutely.

Our data are in line with other publications concerning the higher loneliness prevalence in Mental disorders (depression, anxiety/panic, burnout).³ Confinement restrictions impacted negatively upon habits and addictions.²⁸ Altogether these facts increase loneliness, but, in our data they are not a risk for it, as others observed before COVID.⁴

Before COVID chronic disorders (respiratory diseases, allergies, autoimmune, endocrine, dermatologic headaches, fatigue, tinnitus and dizziness) –were significantly associated with Loneliness, with a special reference of cardiovascular disorders.³

In this study, during COVID, we detected this association with Loneliness in respiratory, allergic and autoimmune diseases, which altogether are in a higher of a viral infection. Headache, fatigue, tinnitus and dizziness and dermatologic disorders share the significant increase in worries versus uncertainty and have high CECL values.

The classification achieved by Binary Logistic Regression in group membership: Loneliness YES /NO was good 90.2% correct for the entire population, with gender differences (93.6% for males and 89.0% for females). For the Omnibus Tests of Model Coefficients, the results were statistically significant and the values of 2log likelihood relatively high. However, values of the markers of goodness of fit, R2 Cox and Snell, R2 Nagelkerke, were relatively low.

For the general population 9 variables were loneliness predictors: CECL; Frequency of sexual activity; Negative attitudes; Positive attitudes; Negative Behaviors; Civil status; Living alone; Sleep latency weekdays; Sleep latency weekends. For Males the loneliness predictors were: CECL, Negative attitudes; Negative behaviors, Civil status and living alone.

For Females the predictors were: CECL, Negative attitudes; Negative behaviors, Civil status, Living alone, Sleep latency weekdays; Sleep latency weekends.

These predictors are in accordance with our second hypothesis concerning broader ecologic risks of loneliness.

Study limitations concerning the type of sample, the limitations of an internet survey and the cross-sectional design, the must be taken into consideration together with the possibility of a reverse causation has it has been observed in the relations between depression and social media.²¹

The prevalence of Loneliness during COVID was 12.3%. The prevalence was higher in females, in emerging adults, in those living alone, in a flat, and in a big city. These data imply recommendations both for supporting the most fragile groups and for architectural planning in big cities.

Loneliness is associated with a dark constellation of symptoms and factors, namely: Stress, depression, anxiety, irritability, worries, CECL, economic problems, Sleep latency and Awakenings, Screen time in TV, Mobile, Social networks, negative attitudes and negative behaviors, dependences from TV, Social networks and Games, morbidities (insomnia, narcolepsy/hypersomnia, shiftwork, depression, anxiety/panic, burnout, headaches, fatigue, respiratory diseases, allergies, endocrine, autoimmune, and dermatologic disorders, tinnitus, dizziness, nightmares).

The study allows us to conclude that loneliness during the COVID-19 pandemic was associated with physical and mental health, psychological, behavioral and lifestyles, and housing factors. There are gender similarities and differences regarding the predictors of loneliness.

It can be concluded that loneliness is explained by the ability to adapt to the pandemic, particularly in relation to the uncertainty and associated constraints. Marital status and negative or positive attitudes towards lockdown also affect the perception of loneliness. In terms of lifestyles, sleeping habits and sexual activity were also important predictors. Sleep was particularly relevant as predictor in women.

In conclusion loneliness is a very important aspect in people's mental health, adaptation and lifestyle. Protective factors in relation to loneliness allow for better adaptation and health. It is important to evaluate the medium and long-term impacts of this perceived loneliness and other less healthy impacts of the pandemic. On the other hand, for future pandemic or calamity situations, promote intervention in the prevention of loneliness and promotion of healthy lifestyles as prevention of psychosocial health risks.

We acknowledge support in data collection to: Conceição Pereira, Alexandra Carreiro, Aurora Lino, Susana Moreira, Ana Bernarda, Susana Gaspar, Lucia Ramiro, Amelia Feliciano, Maria Augusta Machado, Júlio Fonseca, Gabriela Videira.

Author contributions

Conceptualization, Teresa Paiva, Tania Gaspar, Margarida Gaspar de Matos.; methodology, Teresa Paiva, Tania Gaspar, Margarida Gaspar de Matos.; software, Teresa Paiva; validation Gina Tomé.; data curation, all authors.; writing—original draft preparation, Teresa Paiva, Tania Gaspar, Margarida Gaspar de Matos.; writing—review and editing, all authors.

Institutional review board statement

The study was conducted in accordance with the Declaration of Helsinki, and approved by the Ethics Committee of CENC, Lda (Dra Teresa Paiva, Centro de Electroencefalopgrafia e Neurofisiologia Clínica, Lda; approval date 15thJune 2020 for studies involving humans.

Informed consent statement

Informed consent was obtained from all subjects involved in the study.

Data availability statement

Data will be available when research papers publication will end.

None.

The authors declare no conflict of interest.

