Life satisfaction

Life satisfaction is defined as an overall cognitive and judgmental assessment of personal quality of life according to self-selected criteria [1]. Although personal satisfaction may vary across specific life domains, most previous studies have focused on global life satisfaction; defined as a unidimensional characteristic [2]. Life satisfaction is a relatively stable characteristic, more tied to systematic and chronic factors, than to temporary and transitory affective states [3]. It has been shown that cognitive judgment of personal life satisfaction is an essential component of subjective well-being and a valid indicator of overall life quality [4]. As such, life satisfaction is one of the central constructs in the field of positive psychology [5]. Prospective studies suggest that lower life satisfaction is associated with an increased risk of disease-related, injury-related, and all-cause mortality [6–8], suicidal behaviour [9], major weight gain [10], and moderate or severe depression [11]. It has been suggested that low life satisfaction may negatively affect subsequent income, employment and marital status [12]. Furthermore, cross-sectional studies have shown that poor life satisfaction is associated with a range of unhealthy behaviours [13,14] and youth violence [15].

Average life satisfaction among adults across nations ranges between 60% and 80% of its maximum score and significantly varies between individuals [16,17]. Interestingly, more than 75% of young adults are not completely satisfied with their lives [14]. An international study has shown that the average life satisfaction of adolescents varies between ~55% and 75% of its maximum score across different nations [18]. Among children from 36 countries the mean life satisfaction score was found to be 76% of its maximum [19]. Although most children rate their lives positively, there is still a significant proportion who report very low levels of life satisfaction [2]. To effectively identify and influence individuals and groups at the highest risk, public health strategies need to be informed by empirical evidence about the determinants of life satisfaction. Previous studies have investigated the demographic, psychological, social, environmental, cultural, behavioural and health-related determinants of life satisfaction [5,20].

Furthermore, different theories have been developed to explain the determinants of life satisfaction. Activity theory suggests that life satisfaction is determined by the frequency of participation in given activities and the degree of intimacy associated with the activities [21]. Although the theory was originally developed to help explain the process of successful aging, its premises regarding life satisfaction have also been confirmed among a general adult population [22]. Interestingly, in an empirical examination of activity theory Longino and Kart [23] found that life satisfaction is: 1) positively affected by participation in informal activities, 2) not impacted by participation in solitary activities, and 3) negatively affected by participation in formal activities. Need theory implies that life satisfaction is mostly regulated by the ability of an individual to satisfy his/her biological and psychological needs [24]. Rodriguez et al. [22] suggested that ‘fulfilling needs’ is a better predictor of life satisfaction than ‘participation in activities’. Their study identified ‘satisfying the social need’ and ‘satisfying the autonomy need’ as the strongest predictors of life satisfaction.

Relationship between physical activity and life satisfaction

It has been shown that regular physical activity (PA) may be associated with increased life satisfaction among all age groups: children and adolescents [5,25–30]; young adults [14,31–37]; adults [38–43]; and elderly [44–48]. Life satisfaction has been shown to be positively related to overall leisure-time PA [31,34,37,39,42], participation in sports/exercise [14,22,25–30,32,35,38,41,43,44], and participation in several specific types of PA, such as stretching [26], strength training [22,26], jogging [39], and walking [44,45,48]. By contrast, mixed results were observed for non-domain-specific PA (sum of PA levels across two or more domains), indicating its positive [36,37,40,46] or non-significant [33,36,37,48,49] relationship with life satisfaction. Interestingly, a recent study separately examined PA in sporting clubs, gymnasia, and walking among Australian rural-living women and found a significant association with life satisfaction only for type, but not for level of PA [50].

