Association between parents’ physical activity and mode of commuting with their offsprings

Background. Some studies have reported a positive association between parents and their offspring’s physical activity (PA), but few have examined the difference in these associations concerning both genders. The objective of this study was to establish the association between moderate-vigorous physical activity (MVPA) and mode of commuting (MC) of the parents with their offspring’s MC, by gender and age group. Methods. This cross-sectional study included 686 parents (mothers: 52.8%) and their offsprings (33.8% girls). Each participant completed a questionnaire on PA and MC. Chi-square test, Odds Ratio for categorical variables and lineal regressions for continuous variables were used to examine the associations between the parents and their offspring. Results. An inverse association was found between fathers-children in the weekend MVPA in children and between mothers-adolescents in out-of-school and weekend MVPA in adolescents, specically, an inverse association was found in MVPA between mothers-girls and 2) the different parents’ MC to work were positively associated with the MC to school in children and adolescents except for the association AC parents-adolescents and specically, the AC was mainly associated between mothers and girls and boys. positive ripple effect in behaviours.

Moreover, the effect of the parent gender on children's PA has been previously examined. For instance, a positive association has been found between mothers' sport participation and children's leisure-time PA, and it remained signi cant only in girls (Arlinghaus and Johnston, 2017). Another Brazilian study observed a greater positive association between the mother's PA with adolescent boys and girls, but there was no association for fathers (Christofaro et al., 2018). Furthermore, the PA of both parents presented a greater association with girls than with boys. In contrast, a study demonstrated that only the mother's MVPA was associated with their children's MVPA (Tanaka et al., 2018). Another study showed a signi cant positive association between fathers and boys only (Schoeppe et al., 2016). Despite that fact, studies have focused on the mother-child relationship, and relatively little attention has been paid to the role of fathers in their children's PA (Neshteruk et al., 2017;Davison et al., 2016). Mothers play a greater role in planning and organizing children's PA, while fathers are more likely to model children's PA (Lloyd et al., 2014). Until now, the evidence on PA gender-speci c parental in uence (e.g., mother-daughter, mother-son, father-daughter, father-son) is still inconsistent.
It has been established that active commuting to school (ACS) or to work (walking or cycling) is an opportunity to increase PA levels (Martin, Kelly, Boyle, Corlett, and Reilly, 2016; Prince, Butler, Rao and Thompson, 2019). However, around 50% of youths and 75% of adults passively commute to school (Aubert et al., 2018) and to work (Te Velde et al., 2017), respectively. Multiple factors in uence whether schoolchildren actively travel to/from school (Rahman et al., 2020). The educational level and body weight (Te Velde et al., 2017), the socioeconomic level (D'Haese et al., 2014) and the unemployment of the parents (Rodríguez- López et al., 2013) are factors that affect children's AC and have been studied. But few studies have linked the parent's AC with their offspring's AC and even less take into account gender. A Spanish study observed a positive association between parents' AC and children's AC (Rodríguez-López et al., 2013). Indeed, the number of step/day of the parents has been positively associated with the number of step/day of the offsprings (Sigmundová et al., 2014), mainly associated between mothers and daughters (Craig et al., 2013). Speci cally, it has been reported that for every 1000 step/day that parents increase, their offsprings increase 260 steps/day (Stearns et al., 2016).
Therefore, further quality studies analysing the in uence that the fathers' or mothers' AC has on boys' or girls' ACS, as well as taking in account the age (i.e., children or adolescents) are needed. This would allow targeting interventions to more speci c sectors and groups in order to achieve successful results.
According to the previous information, the main objectives of the current study were 1) to establish the association between parents' and their offsprings PA and 2) to establish the association between parents and their offsprings mode of commuting, determining how the parental association is different by gender and by age groups.

