Associations of Leisure-Time Physical Activity Trajectories with Fruit and Vegetable Consumption from Childhood to Adulthood : The Cardiovascular Risk in Young Finns Study

International Journal of Environmental Research and Public Health Article Associations of Leisure-Time Physical Activity Trajectories with Fruit and Vegetable Consumption from Childhood to Adulthood: The Cardiovascular Risk in Young Finns Study Irinja Lounassalo 1,*, Mirja Hirvensalo 1, Anna Kankaanpää 2, Asko Tolvanen 3, Sanna Palomäki 1, Kasper Salin 1, Mikael Fogelholm 4, Xiaolin Yang 2, Katja Pahkala 5,6, Suvi Rovio 5,6, Nina Hutri-Kähönen 7, Olli Raitakari 5,6 and Tuija H. Tammelin 2 1Faculty of Sport and Health Sciences, University of Jyväskylä, 40014 Jyväskylä, Finland; [email protected] (M.H.); [email protected] (S.P.); kasper[email protected] (K.S.) 2LIKES Research Centre for Physical Activity and Health, 40700 Jyväskylä, Finland; [email protected] (A.K.); [email protected] (X.Y.); [email protected] (T.H.T.) 3Methodology Center for Human Sciences, University of Jyväskylä, 40014 Jyväskylä, Finland; [email protected] 4Department of Food and Nutrition, University of Helsinki, 00014 Helsinki, Finland; [email protected] 5Research Centre of Applied and Preventive Cardiovascular Medicine, University of Turku, 20014 Turku, Finland; [email protected] (K.P.); [email protected] (S.R.); [email protected] (O.R.) 6 Department of Clinical Physiology and Nuclear Medicine, Turku University Hospital, 20521 Turku, Finland 7Department of Pediatrics, Tampere University and Tampere University Hospital, 33520 Tampere, Finland; [email protected] *Correspondence: [email protected]; Tel.: +358-50-5468744 Received: 11 October 2019; Accepted: 6 November 2019; Published: 12 November 2019   Abstract: A physically active lifestyle and a diet rich in vegetables and fruits have a central role in promoting health. This study examined the associations between leisure-time physical activity (LTPA) trajectories and fruit and vegetable consumption (FVC) from childhood to middle age. The data were drawn from the Cardiovascular Risk in Young Finns Study with six age cohorts. Participants were 9 to 18 years (n=3536; 51% females) at baseline in 1980 and 33 to 48 years at the last follow-up in 2011. LTPA and FVC were self-reported. LTPA trajectories were identified using latent profile analyses, after which the mean differences in FVC across the trajectories were studied. Active, low-active, decreasingly and increasingly active trajectories were identified for both genders. An additional trajectory describing inactivity was identified for females. Those who were persistently active or increased their LTPA had higher FVC at many ages when compared to their inactive or low-active counterparts (p<0.05). In females prior to age 42 and in males prior to age 24, FVC was higher at many ages in those with decreasing activity than in their inactive or low-active counterparts (p<0.05). The development of LTPA and FVC from childhood to middle age seem to occur in tandem. Keywords: physical activity; diet; trajectory; longitudinal; childhood; adolescence; adulthood 1. Introduction 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 Int. J. Environ. Res. Public Health 2019,16, 4437; doi:10.3390/ijerph16224437 www.mdpi.com/journal/ijerph Int. J. Environ. Res. Public Health 2019,16, 4437 2 of 17 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 , 10 ] 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 , 12 ]. 