The main findings of the present study were that CRF significantly predicted TDS in Norwegian adolescents while controlling for muscular strength, body composition, socioeconomic status, school clustering, sex and domestic/foreign birthplace. The results indicated that CRF was the only aspect of physical fitness associated with TDS as a measure of mental health, thus excluding body composition and muscular strength as predictors.
Muscular strength, body mass index and metal health
A significant association between muscular strength and TDS was initially found. However, when controlling for CRF there was no association between these variables. A possible explanation is that the participants with high CRF were also likely to have a relatively high muscular strength . However, based on the fully adjusted model, it can be postulated that participants with high muscular strength did not necessarily have high CRF. A possible interpretation is that muscular strength in adolescents is generally a natural consequence of the individual’s CRF level, which may represent the true association with mental health.
The present findings support previous studies that have suggested CRF to be the only health-related aspect of fitness associated with mental health outcomes such as quality of life , depression [27, 50] and well-being . Many studies that have found associations between muscular strength and mental health outcomes in adolescents did not measure CRF [24, 52]. The present findings did not show an association between BMI and mental health, independent from controlling for CRF in the regression. This is congruent with the review by Luppino et al. , who found an association between overweight and depression in adults, but not in individuals younger than 20 years. This indicates a different relationship between age groups; however, it is important to point out that none of the reviewed studies controlled for CRF. Although it is possible that muscular strength and BMI are associated with mental health outcomes, studies that do not also measure CRF lack important information. Had we not controlled for all fitness variables in the present study, we would have erroneously concluded that muscular strength was associated with TDS. Opposing findings by Kettunen et al.  showed muscular strength to be more important than CRF to reduce stress in adults. However, this study categorized continuous variables and employed ANOVAs, which has been strongly advised against by Altman and Royston , and might have produced biased results. Additionally, associations between physical fitness and mental health might be different in adult and adolescent populations. For instance, many experimental studies have found effects of strength training on mental health in older adults . Positive effects in older adults are not surprising, considering how strength training can reverse muscle atrophy and improve the daily functioning of older people . In adolescents however, muscular strength is mainly associated with appearance-related mental health outcomes, such as self-perception, perceived physical appearance or physical self-worth . Future studies of associations between health-related physical fitness and mental health should include different mental health outcomes, to gain a better understanding of whether specific aspects of fitness are associated with specific outcomes of mental health.
Cardiorespiratory fitness and mental health
The present findings showed that higher CRF predicted improved mental health. Although a causal direction between CRF and mental health cannot be established from cross-sectional findings, recent evidence has indicated a one-directional causal relationship for physical activity as a protective factor against depression among adults . High-intensity exercise is an important factor for high CRF , hence results from the present study support a hypothesis suggesting that high-intensity exercise might be more favorable for mental health than low-intensity exercise. This is in accordance with the study by Parfitt, Pavey and Rowlands , who found high-intensity exercise to be more favorable for mental health than light-intensity exercise, in a population of children. Furthermore, the meta-analysis by Ahn and Fedewa  found high-intensity exercise RCT interventions to have the most effect on children’s mental health. On the other hand, Helgadóttir et al.  concluded that low-intensity exercise was more effective on depression treatment than high-intensity exercise in an adult population. The low-intensity group exercised with yoga and this type of exercise may have a distinct relationship with mental health. However, the results should be treated with caution, because the intervention had low adherence and did not mention how this differed between exercise groups. Additionally, 12 months after the intervention, there were no significant differences between the low- and vigorous-exercise groups. The study by Helgadóttir et al.  is incongruent with the previously mentioned studies, as well as what Bailey et al.  suggested to treat depression in adolescents: “…aerobic-based activity of moderate-to-vigorous intensity.” It is also possible that intensity might not even be especially crucial, as long as CRF is improved. Shepherd et al.  prescribed high-intensity interval training and moderate-intensity continuous training in two groups of inactive adults and both groups experienced increased CRF and improved mental health. Few studies have examined the causal relationship between increased CRF and improved mental health outcomes, but a recent longitudinal study by Rahman et al.  showed that improved CRF predicted at least a 50% reduction in depression scores for adults. Ruggero et al.  found that high CRF at baseline was associated with lower levels of depression a year later in adolescent girls and suggested that CRF might mediate the effect physical activity has on depression. This was supported by Eddolls et al.  who concluded that CRF mediated the relationship between vigorous physical activity and mental health in adolescents, thus suggesting that physical activity interventions to treat depression may only be effective if they improve CRF.
There may be several potential mechanisms underlying the associations between CRF and mental health. The endocannabinoid system, which mediates high-intensity exercise effect on depression , is a possible neurobiological explanation. Another potential explanatory mechanism has to do with brain function, as discussed in a recent review : Depression is associated with reduced brain connectivity ; therefore, CRF’s protective role against depression might be due to the relationship between CRF and the improved functional connectivity between regions of the brain known as the default mode network [15, 67]. CRF in adolescents is associated with self-esteem and body satisfaction , and with participation in common team sports like football, handball and basketball , which are important arenas for social relationships. It is therefore not unlikely that the psychosocial mechanism  also had a mediating role in the present results. The topic of explanatory mechanisms between physical fitness and mental health requires more research, especially on adolescent populations, in order to fully understand the relationship between the relevant variables. Additionally, future studies need to examine how exercise at different intensities affects different mental health outcomes, and whether the results are influenced by increases in CRF. Such knowledge can be useful in efforts to prevent or treat mental disorders.
4.5 Strengths and limitations
Strengths of the present study include the large sample size from separate geographical regions, the use of three objectively measured health-related aspects of physical fitness, and the control of relevant covariates.
The main limitation of the present study was a large number of missing values; however, the extensive missing value analyses indicated that the main results most likely were unaffected by the dropouts.
The present internal consistency results pertaining to TDS were quite low and are similar to Italian , Finnish  and Dutch  results. Internal consistency results from English speaking populations [71, 72] are usually higher, which suggests that statements are better understood by native English speakers, while non-native English speakers may misinterpret the statements somewhat. Age is also a factor, as the internal consistency is lower for younger adolescents, such as the present population, compared to older adolescents as examined in studies by Bøe et al.  and Sagatun et al. . Finally, the cross-sectional nature of the study limits the ability to make any causal inference.