Distinct effects of domain and intensity-specific PA levels

The World Health Organization (WHO) suggests that moderate or vigorous-intensity PA (MVPA) required for health benefits can be accumulated in any domain—work, domestic, transport, or leisure-time [51]. However, recent studies have shown that different domains of PA may have distinct associations with health-related and overall quality of life [52–56], symptoms of depression [57–59], symptoms of anxiety [60] and all-cause mortality [61]. For example, a study by Jurakic et al. [52] has shown that leisure-time PA may be positively, whilst transport-related and domestic PA are inversely related to health-related quality of life. In another study, lower depression was associated with higher levels of leisure-time PA and lower levels of occupational PA [58]. Furthermore, the magnitude and direction of the relationships between different domains of PA and health outcomes may be moderated by age and gender [52,53,55]. For instance, Peeters et al. [55] found that higher levels of domestic PA are associated with increased well-being among mid-aged and older women and with decreased well-being among young women. Cerin et al. [53] found no association between household and occupational PA and self-perceived mental health among men, but they observed a negative relationship between these variables among women. Furthermore, previous studies have also indicated that different intensity-levels (e.g. moderate-intensity, vigorous-intensity) may have distinct relationships with mental health outcomes [53,57]. For example, self-perceived mental health was positively related to vigorous-intensity leisure-time PA, whilst it showed no significant relationship with moderate-intensity PA in the same domain [53]. It seems, therefore, that non-domain-specific and non-intensity specific PA guidelines may not be most effective for achieving all psychological benefits. The emerging evidence about differential effects of domain and intensity-specific PA calls for further research, particularly regarding psychological outcomes that have not been studied in this context.

Many studies examining PA and life satisfaction have assessed physical exercise or sports participation as specific types of leisure-time activity [14,22,25–30,32,33,35,38,41,43,44,62]. Several studies assessed overall leisure-time PA [31,34,37,39,42] or non-domain-specific PA [33,36,37,40,46,48,49]. Specific types of PA were separately analysed by Valois et al. [26] (walking/riding a bike), Morgan and Bath [45] (walking), and Withall et al. [48] (total number of steps taken). Morgan and Bath [45] also separately assessed PA level in the domestic domain. This shows that most previous studies on the relationship between PA and life satisfaction have focused on total PA and the leisure-time domain, while evidence for work, domestic, and transport domains is scarce. Moreover, none of the studies assessed all PA domains (work, transport, domestic, leisure-time) separately. Furthermore, most studies used single-items or short questionnaires which did not allow to separately examine different intensities (low, moderate, vigorous) or types of PA (e.g., walking, cycling). Although few studies used instruments that would allow them to separately analyse different intensity levels, such as the International PA Questionnaire—Short form (IPAQ-short) [36,37] and accelerometers [36,48], none of them reported separately on intensity-specific PA levels. In addition, none of the studies used instruments that would allow for separate assessment of intensity levels within specific domains of PA, i.e. domain and intensity-specific PA (e.g. vigorous-intensity PA in the work domain, moderate-intensity PA in the work domain, vigorous-intensity PA in the domestic domain, moderate-intensity PA in the domestic domain). Therefore, thorough information about the relationship between life satisfaction and domain and intensity-specific PA levels is still lacking.

Aims and hypotheses

The magnitude and direction of the relationship between PA and life satisfaction may be dependent on the study population and instruments used. In order to compare the relative contributions of different domain and intensity-specific PA levels to life satisfaction, it is essential to assess them all within a single study. Besides, studies have shown that age can act as a moderator of the determinants of life satisfaction [63], which indicates the importance of research within specific age groups. Hence, the aim of this study was to examine the relationship between life satisfaction and intensity-specific PA levels in work, domestic, transport and leisure-time domains among university students. Better understanding of this relationship may increase the public health utility of PA intervention strategies to influence life satisfaction among university students, and help elucidate whether non-domain and non-intensity specific PA recommendations are justified in such interventions. The hypothesized relationships are presented in Fig. 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Hypothesised relationships between domain and intensity-specific physical activity levels and life satisfaction.