Study design and participants
Data were obtained from the study "Cycling and Walk to School" (PACO, for its Spanish acronym), a cross-sectional study focused on promoting PA and, particularly, active commuting to and from school. This study was carried out in Granada (Spain) and Valparaíso (Chile) participating a total of 2,526 children and adolescents and 1,959 of their parents. Twenty schools were invited to participate in the study as a non-randomized sample. Of the total offsprings, 1,807 participants could not be paired with parents' data and 34 did not report their gender and were excluded (72,8% of the total sample). A total of 572 parents could not be paired with children's data and 703 without gender data were excluded (65% of the total sample). Finally, a total of 686 paired parents (52.8% mothers) and their respective offsprings (33.7% girls) were considered. They belonged to fteen schools of Granada, Spain (n = 494) and ve schools of Valparaíso in Chile (n = 192). The age (mean ± standard deviation) of each group was: parents 43.4 ± 6.5 years old), children 9.7 ± 1.7 years old and adolescents 14.0 ± 1.7 years old.

Physical activity
The questionnaire to determine the weekly school and extracurricular PA in children and adolescents was the Youth Activity Pro le (YAP). The YAP provides a simple and low-cost method that have already been calibrated and validated to accurately estimate children's MVPA and sedentary behaviours at the group level (Saint-Maurice & Welk, 2015). The self-reported YAP questionnaire comprises three items about participation in different types of activities and sports in the last seven days. The items were "PA in school" (physical education, recess and active commuting to school), "out-of-school PA" (before and after school and PA at the weekend) and "sedentary time" (not considered in current study). Each item was scored on a scale from 1 (low PA) to 5 (high PA) and the average score denotes the YAP score (1)(2)(3)(4)(5). The questions were categorized into three areas: PA at school, out-of-school PA and weekend PA.
Then, PA in children and adolescents was established by transforming the YAP score to minutes/day in MVPA using the Fairclough equations (Fairclough et al., 2019). We obtained the min/day in MVPA at school, out-of-school MVPA and weekend MVPA, separately for children and adolescents and for boys and girls. The cut-off point to be "physically active" was > 60 minutes/day MVPA (Tremblay et al., 2016) classi ed as meeting MVPA recommendations. Those children who did not comply were considered as "physically inactive".
The International Physical Activity Questionnaire (IPAQ, short version) was used to assess the parents' MVPA. This questionnaire has been validated in 12 countries, (Craig et al., 2003;Roman-Viñas et al., 2010) showing acceptable psychometric properties to measure the MVPA levels in one week. The IPAQ allows determining the PA in min/week. Also, this instrument determines intensity categories as sedentary time, light PA, moderate PA and vigorous PA. According to the MVPA international recommendations, parents were classi ed as physically active when they completed ≥ 150 minutes/week -meeting the recommendations-and physically inactive when they did not reach 150 minutes/week -not meeting the recommendations-(WHO, 2010).

Mode of commuting
The instruments used to assess the mode of commuting were a student (for children and adolescents) and a family (for parents) questionnaire from the PACO Study that have been developed at the University of Granada, Spain, by a group of experts in this topic (https://pro th.ugr.es/paco). The questions about the mode of commuting (MC) to school derive from an exhaustive review of previous studies of the scienti c literature on AC (Herrador-Colmenero, Pérez-García, Ruiz, and Chillón, 2014) and have been validated in Spanish population  and Chilean population (Escobar-Gómez et al., 2020). In children and adolescents, the questions included in the questionnaire were: (1) How do you usually get to school? and (2) How do you usually get home from school?. The possible answers were: walking, cycling, car, motorcycle, school bus, public bus, metro/train and other (the mode description was required). The nal variable to analyse was usual active mode of commuting to school.
The questions about parent's MC to work have gone through an exhaustive reliability process (Herrador-Colmenero et al., 2015). In parents, the questions were: (1) How do you usually get to work? and (2) How do you usually get home from work?. The possible answers were: walking, cycling, car, motorcycle, public bus, metro/train or other (the mode description was required).
MC for parents and offsprings was categorized as "active" (walking and cycling) and "passive" (car, motorcycle, bus, metro/train). Also, the "passive" commuting was divided into private mode of commuting (car, motorcycle) and public mode of commuting (public bus, metro, train). In addition, the nal variable of usual active commuting included one or two trips to school (to, from or both).