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 interand intra-individual LTPA development is associated with FVC. 2. Materials and Methods 2.1. Study Design and Participants The Cardiovascular Risk in Young Finns Study (YFS) is an ongoing, longitudinal, population-based study whose participants have been randomly selected from five Finnish university cities with medical schools: Helsinki, Kuopio Oulu, Tampere, and Turku. Urban and rural areas in and around these cities were included. The study comprises six age-cohorts born in 1962, 1965, 1968, 1971, 1974 and 1977. At baseline in 1980, 3596 boys and girls aged 3 to18 years participated in the study (83% of those invited, n=4 320). Follow-up studies have been conducted in 1983, 1986, 1989, 1992, 2001, 2007 and 2011. Participation rates in the follow-up studies have been satisfactory, ranging from 2060 to 3596, the main reasons for non-participation being lack of interest in the study, the accompanying person being unable to obtain leave from work, child unwilling to participate and fear of clinical examination [ 15 ]. Unexpectedly, many participants lost to follow-up earlier in the study have returned. For the present study, the sample was restricted to participants with at least one measurement of both LTPA and FVC. In addition, since the physical activity data on participants under age nine was parent-reported, we restricted the self-reported LTPA data to ages nine to 48 years (n=3536; 51% females). The representativeness of the study population has been studied by comparing the baseline (1980) characteristics between the sample of the year 2001 and those lost to follow-up [ 15 ]. The results showed that participants were older and more often females than non-participants. However, no significant differences in physical activity, body mass index (BMI) or parental years of education were observed between participants and non-participants. A detailed description of the YFS and the study protocol has been published earlier [15]. 2.2. Measurements Leisure-time physical activity. LTPA was assessed eight times between the years 1980 and 2011 (ages 9 to 48) through a self-administered questionnaire. In the years 1980 to 1989, the questionnaire items concerned the frequency and intensity of LTPA, participation in sports-club training, participation in Int. J. Environ. Res. Public Health 2019,16, 4437 3 of 17 sports competitions, and habitual way of spending leisure-time. In the years 1992 to 2011, the LTPA questionnaire items concerned the frequency and intensity of LTPA, frequency of vigorous LTPA, hours spent on vigorous LTPA, the average duration of an LTPA session, and participation in organized LTPA. All items were first recoded (1 =inactivity or very low activity; 2 =moderately intensive or frequent activity; 3 =frequent or vigorous activity) and then summed to create a physical activity index ranging from 5 to 15 [ 16 , 17 ]. The creation of the original index has been described previously elsewhere [ 16 ]. An index value under seven describes inactivity and value of eleven high activity. Six is scored when, for example, participants do not usually experience breathlessness and sweating during physical activity, do not engage in rigorous LTPA, engage in organized physical activity occasionally, and engage in LTPA sessions lasting for 20-40 minutes. Eleven is scored when, for example, participants experience a lot of breathlessness and sweating during physical activity, engage in rigorous LTPA several times and 2-6 hours per week, participate in organized LTPA at least once a week and engage in LTPA sessions lasting for 40-60 minutes. The criterion validity of the childhood (years 1980-89) and adulthood (year 1992 onward) LTPA index has been tested by studying its correlation with indicators of exercise capacity (hypothetical maximal workload sustainable for 6 minutes) in a subsample (n=102) of YFS participants. The correlations were significant in childhood (girls: r=0.39; boys: r=0.33) and adulthood (women: r=0.49; men: r=0.53) [ 16 ]. In addition, the LTPA index correlated significantly with 7-day pedometer data (r=0.24 for total steps; r=0.31 for aerobic steps) [ 17 ], and also with accelerometer data (r=0.26-0.45) [18]. Fruit and vegetable consumption. FVC was assessed six times using two different self-administered questionnaires. The first questionnaire