Right dashed arrow: hypothesized relationship tested in the present study. Sign above the arrow: hypothesized direction of the relationship (‘+’ = positive; ‘-’ = inverse).

https://doi.org/10.1371/journal.pone.0118137.g001

The hypotheses were proposed based on: 1) activity theory, 2) need theory, and 3) previous empirical findings.

Hypothesis 1: Leisure-time PA is positively associated with life satisfaction

The positive relationship between leisure-time PA and life satisfaction was hypothesized for several reasons. Firstly, we assumed that among young adults the degree of intimacy associated with leisure-time PA is higher than with activities in other domains. Secondly, participation in leisure-time PA may satisfy the social need, autonomy need, and physical fitness need, that are shown to be among strongest predictors of life satisfaction [22]. Thirdly, several previous studies have found a positive relationship between leisure-time PA and life satisfaction among young adults [14,31,32,34,35,37].

Hypothesis 2: Domestic and transport-related PA levels are positively associated with life satisfaction

We assumed that domestic and transport-related activities are mostly solitary for young adults. As such, according to Longino and Kart [23], they would not be expected to show significant association with life satisfaction. Accordingly, it can be assumed that these activities have very little capacity to contribute in satisfying the social need of young adults. Jurakic et al. [52] suggested that some people may even perceive the participation in these activities as a restriction on their autonomy. Some people may participate in such activities because they do not have an alternative (e.g. they have to cycle to work, because they cannot afford to buy a car). This may suggest that according to the need theory these activities are not likely to increase life satisfaction. Despite that, previous studies have indicated that participating in domestic PA, such as gardening and housework, may be positively related to life satisfaction [45,64,65]. Furthermore, a recent review suggested that active transport has potential to indirectly increase subjective well-being through its positive impact on travel satisfaction [66]. It therefore seems possible that these domains of PA are linked to life satisfaction through mechanisms that are not explained by activity theory and need theory. Hence, based on the recent empirical findings, we hypothesized a positive relationship between life satisfaction and domestic and transport-related PA.

Hypothesis 3: Occupational PA is negatively associated with life satisfaction

We assumed that for young people occupational activities are contextually most similar to formal activities described by Longino and Kart [23], which would suggest that they are likely to be inversely related to life satisfaction. It should be noted, however, that this assumption may equally relate to participation in both sedentary and physically demanding occupational activities. Furthermore, a study has shown that a high physical work load is associated with low job control [67]. We therefore assumed that physically demanding occupational activities may reduce young people’s self-perceived autonomy by imposing restrictions and control, and consequently negatively affect their life satisfaction. Furthermore, previous findings about the relationship between occupational PA and psychological health and well-being were equivocal. For example, Laaksonen et al. [67] have shown that higher physical work load is associated with poor self-reported mental health among middle-aged employees, whilst several other studies found no relationship between occupational PA and overall and health-related quality of life [52–54,56]. Despite these equivocal findings, based on the above-mentioned assumptions from the activity theory and need theory, we hypothesized a negative relationship between occupational PA and life satisfaction.

Hypothesis 4: The strength of the relationship between PA and life satisfaction increases with increased intensity of PA

When compared to moderate-intensity and walking, vigorous-intensity PA has shown stronger associations with self-reported mental health (among men) [68], mental well-being (among both genders) [53], and symptoms of depression, anxiety and somatisation (among men) [69]. We assumed that the effect sizes of the relationships between intensity-specific PA levels and life satisfaction may follow a similar pattern.

Additionally, we examined the associations between life satisfaction and: 1) meeting the WHO MVPA recommendations; 2) total (i.e. non-domain-specific and non-intensity specific) PA level; 3) gender; 4) age; 5) disposable income; 6) community size; 7) smoking; 9) alcohol intake; 10) body mass index (BMI); and 11) self-rated health.