Procedures
The questionnaires (paper-and-pencil) were administered at participating schools between 2015 and 2018. The children and adolescents' questionnaires were implemented by the researcher staff during the school hours within the physical education lessons and along approximately 30 minutes. Both research team and schoolteachers presented this questionnaire for clari cation purposes. The family's questionnaire was completed once by parents and it was delivered to children and completed at home by parents.
Parents signed an informed consent that described the objectives and characteristics of this study and allowing their offsprings to participate, in accordance with the Declaration of Helsinki. This study was reviewed and accepted by both the Ethical Committee of the University of Granada, Spain (No.162/CEIH/2016) and the Ethical Committee of the Ponti cia Universidad Católica de Valparaíso, Chile (CCF02052017).

Statistical analysis
Descriptive statistics were calculated for study variables, mean (M) and standard deviation (± SD) for continuous variables and frequency (%) for categorical variables.
The comparison of categorical variables according to gender in each group (parents (mothers and fathers), children (girls and boys), adolescents (girls and boys)) was examined using the Chi-Square test.
To establish the association between the MVPA in parents and offsprings, several lineal regressions through a standardized coe cient (ß) were used for the whole sample, for boys and for girls separately. The min/day of MVPA in children and the dichotomic variable of meeting the recommendations for parents were included. Offsprings' MVPA was established as the dependent variable. For the MC association several binary logistic regressions were performed. Offspring's MC to school was established as the dependent variable. The references for MC were a) "passive" for active commuting; b) "passive + private" for public commuting; c) "passive + public" for private commuting. The associations were adjusted for parents' educational level and age of the offsprings. SPSS® v21 (IBM, New York, NY, USA) was used for all the analyses. Also, a p < 0.05 value was established as statistically signi cant.

Results
Sociodemographic characteristics of the parents and their offspring are presented in Table 1. Parents of children were younger than parents of adolescents (41.9 ± 5.7 vs. 46.7 ± 7.1, respectively; p < 0.001). The maximum educational level reached by mothers of children was higher than by mothers of adolescents (p < 0.05). The monthly income for parents of children was higher than parents of adolescents (p < 0.05). Statistical differences in Chi-square * p < 0.05; ** p < 0.001.
The children and parents' MVPA and MC separated by gender are presented in Table 2. The percentage of physically inactive girls was signi cantly higher than the boys for both, children (p < 0.001) and adolescents (p = 0.003). Parents of children and parents of adolescents were also mainly physically inactive (75% and 84.9%, respectively) in both genders (mother and father).

Discussion
This study aimed to examine the association between the parents with their offspring's MVPA and MC of, determining the differences by gender and age groups. The main ndings of the study were: 1) an inverse association was found between fathers-children in the weekend MVPA in children and between mothers-adolescents in out-of-school and weekend MVPA in adolescents, speci cally, an inverse association was found in MVPA between mothers-girls and 2) the different parents' MC to work were positively associated with the MC to school in children and adolescents except for the association AC parents-adolescents and speci cally, the AC was mainly associated between mothers and girls and boys.

Physical activity
In children, the fathers' MVPA was only positively associated with the girls' weekend MVPA. In addition, an inverse association in MVPA between mothers-boys in children, mothers-offspring's out-of-school MVPA and mothers-girls in the weekend PA in adolescents was found.
Accordingly, a modest and positive association between parent-child PA in a systematic review was found (52% of studies), but no differences were observed in the in uence of mothers or fathers on children's PA (Neshteruk et al., 2017). However, previously a positive association was found for fathers-boys and mothers-girls PA (Gustafson & Rhodes, 2006), but the stronger correlation was observed for father-boy PA and lesser association for mothers-boys and mothers-girls PA (Yao and Rhodes, 2015).
Our results in children showed that fathers' MVPA was positively associated with the weekend MVPA in girls and negatively associated with boys. The positive association between active fathers and active girls can be explained by the role modelling of parents for girls' PA, especially at the weekend. For example, girls whose parents exercised ≥ 3 times weekly can be almost 50% more active than girls with sedentary parents. This especially occurred at the weekend where there is a greater interaction between parents and children (Madsen, McCulloch and Crawford, 2009). In this regard, it has been reported that the fathers were more likely to offer social support to practice PA than mothers, whose main role is taking care of their children during the weekdays because fathers work long hours during the day (Solomon-Moore et al., 2018). Furthermore, the negative association fathers-boys found in the current study was unexpected, but we assume that fathers are not a role modelling but support the practice ( Our results in adolescents showed that the mother' MVPA was negatively associated with out-of-school MVPA in girls and boys and with girls' MVPA during the weekend. The other parents´ associations with boys and gilrls were not signi cant. These associations can be interpreted as the adolescents' PA is more independent of the parents' PA, especially with the mother when the children grow. In this regard, it has been reported in adolescents that, peer, paternal and maternal support decreased with age, whereas independent play increased (Kirby, Levin, and Inchley, 2011). In addition, a qualitative study showed that girl adolescents received more emotional negative support for PA from parents and less tangible support and direct participation from parents in PA with them (Wright et al., 2010). It has been found that it may be that sedentary girls are likely to have sedentary parents, while there may be a weak association between parents and regularly active girls (Bauer et al., 2011).
In general terms, in the current study no positive association parents-offspring's PA was found. In this regard, a recent systematic review showed a weak correlation between the level of parent and child PA and suggests that we should develop a deeper understanding of associations and mechanisms (Petersen et al., 2020). Future studies and interventions should enhance the association between parents and their offspring's PA, especially out-of-school and at the weekend, which seems to be the time where there is a greater association.