was used in 1980, 1983, 1986, and 1989 (at ages 9 to 27 years) and consisted of two items: Frequencies of fruit and fruit juice consumption and vegetable consumption separately during the past month. Participants selected one of six response categories: (1) not at all or hardly ever; (2) once or twice a month; (3) once a week; (4) a few times a week; (5) nearly every day; and (6) every day. An FVC index was created by summing the values from the two items. The index ranged from 2 to 12, with high scores indicating high FVC. In 2007 and 2011 (ages 30 to 48 years), the above questionnaire was replaced with a more comprehensive food frequency questionnaire (FFQ) comprising 131 items on different foods and drinks. The FFQ was developed and validated by the Finnish National Institute for Health and Welfare [ 19 ]. Participants reported their monthly, weekly or daily consumption of different food and drink items during the past year. The reported intakes (frequency and portion sizes) of fruits (including fresh and canned fruits and berries) and vegetables (including fresh and canned vegetables, root vegetables, mushrooms, cabbage, pulses and edible bulbs) were first converted into grams per day and then summed to form a variable indicating total daily FVC. Those in the 0.5% of the sample with the highest or lowest scores of the total daily FVC for the years 2007 and 2011 (n=38) were excluded in order to remove the extreme underand over-reporters from the further analyses. Covariates. Participants’ BMI, total energy intake, education level, and their own and their mothers’ number of years of education were used as covariates. Height was measured with a wall-mounted stadiometer and weight with a digital scale. BMI was calculated as the ratio of weight to the square of height (kg/m 2 ). Total energy intake was assessed based on the FFQ in 2007 and 2011 [ 19 ]. To exclude underand over-reporters of total energy intake, all those in the 0.5% of the sample with the highest or lowest scores of total energy intake data for the years 2007 and 2011 (n=38) were excluded from the further analyses. Participants were asked to state education level (primary school, vocational school, high school, and university or equivalent) and their own and their mothers’ number of years of education. 2.3. Statistical Analysis Descriptive statistics were calculated by using IBM SPSS Statistics for Windows, version 24.0 (IBM Corp. Armonk, NY, USA), and expressed as means and standard deviations. Differences in Int. J. Environ. Res. Public Health 2019,16, 4437 4 of 17 the study variables between males and females were tested by using an independent-samples t-test. Further modelling was performed by using Mplus, version 8.0 [20]. Distinct LTPA trajectory classes from childhood to adulthood have been identified from the YFS data in a previous study for males and females separately by using latent profile analysis, which is a type of finite mixture modelling [ 21 ]. The modelling was performed again for the present study as the composition of the sample had changed slightly owing to the requirement that each participant had at least one measurement of LTPA and FVC. The statistical modelling of the LTPA trajectories has been described in a previous paper [ 21 ] and is presented in Supplement 1. All analyses were performed separately for males and females owing to differences in LTPA previously observed between the sexes [ 21 ]. Missing data were assumed to be missing at random. Model parameters were estimated by using the full information maximum likelihood method with robust standard errors, thus, enabling the use of all the available data. After identifying the LTPA trajectories, the mean differences in FVC from age 9 to 48 years across the LTPA trajectory classes were studied utilizing the Bolck-Croon-Hagenaars (BCH) approach [ 22 – 24 ]. In the BCH approach, the model estimates for the latent classes (here