Introduction

Social relationships are crucial for emotional development and adjustment in adolescence (Schwartz-Mette et al., 2020). Especially peers, but also parents have an influence on adolescent health behavior (Umberson et al., 2010; Anthony, 2014). Adolescents seek help with health concerns from family and friends (Booth et al., 2004), but regarding professional help, adolescents have concerns about confidentiality (Booth et al., 2004) and stigma (Sheppard et al., 2018).

Greater utilization of health care services by adolescents has been related to a higher parental educational level, parental social support (Giannakopoulos et al., 2010), being from a single parent family, and having greater social involvement (Vingilis et al., 2007). Conversely, lower health care service utilization by adolescents has been related to currently attending school or being employed (Vingilis et al., 2007). Furthermore, not living with two biological parents has been identified as a risk factor for mental health service use (Eijgermans et al., 2021).

Loneliness has been associated with greater self-reported health care service utilization in adolescence (Matthews et al., 2023). Adolescent loneliness is also related to symptoms of depression (Schwartz-Mette et al., 2020; Hards et al., 2022) and social isolation (Almeida et al., 2021). Both depression and loneliness have been considered as indicators of emotional adjustment during adolescence (Schwartz-Mette et al., 2020). The association between peer relationships, loneliness, and depression is strongest in younger adolescents (Schwartz-Mette et al., 2020).

Health care services are more commonly used by adolescent females than males (Nordin et al., 2010). According to previous research, among adolescent lifestyle factors, alcohol consumption may increase health care services utilization, especially among females (Kekkonen et al. 2015a).

A key finding from previous research is that in adolescence, seeking help from health care services is challenging without social support (Umberson et al., 2010; Giannakopoulos et al., 2010; Vingilis et al., 2007; Eijgermans et al., 2021). Social isolation (Almeida et al., 2021) and loneliness (Mathhews et al., 2023) in adolescents are related to poor mental health outcomes such as depression. Both loneliness and symptoms of depression reflect emotional adjustment during adolescence (Schwartz-Mette et al., 2020). However, to our knowledge, there have been no previous longitudinal studies on the effects of social relationships and loneliness together with depression on health care service utilization from adolescence to young adulthood.

Our aims were to investigate whether (1) relationships with peers and parents, as well as loneliness, are associated with primary health care utilization in adolescence and (2) the possible association between primary health care utilization and loneliness is mediated by symptoms of depression.

Methods

Participants

The participants were from a follow-up study of cohorts of adolescents aged 13 to 18 years attending comprehensive, upper secondary, and vocational schools in Kuopio, which is a city in Eastern Finland with approximately 121 000 inhabitants. The baseline data were collected using structured self-rating questionnaires that the participants completed during class periods at school.

The original target population comprised 6421 adolescents aged from 11 to 21 years in 2004–2005. The response rate was 65.5%, leading to a sample of 4214 adolescents. Girls responded significantly more often compared to boys. Altogether, 43 participants were excluded due to an age of 12 or younger or 19 or older, leading to a final sample of 4171 adolescents. From this population, 1827 (43.8%) provided their consent to be contacted for a follow-up study. The follow-up data were collected five years later by mail. There were 1585 participants whose addresses were retrieved for recruitment (86.8% of those who consented). Finally, 797 (females 70.9%) participated in the follow-up (50.3% of those whose postal address could be retrieved), and 793 of the participants had eligible data for retrospective health care outpatient register research.

Primary health care outpatient register

Primary health care data were obtained retrospectively from the outpatient registers of the Kuopio public health care center between years 2005 to 2010. Altogether, 862 of follow-up study participants were registered in Kuopio and their health care registers could be retrieved. Health care data comprised the reports of general practitioners, nursing services, school and student health care units, and emergency room services (located at Kuopio University Hospital in collaboration with local public primary, secondary, and tertiary health care). For each visit to primary health care, the main reason for the visit was determined from the medical records and classified according to the International Classification of Primary Care (ICPC) (Lamberts & Wood, 1987). Psychiatric reasons for visits included all mental health-related visit reasons, such as anxiety, depression, and sleep disturbances.

