Autonomy and health-related quality of life in adolescents

Health and well-being are essential in all human life. Health-related quality of life (HRQoL) is a multi-dimensional measure that relates to individuals’ perception and subjective evaluation of their health and well-being within their specific culture, following the intentions from WHO [1, 2]. Many measures of HRQoL concern special patient groups while others are generic, that is directed towards a general population [3]. The role of different aspects of autonomy for HRQoL among adolescents are however still poorly understood.

KIDSCREEN is a generic instrument measuring HRQoL in children and adolescents 8–18 years old. It was developed as an EU project including 13 European countries and was tested on over 3000 children. The instrument comes in 52, 27 and 10-item versions. All versions cover 10 dimensions: physical well-being, psychological well-being, moods and emotions, self-perception, parent relation and home life, financial resources, peers and social support, bullying, and autonomy [4]. Hence, autonomy is part of the very definition of the health-related quality of life of children and adolescents.

A close relation between autonomy and different measures of well-being is a common finding. Perceived autonomy in adolescence has been found to be negatively related to depression [5] and positively related to well-being [6]. Autonomy is also central to self-determination theory (SDT) in which autonomy, competence and relatedness constitute basic psychological needs and the satisfaction of these needs leads to increased well-being in both adolescents and adults [7], something that also has been verified in numerous empirical studies [8].

However, autonomy can be defined in many ways, for example as emotional independence, as volition versus pressure, functional independence and detachment, and its effect on health and well-being is therefore dependent on how it is defined [9]. Van Petegem and colleagues [9] investigated in two studies whether 8 and 14 scales related to autonomy could be reduced to a few dimensions. Indeed, the authors found that the different measures could be reduced to two approximately orthogonal dimensions labeled as “volition” and “distance and proximity”. Volition concern to what extent the adolescents experience they are acting from free will or coerced by someone else, and is influenced by SDT. Distance and proximity concern an emotional distance in the relations to the parents and is influenced by attachment theory [9]. While a high score on autonomy as volition was positively correlated to positive outcomes including well-being, a high score on autonomy as distance were negatively related to life-satisfaction and positively related to various problem behaviors [9]. Recently [10] it has become more common to distinguish between volition and independence where independence “is mainly about the question how much adolescents depend on others and who is regulating a certain behavior or goal (i.e., the parent, the adolescent or both)”. Volition and independence are often correlated, and it is a challenging research task to describe their interaction and under what conditions the outcome is beneficial [10].

The autonomy dimension in the KIDSCREEN instrument concern children’s opportunities to create social and leisure time [11] without further specification. Yet, it is quite clear from the items (i.e., Do you have enough free time to be with your mates?) that they concern autonomy as volition, and this is also supported by their positive relationship to the other dimensions in KIDSCREEN.

This conception of autonomy is compared to autonomy from a control perspective in the present study. Autonomy from this perspective emphasizes the importance of power and control for children’s development, in particular their experience of being seen and listened to and their opportunities to take part in decisions that concern themselves [12]. Yet, there is no available scale that directly address this aspect of autonomy in adolescence, although it is conceptually related to autonomy as independence [10]. To be clear, the concept of control is not new in relation to autonomy in adolescence, but it has been used to refer to parental control versus adolescence freedom of choice [13], and thus is akin to the volition-pressure dimension. The present measure concerns the adolescents’ sense of control and influence over things that happens in their lives.

The first aim of the present study was therefore to investigate the psychometric properties of a new scale measuring autonomy in adolescence from the control perspective (AAC) [12] including its relation to the autonomy dimension included in KIDSCREEN. A second aim was to investigate the relation between AAC and HRQoL. The second aim will be undertaken by investigating AACs ability to predict each of the 10 dimensions constituting KIDSCREEN-52 after controlling for extraneous variables.

Means (SD) for girls and boys in schools located in high and low SES areas for the AAC scale and all 10 dimensions of HRQoL from KIDSCREEN are shown in Table 1. A factor analysis was computed on all items measuring autonomy, that is, the 6 items constituting AAC and the 5 items constituting KSA. A scree plot indicated a two-dimensional solution explaining together 58% of the variance (36% + 22%). An oblique rotation using promax converged after 3 iterations and indicated a small correlation between the axes, r(153) = 0.205, p = 0.011. All five items from KSA loaded highest in the first dimension and all six items from AAC loaded highest in the second dimension (Table 2). The internal consistency yielded an alpha of 0.89 for KSA and 0.78 for AAC. The correlations for items within both scales were all significant (p < 0.01) although there were also some significant correlations between items across the scales (Table 3).

Next, a series of hierarchical regression analyses were computed in which gender and SES were inserted in a first block, KSA in a second block and ACC in a third block. Each dimension of HRQoL as measured by KIDSCREEN-52 served as dependent variable. An exception was the autonomy dimension of KIDSCREN for which the second block (KSA) was omitted. Collinearity statistics revealed that tolerance for all analyses were above 0.9 and VIF were below 1.1. Hence, multicollinearity was no problem.