Participants

The study was carried out in a stratified random sample of students living in the university residence halls in Zagreb, Croatia. A two-stage stratification was performed according to gender and all five university residence halls in Zagreb. During the second week of June 2009, 1750 students were invited by mail to take part in the study. A week later a second invitation was sent to the students who did not respond initially. In total, 71.7% of invited students consented to participate (n = 1254), and were surveyed in the official premises of their resident halls. Eight trained surveyors participated in the process of data collection. The survey took place during the exams period. The sample included students from all 29 faculties of the University of Zagreb and 7 private colleges. With approximately 70,000 students enrolled in its programs, the University of Zagreb is the largest university in Croatia and one of the largest higher education institutions in the European Union. The random sample covered students from all major academic disciplines in natural sciences, social sciences, arts and humanities, formal sciences, and applied sciences and professions. A total of 1163 participants (62.4% females; mean ± standard deviation age, 21.5 ± 1.8 years; 2.6% enrolled in physical education or sport science academic degrees) with data on PA comprised the final analytic sample. A post-hoc power analysis indicated that the sample size was sufficient to detect a significant multiple correlation coefficient at p < 0.05 with the power of 85%, if the population effect size was at least small (Cohen f² ≥ 0.02) [70]. Participation was voluntary and all participants gave written informed consent before entering the study. The study was approved by the Institutional Review Board at the Faculty of Kinesiology, University of Zagreb. Socio-demographic and other relevant characteristics of the sample are presented in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Sample characteristics (n = 1163).

https://doi.org/10.1371/journal.pone.0118137.t001

Instruments

PA levels were assessed using the International Physical Activity Questionnaire (IPAQ)—long form [71]. The questionnaire consists of 27 items about the frequency and usual duration of PA and sedentary behaviour in the last 7 days. It provides intensity-specific data on PA in the domains of: 1) work (vigorous intensity, moderate intensity, and walking); 2) transport (walking and cycling); 3) domestic and garden (vigorous-intensity yard chores, moderate-intensity yard chores, and moderate-intensity household chores); and 4) leisure-time (vigorous intensity, moderate intensity, and walking). In the present study, the domain and intensity-specific PA levels were expressed in minutes per week. Intensity-specific scores were calculated by summing all the minutes spent: 1) walking; 2) in moderate-intensity PA; and 3) in vigorous-intensity PA. Domain-specific scores were calculated by summing all the minutes spent in: 1) work-related PA; 2) transport-related PA; 3) domestic PA; and 4) leisure-time PA, regardless of their intensity. The total PA score was calculated as a sum of all domain and intensity specific PA minutes weighted by their metabolic equivalents (METs) and expressed in MET-minutes per week, according to the official IPAQ data processing guidelines [72]. Additionally, we calculated the total time spent in PA as a simple sum (non-intensity-weighted) of all domain and intensity specific PA minutes. The achievement of the WHO recommendations in terms of: 1) standard MVPA recommendation (≥ 75 minutes/week of vigorous-intensity PA or ≥ 150 minutes/week of moderate-intensity PA or an equivalent combination of both); and 2) MVPA recommendation for additional health benefits (≥ 150 minutes/week of vigorous-intensity PA or ≥ 300 minutes/week of moderate-intensity PA or an equivalent combination of both) [51], was also calculated from the IPAQ data. Previous studies have demonstrated good measurement properties of the IPAQ—long form [71,73]. A detailed description of the questionnaire can be found elsewhere [72].

Life satisfaction was measured using the Satisfaction With Life Scale (SWLS) [1]. The scale consists of 5 statements (for example, “In most ways my life is close to my ideal.”) and participants rated their agreement or disagreement with each statement using a 5-point Likert scale. Overall life satisfaction (on the scale from 5–25) was calculated as the sum of responses to all items, where a higher score indicated a higher level of life satisfaction. Previous studies have consistently demonstrated high reliability and a single-factor structure of SWLS items [74,75].