Ethical considerations

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. The study design was approved by the Research Ethics Committee of Kuopio University Hospital (no. 77/2010) and the Finnish Institute for Health and Welfare (no. THL/1628/5.05.00/2010). Before the study, written informed consent was obtained from all individual participants included in the study and from the parents of those aged below 15 years. All participants had the possibility to withdraw from the study at any time without explanation.

Measurements

Data collected at baseline included age, gender, and psychosocial background characteristics. Self-rated questions were derived from a standardized self-rated questionnaire, the Youth Self-Report for ages 11–18 years (YSR) (Achenbach & Rescorla, 2001), and included questions on the quality of relationships of the participants with peers and parents and on loneliness.

Depressive symptoms were assessed with the 21-item Beck Depression Inventory (BDI), in which questions target cognition, behavior, emotions, and somatic complaints (Beck et al., 1961). The BDI-21 has been validated for adolescents (Stockings et al., 2015).

Alcohol consumption was assessed with the consumption questions of the Alcohol Use Disorders Identification Test (AUDIT) (Saunders et al., 1993), AUDIT-C (Bush et al., 1998), which has been shown to be an effective screening tool for identifying risky alcohol consumption in young adults (Kelly et al., 2009).

Statistics

First, the characteristics of the study sample were investigated. Age was normally distributed, while the baseline BDI and AUDIT-C scores and the numbers of health care service visits were non-normally distributed. The Student’s t-test was used in the comparisons of normally distributed variables (age) and the non-parametric Mann–Whitney U-test in the comparisons of continuous variables with a non-normal distribution. The chi-squared test was used to analyze the group differences in the categorical variables. The associations between continuous variables were measured with Spearman’s rho (ρ) values.

Next, the number of primary health care visits during the five-year follow-up and associated factors were investigated with generalized linear models with a negative binomial distribution and logarithm link function due to the high number of zero values in the data, indicating no visits, and the non-normal distribution of some independent values. The number of primary health care visits was used as a dependent variable and all other variables, such as age, gender, social relationships, loneliness, BDI score, and AUDIT-C score, were used as independent variables. Loneliness was recategorized as a binary variable (0 = not lonely, 1 = lonely or somewhat lonely). Analyses were performed with all variables entered in the same model. Generalized linear models were repeated with only non-psychiatric primary health care visits as the dependent variable to avoid possible bias from psychiatric visits, i.e., psychiatric symptoms and intoxication as reasons for visits were excluded from the models.

Finally, the connections between loneliness, symptoms of depression, and health service utilization were assessed using the PROCESS macro (v. 4.1) for SPSS (Hayes, 2018). We used a simple mediation model (model 4) to test whether the connection between loneliness and health service utilization was mediated by depressive symptoms. According to the Breusch-Pagan test, the homoscedasticity assumption was violated (chi-squared (df) = 17.463 (1), p < 0.001), and in mediation analyses we therefore used the heteroscedasticity consistent standard error and covariance matrix estimator (Davidson-McKinnon). For mediation analyses, 5000 percentile bootstrap samples were used. If the upper and lower bounds of the 95% confidence interval did not contain zero, the result was considered statistically significant.

P-values below 0.05 were considered to indicate statistical significance. All the models were tested for multicollinearity, and all variance inflation factors (VIF) were less than 5. All analyses were conducted with IBM SPSS (version 27) statistical software.

Results

Characteristics of the study sample

There were 793 participants, and for 441 (55.6%; 303 (68.7%) females) of these, data on visits to primary health care services could be retrieved during 2005–2010, with a total of 15 329 visits. The mean age of the participants was 15.4 years (standard deviation (SD) 1.5) at baseline. During the follow-up period, the mean number of all primary health care visits was 19.3 (SD 28.9) (12.3 (SD 16.2) for males and 22.3 (SD 32.3) for females), with a non-significant difference between the groups (p = 0.068). Altogether, there were 659 visits for psychiatric reasons (mean 0.8, SD 2.6) and 4 visits with intoxication as the reason (mean 0.0, SD 0.1) for 144 participants.

