In 2010, low fruit intake was ranked the fifth, physical inactivity the tenth, and low vegetable intake the seventeenth global risk factor for disease burden [1
]. For example, higher physical activity [2
], as well as high fruit, vegetable, and legume intake [3
] are associated with a lower risk of cardiovascular diseases and all-cause mortality. Additionally, a physically active lifestyle throughout childhood and adolescence has been found to be a factor preventing obesity in young adulthood [4
], while fruit and vegetable intake seems to be inversely, but weakly, associated with weight gain later in life [5
]. Since these two behaviors are relevant in regard to health, understanding how they are associated with one another may help in improving public health.
Whereas previous cross-sectional [6
] and prospective studies [7
] have shown that being physically active is associated with higher consumption of fruits and vegetables, the relationship between the pathways of leisure-time physical activity (LTPA) and fruit and vegetable consumption (FVC) from childhood to adulthood have rarely been researched. Since physical activity [8
] and dietary behaviors [9
] established in childhood and adolescence may track into adulthood, studying the associations between these two behaviors during transition phases from childhood to adolescence and from adolescence to adulthood is important.
A disadvantage of studying the tracking of behaviors is that it does not provide detailed information on subgroups that change their behavior over time. The recent advances in trajectory modelling techniques enable the study of behavioral heterogeneity at different phases of life in a data-driven way [11
]. For example, in the present instance, it is possible to identify distinctive trajectory classes (i.e., subgroups) of physical activity [13
] and FVC [14
]. Identifying the key groups of individuals and critical windows during the life course that would be the most receptive to physical activity and dietary promotion would contribute to the enhancement of public health.
The main objective of the present study, with a follow-up lasting over 30 years, was to examine the links between different LTPA trajectories and FVC from childhood to middle age. The study contributes to understanding more profoundly how LTPA develops between and within individuals and how inter- and intra-individual LTPA development is associated with FVC.
The aim of this study was to gain insight into the relationship between LTPA trajectories and FVC from childhood to middle age. The findings suggest that FVC is likely to be higher in individuals, especially males, who are persistently active from childhood to adulthood when compared to their persistently, less-active peers. These results mirror the findings from a previous longitudinal study suggesting that physical inactivity and unhealthy diet are predictive of each other in men [27
]. While FVC in females under 15 and males under 21 on the increasingly active trajectories resembled that of their inactive and low-active counterparts, the increasers of LTPA ended up having the highest FVC in adulthood. This tendency was particularly evident among females. In turn, FVC was higher among the decreasingly active than inactive females up to age 39 after which their FVC no longer differed from one another. Similarly, after males turned 27, the FVC of the decreasingly active males was no longer significantly higher when compared to their low-active peers. The results are in line with previous findings, both cross-sectional [6
] and longitudinal [14
], confirming that higher FVC and LTPA tend to occur in the same individuals. The results also add to previous knowledge by showing how the changes in FVC closely parallel the distinctive contours of each LTPA trajectory over time at many age points.
The general tendency of FVC was declining across all the LTPA trajectory classes during teenage years after which the tendency changed. The only exception was the females following the decreasingly active trajectory with their FVC declining later. The same declining tendency of FVC during adolescence and young adulthood has been observed in a previous longitudinal study [10
]. Even though the tendency of FVC might be declining during teenage years, recent evidence from time trend analyses shows how the overall daily FVC between years 2002–2010 has increased among European and US adolescents aged 11–, 13– and 15 years [29
The present study suggests that FVC level is associated positively not only with persistent activity level, but also with changes in LTPA from childhood to middle age. For example, an increase in LTPA among the increasingly active males and females was observed in young adulthood simultaneously with the increase in FVC. However, among the increasingly active males, LTPA increased only during young adulthood (ages 18–24) with a momentary high peak of FVC at this age, while another, more continuous change towards higher FVC was not observed until later in adulthood. In Finland, most young adults move from the family home just between 18 and 24 years of age [30
]. This sort of transition periods of life may change food choices [31
] which might explain the momentary peak of higher FVC among the increasingly active males and possibly also among the persistently active females at the age of 21. In contrast, the increase initiated at the age of 18 in FVC among females in the increasingly active trajectory continued until the end of the study period. Women tend to abandon unhealthy behaviors in young adulthood more often than men [32
], which might explain why the favorable co-variance in the two behaviors was continuous for females starting already from the age of 18. Additionally, adult women have more pronounced health beliefs, better nutritional knowledge and attach greater importance to a healthy diet than adult men [33
] which might explain the more consistent increase between the two behaviors among the increasingly active women. These results reflect the findings from a previous longitudinal study concluding that an increase in physical activity, in contrast to a decrease in physical activity, is associated with greater improvements in diet quality [36
Reciprocally, the FVC of the decreasingly active participants dropped to the level of the persistently inactive participants among females and low-active participants among males in adulthood. Thus, decreasing LTPA may pose an additional health risk, due to simultaneous detrimental changes in diet. All in all, only 22% of women and 14% of men in Finland attain the recommended intake of fruits and vegetables [37
] with FVC remaining low also worldwide when compared to the recommended level [38
]. However, the decreasingly active males of the current study managed to increase their FVC again after turning 40. Indeed, an increasing trend in FVC was not observed only among the increasingly active participants, but also among the persistently active females and persistently low-active and decreasingly active males in adulthood. An increasing trend in FVC during recent decades has been reported among Finnish adults [39
]. The improved availability of fruits in Northern Europe [41
] and the affordability of fruits and vegetables in high-income countries when compared to low-income countries [38
] probably explains this positive trend in FVC. Also, having lunch in a staff canteen has been found to be associated with higher FVC and improved diet among Finnish adult employees [42
], suggesting that endeavors put into developing healthier catering services in Finland may also play a role in improving adults’ dietary behavior. Furthermore, the consistent effort put in developing nutrition policy and Finnish nutrition guidelines [43
] might have an effect on the favorable development of FVC. For example, when comparing the food pyramid in the national dietary recommendations in 1998 [44
] to the food pyramid in the latest recommendations in 2014 [45
], the latter recommends higher consumption of fruits and vegetables. Also, the guidelines in 2005 [46
] recommended a level of 400 grams of fruits and vegetables per day, while the recommended level is nowadays 500 grams [45
]. Diet following Finnish nutrition recommendations has been found to be inversely associated with waist circumference and body fat percentage [47
] suggesting that consuming fruits and vegetables according to recommendations improves health.