1. Cacioppo JT, Cacioppo S. Loneliness in the modern age: an evolutionary theory of loneliness (ETL). Adv Exp Soc Psychol. 2018;58:127–197.
2. Holt-Lunstad J. Why Social relationships are important for physical health: a systems approach to understanding and modifying risk and protection. Annu Rev Psychol. 2018;69(1):437–458.
3. Quadt L, Esposito G, Critchley HD, et al. Brain-body interactions underlying the association of loneliness with mental and physical health. Neurosci Biobehav Rev. 2020;116:283–300.
4. Beutel ME, Klein EM, Brähler E, et al. Loneliness in the general population: prevalence, determinants and relations to mental health. BMC Psychiatry. 2017;17(1):97.
5. Farooq S, Tunmore J, Wajid Ali M, et al. Suicide, self-harm and suicidal ideation during Covid-19: A systematic review. Psychiatry Res. 2021;306:114228.
6. Megalakaki O, Kokou-Kpolou CK. Effects of biopsychosocial factors on the association between loneliness and mental health risks during the COVID-19 lockdown. Curr Psychol. 2022;41(11):8224–8235.
7. McGinty EE, Presskreischer R, Han H, et al. Psychological distress and loneliness reported by US adults in 2018 and April 2020. JAMA. 2020;324(1):93–94.
8. Freeman D, Sheaves B, Waite F, et al. Sleep disturbance and psychiatric disorders. Lancet Psychiatry. 2020;7(7):628–637.
9. Kuiper JS, Zuidersma M, Oude Voshaar RC, et al. Social relationships and risk of dementia: A systematic review and meta-analysis of longitudinal cohort studies. Ageing Res Rev. 2015;22:39–57.
10. Jang Y, Choi EY, Park NS, et al. Cognitive health risks posed by social isolation and loneliness in older Korean Americans. BMC Geriatr. 2021;21(1):123.
11. Perez LG, Siconolfi D, Troxel WM, et al. Loneliness and multiple health domains: associations among emerging adults. J Behav Med. 2022;45(2):260–271.
12. Patulny R, Bower M. Beware the “loneliness gap”? Examining emerging inequalities and long‐term risks of loneliness and isolation emerging from COVID‐19. Aust J Soc Issues. 2022;57(3):562–583.
13. Caro JC, Clark AE, D’Ambrosio C, et al. The impact of COVID-19 lockdown stringency on loneliness in five European countries. Soc Sci Med. 2022;314:115492.
14. Xia N, Li H. Loneliness, social isolation, and cardiovascular health. Antioxid Redox Signal. 2018;28(9):837–851.
15. Crespo-Sanmiguel I, Zapater-Fajarí M, Garrido-Chaves R, et al. Loneliness and health indicators in middle-aged and older females and males. Front Behav Neurosci. 2022;16:809733.
16. Ruano-Ravina A, López-Vizcaíno E, Candal-Pedreira C, et al. COVID-19 variability within european countries sourced from ecdc data. is variability explained by specific country policies? Front Public Heal. 2022;9:737133.
17. Muggenthaler P, Schroth J, Sun Y. ECB Economic Bulletin. Issue 5. 2021.
18. OECD/European Union. Share of adults in the EU at risk of depression among selected groups, 2020‑22. In: Health at a Glance: Europe 2022: State of Health in the EU Cycle. Paris: OECD Publishing; 2022.
19. OECD/European Union. Share of adults at risk of depression, 2020‑22. In: Health at a Glance: Europe 2022: State of Health in the EU Cycle. Paris: OECD Publishing; 2022.
20. Tomé G, Matos MG. Cognitive and behavioral therapy in solidarity. In: Neufeld CB, Falcone EMO, Rangé BP, editors. Procognitiva Programa de Atualização em Terapia Cognitivo-Comportamental: Ciclo 9. Porto Alegre: Artmed Panamericana; 2023. p. 956.
21. Hartanto A, Quek FYX, Tng GYQ, et al. Does social media use increase depressive symptoms? a reverse causation perspective. Front Psychiatry. 2021;12:641934.
22. Tomé G, Branquinho C, Cerqueira A, et al. COVID-19, social distance and adolescents’ risk behaviours, wellbeing and life satisfaction: a proxy study drawn from HBSC study. Análisis y Modif Conduct. 2021;47(176):3–17.
23. Tomé G, Paiva T, Ramiro L, et al. A new Calamity Experience Checklist (CECL) for pandemic individual and group impact evaluation. In: Milcu M, Stevens M, Dhal L, editors. The psychology of pandemics: new trends and prospects, XIII; 2022:186–193.
24. Paiva T, Feliciano A, Machado MA, et al. COVID-19, professionals, disorders, health related behaviors, recommendations. Eurosurveillance.
25. Jahrami HA, Alhaj OA, Humood AM, et al. Sleep disturbances during the COVID-19 pandemic: A systematic review, meta-analysis, and meta-regression. Sleep Med Rev. 2022;62:101591.
26. Almondes K, Araujo E, Paiva T. Sleep habits, quality of life and psychosocial aspects in the older age: before and during COVID-19. Front Neurosci. 2022;16:694894.
27. Canas-Simião H, Reis C, Carreiras D, et al. Health-related behaviors and perceived addictions. J Nerv Ment Dis. 2022;210(8):613–621.
28. Freedman A, Nicolle J. Social isolation and loneliness: the new geriatric giants: Approach for primary care. Can Fam Physician. 2020;66(3):176–182.