Loneliness and symptoms of depression were more prevalent among females compared to males. Characteristics of the participants are presented in Table 1.

Table 1 Characteristics of the study sample and baseline differences between genders
Full size table

The number of all primary health care service visits correlated with age (ρ = -0.27, p < 0.001) and BDI scores (ρ = 0.10, p = 0.007), but not with AUDIT-C scores (ρ = 0.06, p = 0.114).

Results from the general linear model analysis

Female gender, young age, and symptoms of depression associated with a higher number of primary health care visits in both genders. Being lonely in males and poor relationships with peers in females associated with a lower number of primary health care visits. In males, a higher AUDIT-C score associated with a higher number of primary health care visits (Table 2).

Table 2 Generalized linear model analysis of all visits to primary health care services and the quality of social relationships, BDI, AUDIT-C scores at baseline
Full size table

Results from the general linear model analysis remained essentially similar after excluding all the visits for psychiatric reasons from the analyses (Supplementary table).

Results from the mediation analysis

lts of a simple mediation model of the connection betweeSimple mediation analysis (Tables 3 and Fig. 1) revealed that higher levels of loneliness were associated with higher levels of depressive symptoms (a = 4.696, p < 0.001), which in turn associated with more health care visits (0.706, p = 0.001). The bootstrap confidence interval for the indirect effect (ab = 3.315) based on 5000 bootstrap samples was entirely above zero (1.528 to 5.497). The direct effect between loneliness and health care visits did not remain statistically significant, indicating that loneliness independent of depressive symptoms did not affect the number of health care visits (c’ = -0.015, p = 0.994).

Table 3 Model coefficients for the simple model. Depressive symptoms mediate the effect that loneliness has on health service utilization
Full size table
Fig. 1
figure 1

Results of a simple mediation model of the connection between loneliness and health care service use, mediated by symptoms of depression

Full size image

Discussion

Main results

Our aims were to investigate whether social relationships were associated with primary health care utilization in adolescence, and whether the possible association between primary health care utilization and loneliness was mediated by symptoms of depression. In generalized linear models, young age, female gender, and symptoms of depression were associated with a higher number of primary health care visits in both genders. Being lonely in males and poor relationships with peers in females associated with a lower number of primary health care visits. Furthermore, alcohol consumption in males associated with a higher primary health care visit number. All the results from the general linear model analysis remained similar after excluding all the visits for psychiatric reasons from the analysis. In simple mediation analysis, loneliness associated with a higher number of primary health care visits only via symptoms of depression.

Comparison with the previous literature

Young age and female gender associated with a higher number of primary health care visits. Younger participants of this study had a relatively longer time to access school and student health care, which might explain their higher utilization rate. Female overrepresentation compared to males was consistent with previous research on gender differences in health care use in adolescence (Nordin et al., 2010) and in young adulthood (Blankson & Roberts, 2014). Furthermore, higher vulnerability to psychological problems in females (Pfeifer & Allen, 2021) and reproductive health care visits might explain these gender differences in health care utilization.

In males, the experience of loneliness associated with a lower number of primary health care visits. Contrary to this, in a previous six-year follow-up study, a recurrently self-reported (at the ages of 12 and 18 years) experience of being lonely associated with greater health care service use by adolescents (Matthews et al., 2023), although comparisons between genders were not reported. Some possibly tangential similarity with our results was found among 14–16-year-old adolescents, whose need for mental health services was negatively associated with withdrawn problem behavior in boys (Kim et al., 2014).

In females, poor relationships with peers associated with a lower number of primary health care visits. To our best knowledge, there has been no previous research on adolescent peer relationships and health care use. However, it has been suggested that a lack of social (Egenhuis et al., 2021) and peer support (Booth et al., 2004) might affect the willingness of adolescents to seek professional help for their health concerns. In addition, a lack of social relationships is related to higher levels of distress (Umberson et al., 2010). There are also gender differences in peer relationship qualities, suggesting that positive support and intimacy appear to be more common among adolescent females than males (Schwartz-Mette et al., 2020).