LTPA) are not affected by the auxiliary variable (FVC), thereby avoiding the class membership changes [ 24 ]. First, the BCH weights from the latent profile analysis run with the optimal number of LTPA trajectory classes were saved. BCH weights are group-specific weights computed for each participant during the latent profile model estimation. In the second run, the BCH weights were used as training data, and a multiple group regression model was estimated. FVC was regressed on age-specific covariates within the distinct LTPA trajectory classes, and differences in the regression intercepts (i.e., adjusted means of FVC) across the trajectory classes were studied. When, for example owing to small class size, an error in the computation was reported in Mplus, the residual variances were fixed to a value with the lowest Akaike’s Information Criterion, which indicates the best model fit. The models were adjusted for the participant’s BMI at all ages, mother’s years of education (at ages 9, 12, 15, 18, and 21), the participant’s education level (at ages 24 and 27), the participant’s years of education (at ages 30, 33, 36, 39, 42, 45, and 48), and total energy intake (at ages 30, 33, 36, 39, 42, 45, and 48). Standardized values of the covariates were used when adjusting the models. 2.4. Quality Assessment To enhance reporting quality, the study was conducted according to the Strengthening the Reporting of Observational Studies in Epidemiology–nutritional epidemiology (STROBE-nut) checklist [ 25 ] (Supplement 2). The Guidelines for Reporting on Latent Trajectory Studies (GRoLTS) checklist [26] was applied to ensure the quality of the trajectory modelling (Supplement 3). 2.5. Availability of Data and Materials The datasets analysed during this study are not publicly available for ethical and legal reasons, butareavailablefromthePublicationCommitteeoftheYFSonreasonablerequest.Formoreinformation on dataset access, please contact Professor Olli Raitakari, Project Director of the YFS, University of Turku, Finland, [email protected]. 2.6. Ethics Approval and Consent to Participate All subjects gave their informed consent for inclusion before they participated in the study [ 15 ]. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the ethics committees of each of the five participating universities (ETMK:68/1801/2017). Int. J. Environ. Res. Public Health 2019,16, 4437 5 of 17 3. Results 3.1. Participants and Their Characteristics The sample size of the present study was 3536 (51% females). All eight LTPA measurements had been completed by 508 participants (14.4%), seven by 579 (16.4%), six by 666 (18.8%), five by 572 (16.2%), four by 453 (12.8%), three by 378 (10.7%), two by 241 (6.8%) and one by 139 (3.9%). For FVC, the corresponding figures were 752 (21.3%), 944 (26.7%), 828 (23.4%), 534 (15.1%), 321 (9.1%), 156 (4.4%) and one (0.03%). Table 1shows descriptive characteristics and missing data for each study variable at participants’ youngest age, nine years of age, and at 45 years. Since the sample size was small at the participant’s oldest age, 48 years of age, the age of 45 was chosen to describe the descriptive characteristics of the participants at the end of the study. Table 1. Descriptive statistics of the study sample at age 9 and 45. Descriptive Variable at Age 9 and 45 Males Females pa Mean (sd) n(Missing) Mean (sd) n(Missing) At age 9: LTPA (index, range 5–15) 9.9 (1.6) 798 (100) 8.9 (1.4) 807 (101) <0.001 FVC frequency (index, range 2–12) 10.2 (1.7) 778 (120) 10.4 (1.5) 803 (105) 0.010 BMI (kg/m2)16.7 (2.3) 814 (84) 16.7 (2.3) 831 (77) 0.612 Mothers’ education (years) 10.9 (3.3) 798 (100) 10.7 (3.2) 804 (104) 0.315 At age 45: LTPA (index value, range 5–15) 8.8 (1.9) 305 (239) 8.9 (1.7) 380 (215) 0.405 FVC (grams per day) 390 (206) 297 (247) 494 (214) 367 (228) <0.001 Total energy intake (kcal/day) 2629 (799) 297 (247) 2139 (604) 367 (228) <0.001 BMI (kg/m2)27.3 (4.2) 318 (226) 26.3 (5.6) 391 (204) 0.004 Education (years) 14.5 (3.6) 315 (229) 15.4 (3.4) 388 (207) 0.001 sd standard deviation; LTPA Leisure-time physical activity; FVC Fruit and vegetable consumption; BMI Body mass index. ap-value for sex difference (t-test). 