The regression analyses showed that gender and SES (block 1) accounted together for between 1–16% of the variance in each dimension of HRQoL. They contributed significantly to Physical well-being (8%), Moods and emotions (16%), Self-perception (9%), Autonomy (5%), and Social acceptance and bullying (6%) (Table 4).

KSA (block 2) accounted for between 0–20% of the variance in addition to gender and SES (Δ R²). KSA contributed significantly to all HRQoL dimensions of the KIDSCREEN-52 instrument except to Social acceptance and bullying (Table 4), and was particularly high for Physical well-being (20%), Psychological well-being (19%), Moods and emotions (11%), Self-perception (10%), Parent relations and home life (16%), and School environment (19%).

AAC (block 3) accounted for between 3–16% of the variance in addition to gender, SES and KSA. High AAC contributed significantly to all HRQoL dimensions of the KIDSCREEN-52 instrument including Autonomy (for which only two blocks were computed). Although the unique contribution of AAC was significant for all dimensions, the contribution was small in half of the cases (5% or less). However, the contributions to the explained variance for Moods and emotions (14%), Self-perception (16%), Parent relations and home life (15%), Social acceptance and bullying (9%), and Financial resources (14%) were substantial. See Table 4. In agreement with the low correlation between the two dimensions of autonomy, volition versus control, the correlation between the explained variance by KSA and AAC was small, (r(8) = 0.038, p = 0.92). Hence, their relation is neither antagonistic nor concurrent.

Results from the full models including the third block (2^nd for Autonomy) in the regressions revealed that gender predicted HRQoL relative to Physical well-being, Moods and emotions, Self-perception, and Autonomy. Boys reported significantly higher HRQoL than girls in all four of the above dimensions. SES predicted HRQoL relative to Self-perception, Parent relations and home life, School environment, and Social acceptance and bullying. Adolescents from low SES district schools reported significantly higher HRQoL than adolescents from high SES schools for Self-perception, Parent relations and home life, and School environment, whereas the pattern was reversed for HRQoL related to and Social acceptance and bullying (Table 4).

The present study has shown that AAC, a new scale measuring autonomy in adolescence, based on their sense of control [12] has sound psychometric properties and is different from a measure of autonomy based on volition and pressure as expressed in the autonomy part of KIDSCREEN-52, here named KSA. Moreover, AAC predicted HRQoL in all 10 dimensions of the KIDSCREEN-52 instrument in addition to gender, SES and KSA. AAC accounted uniquely for 3–16% of the variance in HRQoL. The contribution from ACC was always positive, the more autonomy in the control sense experienced by the adolescence, the higher was their HRQoL. This finding supports earlier claims that autonomy is important for our well-being [5‐8] and extend it by showing that it is not only true for autonomy as volition but also for autonomy in the control sense, for example experiencing influence over what happens in the adolescent’s life [12].

In agreement with previous studies, we found that HRQoL was higher in 15–16-year-old boys as compared with same aged girls [14‐16], and that HRQoL was affected by SES [17‐19]. However, whereas other studies on SES have reported that adolescents from high SES families consistently have higher HRQoL, only one dimension (social acceptance and bullying) in the present study was in this direction. For the other three dimensions with significant contributions from SES (self-perception, parent relations and home life and school environment), low SES adolescents reported the highest HRQoL. One reason for this difference might be that SES in the present study was measured indirectly by school location and not on the individual level. This might have weakened the predictive power of SES. Yet, effects of gender and SES were included in the present study for the purpose of control and the sample size was too small to investigate their effect on HRQoL more thoroughly.

The small sample size was a major limitation to the main results of this study. However, the relation between AAC and a multidimensional measure of well-being such as KIDSCREEN-52 was a major strength of the present study, demonstrating its associations to many areas in life. It is plausible that the control aspect of autonomy is more important for adolescents than for younger children. Inclusion of younger age groups and ideally longitudinal studies are therefore needed in future studies.

To give a young person power and control over her or his life should not be confused with parental neglect. A multitude of sources has shown that parental neglect has aversive effects on children’s development [20]. However, a high score on the ACC scale indicate that parents care about their children. They listen to them and treat them justly and seriously. This is quite different from parent neglect. Yet, an empirical corroboration of this assumption would be valuable.

The present research also raises the question of whether HRQoL measures such as KIDSCREEN would be improved if more aspects than volition was measured in the autonomy dimension. This is a challenging thought but must stay as such until some of the limitations of the present study on AAC has been overcome. The KIDSCREEN group might also have some ideas about the relative contribution to HRQoL from each of the included dimensions.

Autonomy is a multifaceted concept and has been measured in many ways. A scale for measuring adolescent’s sense of control and influence over things that happen in their lives has been presented and shown to differ from autonomy as volition in adolescence 15–16 years old. Both forms of autonomy are positively related to HRQoL in adolescence.