Relationships with parents did not affect the number of primary health care visits. We are aware of no previous studies on health care use and relationships between parents and adolescents. However, it is known that adolescents’ relationships with their parents might promote health behavior in adolescence (Umberson et al., 2010). In several studies, an insecure parental attachment style has been linked with more frequent somatic complaints among adolescents (27). In addition, family disintegration has been associated with a greater use of mental health care services in adolescence (Vingilis et al., 2007; Eijgermans et al., 2021).

Symptoms of depression associated with a higher number of primary health care visits in both genders. Depression is one of the most common mental health problems in adolescence (Silva et al., 2020), and becoming depressed in adolescence has been associated with greater use of health care resources in adulthood (Ssegonja et al., 2019). According to a Finnish health survey of young adults, mental symptoms such as depression increased the use of health care services, and health care use was more prevalent among young women compared to men (Kestilä et al., 2007). Overall, depression in primary health care patients has been found to be related to more severe somatic illnesses and disability (Berardi et al., 2002), which might reflect in a higher number of health care visits.

A higher score in AUDIT-C associated with a higher number of primary health care visits in males. Alcohol consumption in adolescence is associated with risky health behavior, comorbid mental health problems, and increased self-harm (Skala & Walter, 2013). In adolescence, alcohol use norms are absorbed from peers and parents (Kantawong et al., 2022), binge drinking is common (Skala & Walter, 2013), and is more frequent among alcohol consuming peers (Skala & Walter, 2013). Therefore, it is possible that alcohol consumption in adolescence reflects various risky health behaviors and health care service use.

Symptoms of depression mediated the association between loneliness and a higher number of primary health care visits through symptoms of depression. The relationship between depression and loneliness is well known (Schwartz-Mette et al., 2020; Hards et al., 2022). Loneliness is also related to wide range of other mental health problems (Matthews et al., 2023) and indicates problems in emotional adjustment in youth (Schwartz-Mette et al., 2020).

Strengths and limitations

The main strength of this study was its longitudinal setting and the examination of a large variety of possible confounding variables, combined with the use of retrospective medical health care registers. The high study dropout rate and female over-representation can be considered as limitations. Depression is a transitory and fluctuating phenomenon, and its measurements in several time points would have added to the validity of our findings. Loneliness, however, is a relatively stable feature in most people (Mund et al., 2020), and it is not likely that its changes over five years of follow-up would have biased our findings essentially. Only differences in the utilization of primary health care services were examined in this study. Despite the modern medical record systems, documentation of the reasons for health care visits was incomplete. Possible differences between professionals in their way of reporting medical examinations in patient files might have caused some bias. Utilizing structured clinical interviews to obtain diagnoses would have been a more exact means to evaluate the mental health status than self-reported questionnaires. However, due to the large sample size and limited resources, we were unable to utilize these types of tools with this dataset. According to previous study on sample selection biases, study participation associated with female gender, a higher number of hobbies, and better school performance, as well as symptoms of depression and anxiety (Kekkonen et al., 2015b).

Summary and conclusions

We investigated how health care service utilization in adolescence associated with quality of social relationships, symptoms of depression, and alcohol consumption in a longitudinal setting. Young age, female gender, and symptoms of depression associated with a higher number of primary health care visits in general. Loneliness in males and poor relationships with peers in females both associated with a lower primary health care visit number. Alcohol consumption associated with more primary health care visits in males. The results remained unchanged even when visits for psychiatric reasons were excluded from the analysis. In addition, loneliness mediated the association between symptoms of depression and a higher number of primary health care visits.

The results suggest that the quality of social relationships and depression might independently affect health care service utilization in adolescence. As the effect of loneliness has been the opposite in previous literature, our results suggest possible discrepancies in health care utilization from adolescence to young adulthood related to the quality of social relationships, and further longitudinal research is therefore needed. Seeking health care services in adolescence might be the most challenging without sufficient social support. Early identification and intervention in adolescent social isolation should be emphasized, and symptoms of depression detected by health care providers.