3.2. Fruit and Vegetable Consumption of Males Across the Leisure-Time Physical Activity Trajectories Four LTPA trajectories were identified for males: Persistently low-active (40.9%), decreasingly active (15.7%), increasingly active (31.1%), and persistently active (12.3%) (Figure 1A). More detailed description of the LTPA trajectories is presented in Supplement 1 and the selection of the final number of classes is presented in Table 2. The lowest level of FVC was found for males on the persistently low-active trajectory at nearly all ages (Figure 2A,B). Compared to those following the persistently active trajectory, the FVC of the persistently low-active males was significantly lower (p=[0.000, 0.015]) at half the ages studied (12, 15, 18, 21, 24, 36, and 39 years) (Table 3). In general, the level of FVC declined during adolescence (age 12-18 years) across all the LTPA trajectories (Figure 2A), but increased among the increasingly active, as well as low active (age 33–42) and decreasingly active males (age 42–48) in adulthood (Figure 2B). Int. J. Environ. Res. Public Health 2019,16, 4437 6 of 17 Int. J. Environ. Res. Public Health 2019, 16, x 6 of 16 (A) Males (n = 1727) (B) Females (n = 1809) Figure 1. Leisure-time physical activity trajectories for males (n = 1727) (A) and females (n = 1809) (B). Figure 1. Leisure-time physical activity trajectories for males (n=1727) ( A ) and females (n=1809) ( B ). Int. J. Environ. Res. Public Health 2019,16, 4437 7 of 17 Table 2. Latent profile analyses for leisure-time physical activity in males (n=1727) and females (n=1809). AIC BIC ABIC VLMR LMR BLRT Entropy Class Sizes (%) aAvePP The Number of Random Start Values and Final Iterations Males 1 32128 32281 32192 - - - - - - 500, 20 2 30662 30897 30760 <0.001 <0.001 <0.001 0.78 73.7%, 26.3% 0.95, 0.90 500, 20 3 30342 30658 30474 0.01 0.011 <0.001 0.63 43.4%, 40.2%. 16.4% 0.83, 0.78, 0.89 500, 20 4 30139 30537 30305 <0.001 0.001 <0.001 0.64 40.9%, 31.1%, 15.7%, 12.3% 0.80, 0.75, 0.72, 0.85 1000, 40 5 30082 30562 30282 0.198 0.202 <0.001 0.59 32.2%, 23.6%, 17.2%, 15.7%, 11.3% 0.65, 0.75, 0.72, 0.73, 0.85 1000, 40 6 30021 30583 30256 0.494 0.498 <0.001 0.58 31.3%, 16.9%, 16.4%, 16.2%, 9.9%, 9.2% 0.64, 0.72, 0.68, 0.69, 0.81, 0.74 1000, 40 Females 1 34757 34911 34822 - - - - - - 500, 20 2 33634 33871 33734 <0.001 <0.001 <0.001 0.83 84.0%, 16.0% 0.96, 0.89 500, 20 3 33268 33587 33403 <0.001 <0.001 <0.001 0.64 57.5%, 30.6%, 11.9% 0.82, 0.81, 0.89 500, 20 4 33147 33548 33316 0.615 0.617 <0.001 0.66 49.8%, 33.5%, 12.5%, 4.2% 0.79, 0.80, 0.77, 0.82 1000, 40 5 33018 33502 33223 0.183 0.184 <0.001 0.66 52.4%, 16.8%, 15.1%, 12.3%, 3.4% 0.77, 0.79, 0.76, 0.77, 0.87 2000, 80 6 32963 33530 33202 0.322 0.324 <0.001 0.60 41.7%, 15.7%, 15.0%, 14.1%, 10.0%, 3.5% 0.69, 0.76, 0.67, 0.72, 0.75, 0.85 2000, 80 a Final class proportions for the tested latent class models based on estimated posterior probabilities. AIC Akaike’s information criterion; BIC Bayesian information criterion; BIC sample-size adjusted Bayesian information criterion; VLMR Vuong-Lo-Mendell-Rubin likelihood ratio test; LMR Lo-Mendell-Rubin adjusted LRT test; BLRT Parametric bootstrapped likelihood ratio test. AvePP Average posterior probabilities for most likely latent class membership. The class solution considered optimal is presented in bold. Int. J. Environ. Res. Public Health 2019,16, 4437 8 of 17 Int. J. Environ. Res. Public Health 2019, 16, x 10 of 17 (A) In childhood and adolescence (B) In adulthood Figure 2. Mean fruit and vegetable consumption in different leisure-time physical activity trajectories among males from ages 9 to 27 years (A) and 30 to 48 years (B). 