Previous studies highlight the importance of, e.g., BMI, socioeconomic status, living environment, perceived health, life events, and other factors in explaining levels of physical activity or FVC [38
]. Although the models used in the present study were adjusted for BMI, education and total energy intake (only in adulthood), it should be borne in mind that factors other than the two behaviors studied here may explain why FVC was higher among the persistently and increasingly active than low-active or inactive participants. For example, being on an increasingly or persistently active trajectory may be an indicator of an overall positive health orientation [6
], while being on an inactive or low-active trajectory could be an indicator of an overall negative health orientation. A previous study on participants aged 13 to 30 years of age supports this interpretation: The authors identified an overall unhealthy trajectory in which daily fruit intake and regular exercise decreased with higher rates of smoking and inebriation [14
Since the development of the two behaviors seems to occur in tandem, future interventions should study whether to target these two behaviors simultaneously, or is it enough to target one of them and improvements in the other behavior will follow. So far, short term interventions studying the issue have had inconsistent findings with two concluding that improvements in physical activity do not lead to healthier diet [53
], and another one showing that by increasing physical activity improvements in diet can be achieved, but only among boys [55
]. Future studies should also investigate the factors that determine why certain people end up on unfavorable and others on favorable dietary and physical activity trajectories and what are the factors enabling the favorable changes of the two behaviors during the lifespan.
This study has several strengths, including a large sample size comprising six age cohorts, recruitment of participants from across Finland, a 30–year follow-up starting from childhood, and multiple follow-up measurements of LTPA and FVC. Together, these features of the YFS made it possible to study the complex relationship between LTPA and FVC during the life course. The use of a novel, data-driven statistical method enabled the identification of LTPA trajectories in which participants’ LTPA developed similarly, while differing from those in the other trajectories [11
]. As recommended in the previous literature, whole foods, here fruits and vegetables [56
], were used as units of healthy nutrition [57
] instead of specific nutrients (e.g., vitamin C). Moreover, the study utilized the STROBE-nut checklist (Supplement 2
) and a modified version of the GRoLTS checklist (Supplement 3
) for ensuring appropriate quality in reporting the results. The GRoLTS checklist was modified as not all the items in the list were suited to latent profile analyses, but instead to latent growth mixture modelling and latent class growth analyses.
This study has its limitations. LTPA and FVC were self-reported, which may produce biased results. There is a possibility for recall bias with light leisure-time physical activities being generally harder to recall than vigorous ones [58
]. Social desirability bias, meaning the tendency to over-report desired behaviors and underreport undesired ones, is also present with self-reported data [59
]. Self-reported physical activity is commonly estimated higher than objectively measured physical activity [60
], and study participants, especially women, tend to overestimate their consumption of foods that are considered to be healthy [61
]. After participants turned 30, the method used to assess diet changed from a simple 19–item form to a more comprehensive FFQ. The validity of the latter questionnaire is presumably higher than that of the former one as a previous study showed how a similar comprehensive FFQ as used in the current study performed more accurately than 7– and 16–item questionnaires [62
]. However, these biases do not necessarily invalidate the results since the measurement instruments are used for ranking individuals and not analyzing the exact amount of fruits and vegetables consumed nor the exact intensity (low, moderate or vigorous) of physical activity. The sample was representative of the general Finnish population. Hence, the present results cannot be generalized to other populations, especially those in low- or middle-income countries or with diverse ethnic groups.
Trajectory modelling also has its limitations. For example, each trajectory is a description of the subgroup’s mean behavior, and thus, no participant is likely to follow the identified trajectory precisely [63
]. This is a reliability issue, especially in the small trajectory classes. Also, the mean FVC across the LTPA trajectories might not be reliable at those five age points where the residual variances of the FVC needed to be fixed to a value with the lowest Akaike’s Information Criterion in order to get a result (see FVC mean values marked with c in Table 3
). Especially the persistently active women at the age of 21 and at the age of 48 seemed to have a deviant shift in their FVC mean values (Figure 3
). On the other hand, this trajectory class was extremely small (3.4% of the sample) which is why a change in the answers of just a few participants may, indeed, change the mean FVC considerably. Finally, although the final number of LTPA trajectory classes was based on objective index values, subjective interpretation was also used, a procedure that could induce selection bias.