A separate set of questions asked about: gender; age; size of community, i.e. the number of inhabitants of city/town/village of participant’s origin ({1} < 2000, {2} 2001–10 000, {3} 10 001–100 000, and {4} > 100 000); disposable income; smoking status ({1} never smoked, {2} quit smoking, {3} smoking <5 cigarettes/day, {4} smoking 5–10 cigarettes/day, {5} smoking 10–20 cigarettes/day, or {6} smoking >20 cigarettes/day); alcohol intake ({1} never, {2} <1 day/week, {3} 1–3 days/week, {4} 4–6 days/week, {5} everyday); self-reported height and weight (used to compute body mass index (BMI; kg/m²); and self-rated general health (single item “In general, how would you rate your health?” with the following response scale: {1} Poor, {2} Fair, {3} Good, {4} Very Good, and {5} Excellent).

Data analysis

The data on socio-demographic and lifestyle characteristics, life satisfaction, and self-rated health were presented as mean ± standard deviation for quantitative variables, and percentages for qualitative variables. Due to their non-normal distribution, the continuous PA data were presented using medians and interquartile ranges, which is in accordance with the official guidelines for data processing and analysis of the IPAQ [72]. The relationship between PA behaviour (independent variables) and life satisfaction (dependent variable) was assessed by 7 separate linear regression models. The independent variables in the regression models were: (i) meeting the WHO standard MVPA guidelines (Model 1); (ii) meeting the WHO MVPA guidelines for additional health benefits (Model 2); (iii) total volume of PA expressed in MET-hours/day (Model 3); (iv) total duration of PA expressed in hours/day (Model 4); (v) intensity-specific PA levels (vigorous-intensity, moderate-intensity, and walking) (Model 5); (vi) domain-specific PA levels (work, transport-related, domestic, and leisure-time) (Model 6); and (vii) domain and intensity-specific PA levels (vigorous-intensity at work, moderate-intensity at work, walking at work, cycling for transport, walking for transport, vigorous-intensity yard chores, moderate-intensity yard chores, moderate-intensity house chores, vigorous-intensity in leisure-time, moderate-intensity in leisure-time, and walking in leisure-time) (Model 7), whilst the total SWLS score was used as the dependent variable in all models. In the first step, Models 1–4 and Models 5–7 were analysed using the simple linear regression (with no adjustments) and the multiple linear regression (with mutual adjustments of PA variables within the model), respectively. In the second step, the regression models were additionally adjusted for gender, age, size of community, disposable income, smoking status, alcohol intake, BMI and self-rated health, because previous studies have indicated that these variables could act as confounders of the proposed relationship [63,76–79]. Gender was entered to the model as a dichotomous variable, size of community, smoking status, alcohol intake and self-rated health as ordinal variables, and age, disposable income and BMI as continuous variables. Prior to the regression analysis, reported alcohol intakes on ‘4–6 days/week’ and ‘everyday’ were collapsed into a single category, due to the low number of such responses. Standardized and unstandardized regression coefficients, partial correlations, and coefficients of multiple determination were calculated and tested for statistical significance. The unstandardized regression coefficients for continuous PA variables were presented in hours/day instead of minutes/week. The partial correlations were interpreted as a measure of relative effect size, and classified as: small (0.10–0.29); medium (0.30–0.49); and large (≥0.50), according to Cohen [70]. Further, a previous study has shown that gender may act as a moderator of the relationship between PA and mental well-being [53]. Therefore, using the General Linear Models (GLM) procedure we tested all the unadjusted and adjusted models for the possible interaction effect between gender and PA variables. Prior to the analysis, all regression models were inspected for the normality of residuals, linearity of relationships, multicollinearity, and heteroscedasticity using the normal probability plot, bivariate scatterplot of each independent variable vs. residuals, variance inflation factors, and predicted vs. residuals plot, respectively. All analyses were performed using IBM SPSS Statistics 21 (SPSS Inc. an IBM Company, Chicago, IL, USA), with the level of statistical significance set at p < 0.05.