3.3. Fruit and Vegetable Consumption of Females Across the Leisure-Time Physical Activity Trajectories Among females, five LTPA trajectories were identified: Persistently inactive (16.8%), persistently low-active (52.4%), decreasingly active (12.3%), increasingly active (15.1%), and persistently active (3.4%) (Figure 1B). For a more detailed description of the LTPA trajectories and the selection of the final number of classes, see Supplement 1 and Table 2. The mean values of FVC were the lowest among the females on the persistently inactive trajectory at almost all ages (Figure 3A and 3B). These values were significantly lower than among those on the persistently active trajectory at ages 9, 15, 21, 39 and 45 (p = [0.007, 0.049]) (Table 3). (A) In childhood and adolescence (B) In adulthood Figure 3. Mean fruit and vegetable consumption in different leisure-time physical activity trajectories among females from ages 9 to 27 years (A) and 30 to 48 years (B). Figure 3 shows that the FVC was highest among the females on the increasingly active trajectory in middle age (Figure 3B). From age 15 onward, FVC was higher among the females on the increasingly active trajectory when compared to the inactive trajectory (p = [0.000, 0.005]) (Table 3). FVC was also higher among the increasingly active than low-active females at ages 15, 21, 24, 27, 42, 45 and 48 (p = [0.002, 0.048]). Females following the decreasingly active trajectory had higher FVC than their inactive peers at ages 9, 15, 18, 21, 27, 36, and 39 (p = [0.000, 0.041]) or their low-active peers at ages 15, 18, and 27 (p = [0.000, 0.044]). However, these differences were no longer evident in late middle age. FVC was higher among females identified to the increasingly active trajectory than Figure 2. Mean fruit and vegetable consumption in different leisure-time physical activity trajectories among males from ages 9 to 27 years (A) and 30 to 48 years (B). During childhood and adolescence (at ages 9–18), the highest mean values of FVC were found for males on the persistently or decreasingly active trajectory (Figure 2A). However, from age 33 onward, the highest mean values of FVC was observed in the increasingly active males (Figure 2B). When compared to the increasingly active males, those who were persistently active had significantly higher FVC at ages 15 and 18 (p=[0.003, 0.032]), after which no significant difference in FVC was observed between these two trajectories (Table 3). The participants on the increasingly active and persistently low-active trajectories showed similar levels of FVC in childhood and adolescence (at ages 9–18). However, the two trajectories showed a significant difference in FVC (p=[0.000, 0.020]) in favor of the increasingly active trajectory in adulthood (ages 21, 24, 36, 39, and 45) (Table 3). Concurrently, males on the decreasingly active trajectory showed higher FVC than those on the low-active trajectory up to age 24 (p=[0.006, 0.042]), but no longer in middle age (Table 3). Int. J. Environ. Res. Public Health 2019,16, 4437 9 of 17 Table 3. Mean fruit and vegetable consumption across the leisure-time physical activity trajectories. Mean FVC in LTPA classes in 1980–1989 a(SE) Mean FVC in LTPA classes in 2007–2011 b(SE) Age in years 9 12 15 18 21 24 27 30 33 36 39 42 45 48 Sample size, males 777 1022 957 837 487 491 325 140 244 240 287 276 296 116 Sample size, females 802 1038 1068 987 614 618 443 156 323 341 362 375 365 169 Males: 1 Persistently active 10.3 (0.2) 10.4 (0.2) 10.2 (0.2) 9.9 (0.2) 9.7 (0.3) 9.5 (0.3) 9.6 (0.6) 452.8 (55.7) 437.9 (92.6) 467.5 (37.5) 413.1 (35.0) 403.1 (55.8) 442.7 (67.6) 405.3 (101.4) 2 Increasingly active 10.1 (0.1) 10.0 (0.2) 9.5 (0.2) 9.3 (0.2) 10.1 (0.3) 9.3 (0.3) 9.5 (0.6) 315.1 (48.9) 488.7 (87.3) 483.1 (34.3) 504.1 (47.8) 424.8 (33.1) 475.1 (33.5) 565.8 (109.3) 3 Decreasingly active 10.6 (0.2) 10.4 (0.2) 9.8 (0.2) 9.4 (0.3) 9.6 (0.3) 9.3 (0.3) 9.4 (0.8) 396.4 (59.2) 329.2 (68.9) 343.3 (54.3) 374.4 (47.4) 347.9 (74.0) 430.0 (75.7) 472.5 (99.5) 4 Persistently low-active 10.1 (0.1) 9.9 (0.1) 9.1 (0.1) 8.7 (0.1) 8.7 (0.2) 8.4 (0.2) 8.4 (0.3) 326.1 (57.9) 234.4 (115.5) 259.3 (41.4) c295.8 (22.5) 368.6 (23.4) 352.4 (22.0) 310.2 (61.3) Statistically significant mean differences in FVC between LTPA classes 1–4 (Z-score test) 2<1 ** 2 <1 * 4 <3 * 4 <3 * 3 <2 * 4<3 * 4 <3 ** 4 <3 * 4 <2 *** 4 <2 ** 4 <2 *** 4 <2 ** 4<1 * 4 <1 *** 4 <1 *** 4 <1 ** 4 <1 ** 4 <1 ** 4 <1 ** 4 <2 ** Females: 1 Persistently active 11.0 (0.3) 10.7 (0.3) 10.4 (0.3) 9.9 (0.4) 11.6 c (0.8) 9.5 (0.6) 9.6 (0.7) 649.2 (156.5) 537.2 (96.5) 514.2 (46.4) 556.3 (83.1) 584.8 (103.8) 666.5 (75.7) 490.4 c (58.0) 2 Increasingly active 10.4 (0.2) 10.5 (0.2) 10.4 (0.2) 10.2 (0.3) 10.6 (0.4) 10.8 (0.3) 10.8 c (0.3) 672.8 (90.6) 686.4 (67.2) 617.7 (52.5) 628.2 (67.1) 680.1 (59.1) 771.1 (73.0) 787.2 c (75.0) 3 Decreasingly active 10.7 (0.2) 10.4 (0.1) 10.6 (0.2) 10.3 (0.2) 10.2 (0.5) 10.1 (0.3) 10.7 (0.3) 495.5 (69.7) 548.8 (55.7) 603.7 (43.6) 557.7 (53.3) 499.5 (65.5) 408.3 (82.3) 478.3 (111.0) 4 Persistently low-active 10.4 (0.1) 10.1 (0.1) 9.6 (0.1) 9.7 (0.1) 9.6 (0.2) 9.8 (0.1) 10.0 (0.2) 494.4 (47.0) 441.3 (68.5) 485.3 (37.4) 513.5 (37.2) 476.8 (30.1) 517.8 (24.4) 515.4 (39.9) 5 Persistently inactive 10.2 (0.2) 10.0 (0.2) 9.3 (0.2) 9.3 (0.2) 8.9 (0.3) 9.6 (0.3) 9.4 (0.4) 378.1 (233.4) 383.3 (73.8) 392.3 (50.7) 377.4 (32.1) 464.4 (41.4) 438.8 (40.7) 415.4 (74.1) Statistically significant mean differences in FVC between LTPA classes 1–5 (Z-score test) 4<1 * 4<2 ** 4 <1 * 3 <1 * 4<3 *** 4 <2 * 4 <2 * 5 <1 * 3 <2 ** 1 <2 ** 5<1 ** 4 <3 * 5 <1 * 4 <3 * 5 <4 * 3 <2 * 4 <2 ** 3 <2 * 5<1 * 5 <2 *** 5 <2 ** 5 <2 *** 4 <2 ** 5 <2 ** 3 <2 * 5 <2 ** 5 <2 ** 4 <2 ** 5 <1 ** 4 <2 ** 5<3 * 5 <3 *** 5 <3 ** 5 <3 * 5 <2 ** 5 <3 ** 5 <2 ** 5 <3 ** 5 <3 ** 5 <2 ** 5 <2 *** 5 <2 ** FVC Fruit and vegetable consumption; LTPA Leisure-time physical activity. * p<0.05; ** p<0.01; *** p<0.001. a At ages 9-27 FVC was expressed with index (range: 2-12). The model was adjusted for BMI at ages 9-27, for mothers’ education (years) at ages 9-18, and for participants’ education (level) at ages 21-27. b At ages 30-48 FVC was expressed in grams of fruits and vegetables consumed in a day. The model was adjusted for participants’ education (years), BMI, and total energy intake at ages 30-48. c The residual variances were fixed to a value with lowest Akaike’s Information Criterion. Int. J. Environ. Res. Public Health 2019,16, 4437 16 of 17 42. Roos, E.; Sarlio-Lähteenkorva, S.; Lallukka, T. Having Lunch at a StaffCanteen Is Associated with Recommended Food Habits. Public Health Nutr. 2004,7, 53–61. [CrossRef] [PubMed] 43. Prättälä, R. Dietary Changes in Finland—Success Stories and Future Challenges. Appetite 2003 ,41, 245–249. [CrossRef] [PubMed] 44. Lahti-Koski, M. Report on Nutrition 1998 [Ravitsemuskertomus 1998]. National Public Health Institute: Helsinki, Finland, 1999. Available online: https://www.julkari.fi/bitstream/handle/10024/78259/1999b02.pdf? sequence=1&isAllowed=y(accessed on 7 November 2019). 45. Fogelholm, M.; Hakala, P.; Kiuru, S.; Kurppa, S.; Kuusipalo, H.; Laitinen, J.; Marniemi, A.; Misikangas, M.; Roos, E.; Sarlio-Lähteenkorva, S.; et al. Health from Food: Finnsih Food Recommendations. Available online: https://www.ruokavirasto.fi/globalassets/teemat/terveytta-edistava-ruokavalio/kuluttaja-jaammattilaismateriaali/julkaisut/ravitsemussuositukset_2014_fi_web_versio_5.pdf (accessed on 5 July 2019). 46. Finnish Food Recommendations—Balance of Nutrition and Physical Activity [Suomalaiset ravitsemussuositukset —Ravinto ja liikunta tasapainoon]. Finnish Food Authority: Helsinki, Finland, 2005. Available online: https://asiakas.kotisivukone.com/files/luontoemo.kotisivukone.com/tiedostot/ suomalaisetravitsemussuositukset.pdf (accessed on 7 November 2019). 47. Kanerva, N.; Kaartinen, N.E.; Ovaskainen, M.L.; Konttinen, H.; Kontto, J.; Männistö, S. A Diet Following Finnish Nutrition Recommendations Does Not Contribute to the Current Epidemic of Obesity. Public Health Nutr. 2013,16, 786–794. [CrossRef] [PubMed] 48. O’Donoghue, G.; Kennedy, A.; Puggina, A.; Aleksovska, K.; Buck, C.; Burns, C.; Cardon, G.; Carlin, A.; Ciarapica, D.; Colotto, M.; et al. Socio-Economic Determinants of Physical Activity across the Life Course: A DEterminants of DIet and Physical ACtivity (DEDIPAC) Umbrella Literature Review. PLoS ONE 2018 ,13, 1–24. [CrossRef] 49. Condello, G.; Puggina, A.; Aleksovska, K.; Buck, C.; Burns, C.; Cardon, G.; Carlin, A.; Simon, C.; Ciarapica, D.; Coppinger, T.; et al. Behavioral Determinants of Physical Activity across the Life Course: A DEterminants of DIet and Physical ACtivity (DEDIPAC) Umbrella Systematic Literature Review. Int. J. Behav. Nutr. Phys. Act. 2017,14, 58. [CrossRef] 50. Granner, M.L.; Evans, A.E. Variables Associated With Fruit and Vegetable Intake in Adolescents. Am. J. Heal. Behav. 2011,35, 591–602. [CrossRef] 51. Hall, J.N.; Moore, S.; Harper, S.B.; Lynch, J.W. Global Variability in Fruit and Vegetable Consumption. Am. J. Prev. Med. 2009,36, 402–409.e5. [CrossRef] 52. Hirvensalo, M.; Lintunen, T. Life-Course Perspective for Physical Activity and Sports Participation. Eur. Rev. Aging Phys. Act. 2011,8, 13–22. [CrossRef] 53. Dutton, G.R.; Napolitano, M.A.; Whiteley, J.A.; Marcus, B.H. Is Physical Activity a Gateway Behavior for Diet? Findings from a Physical Activity Trial. Prev. Med. 2008,46, 216–221. [CrossRef] [PubMed] 54. Wilcox, S.; King, A.C.; Castro, C.; Bortz, W. Do Changes in Physical Activity Lead to Dietary Changes in Middle and Old Age? Am. J. Prev. Med. 2000,18, 276–283. [CrossRef] 55. Prochaska, J.J.; Sallis, J.F. A Randomized Controlled Trial of Single Versus Multiple Health Behavior Change: Promoting Physical Activity and Nutrition among Adolescents. Heal. Psychol. 2004 ,23, 314–318. [CrossRef] [PubMed] 56. Jacobs, D.R.; Tapsell, L.C. Food, Not Nutrients, Is the Fundamental Unit in Nutrition. Nutr. Rev. 2007 ,65, 439–450. [CrossRef] 57. Slavin, J.; Lloyd, B. Health Benefits of Fruits and Vegetables. Adv. Nutr. 2012,3, 506–516. [CrossRef] 58. Kurtze, N.; Rangul, V.; Hustvedt, B.E.; Flanders, W.D. Reliability and Validity of Self-Reported Physical Activity in the Nord-Trøndelag Health Study—HUNT 1. Scand. J. Public Health 2008,36, 52–61. [CrossRef] 59. Olds, T.S.; Gomersall, S.R.; Olds, S.T.; Ridley, K. A Source of Systematic Bias in Self-Reported Physical Activity: The Cutpoint Bias Hypothesis. J. Sci. Med. Sport 2019,22, 924–928. [CrossRef] 60. Downs, A.; Van Hoomissen, J.; Lafrenz, A.; Julka, D.L. Accelerometer-Measured versus Self-Reported Physical Activity in College Students: Implications for Research and Practice. J. Am. Coll. Health 2014 ,62, 204–212. [CrossRef] 61. Männistö, S.; Virtanen, M.; Mikkonen, T.; Pietinen, P. Reproducibility and Validity of a Food Frequency Questionnaire in a Case-Control Study on Breast Cancer. J. Clin. Epidemiol. 1996,49, 401–409. [CrossRef] Int. J. Environ. Res. Public Health 2019,16, 4437 17 of 17 62. Thompson, F.E.; Kipnis, V.; Subar, A.F.; Krebs-Smith, S.M.; Kahle, L.L.; Midthune, D.; Potischman, N.; Schatzkin, A. Evaluation of 2 Brief Instruments and a Food-Frequency Questionnaire to Estimate Daily Number of Servings of Fruit and Vegetables. Am. J. Clin. Nutr. 2000,71, 1503–1510. [CrossRef] 63. Warren, J.R.; Luo, L.; Halpern-Manners, A.; Raymo, J.M.; Palloni, A. Do Different Methods for Modeling Age-Graded Trajectories Yield Consistent and Valid Results? Am. J. Sociol. 2017 ,120, 1809–1856. [CrossRef] [PubMed] 64. Fleig, L.; Küper, C.; Lippke, S.; Schwarzer, R.; Wiedemann, A.U. Cross-Behavior Associations and Multiple Health Behavior Change: A Longitudinal Study on Physical Activity and Fruit and Vegetable Intake. J. Health Psychol. 2015,20, 525–534. [CrossRef] [PubMed] © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).