Mode of commuting
In the current study, several positive associations between parents and offsprings in the different MC (active, public and private) were found. A greater association was found between mothers' AC to work with girls' and boys' ACS than fathers. The fathers' AC to work was only associated with the ACS in girls. This nding could be related to a greater degree of care or responsibility assumed by mothers with their children (Zahra, Sebire and Jago, 2015), where mothers were mainly responsible for taking them to school. Another study found that greater mothers' work exibility was related to a greater number of children's active trips (Sener, Lee and Sidharthan, 2019). Furthermore, it was possible to associate the PA of the mothers, but not the fathers, with the ACS of their offsprings (Kobel, Wartha and Steinacker, 2019).
In our study, the association between parents' AC to work and offspring' ACS was stronger in children than adolescents. This could be explained by the higher degree of autonomy and independence of adolescents who no longer depend heavily on parents for commuting (Garcia-Cervantes et al., 2016). In addition, it has been studied that the parents perceived lower barriers in adolescents than children ( The general results of the current study showed a positive association between public commuting of parents and girls. But no associations between mothers-boys or fathers-boys were observed. These results could be explained by a possible greater independence of boys compared to girls, which has been previously reported (Barnett et al., 2019). Girls would prefer to be accompanied by their parents, or parents impose more barriers for independent commuting to girls (Larsen et al., 2009). Using public transport is important, since along with walking, they are considered the two best options for independent mobility in children and adolescents (Sarjala, Broberg and Hynynen, 2016 In the current study, a positive association was observed between the parents' private MC and their offspring's MC. This association was even higher than the observed with AC, especially among parents and adolescents. A greater positive association between mothers and boys' adolescents in private commuting was found. Likewise, a strong association between fathers with adolescents' girls in this mode was observed. In this regard, a Spanish study in schoolchildren of 9-12 years (Solana et al., 2018) observed an inverse association between the fact that it is more convenient to drive and the accompaniment of their offspring on foot to school. Parents' convenience of using the car to carry their children has been previously reported as one of the main barriers to ACS (Faulkner et

Active families
Given that the associations in MVPA and MC could also be bi-directional, that is, they can be bene ting parents and offspring (Niermann, Gerards and Kremers, 2018). Accordingly, we propose a classi cation of families according to whether they have reached the MVPA recommendations and according to their use AC (Fig. 1). It is important to note that compliance with the MVPA recommendations has been considered as a behaviour that is above the AC to be considered active (Mora-González et al., 2017). However, the contribution that the AC has as a mean to achieve the recommendations cannot be ignored either.
Therefore, the proposal includes "Active family" as those parents and offsprings that meet the MVPA recommendations and use AC; "Moderately active family", those who use AC, but parents or offspring do not meet MVPA recommendations; "Moderately passive family", the families where neither parents nor offsprings meet the MVPA recommendations, but some perform AC; and "Passive family", that includes the families where parents and offsprings do not meet MVPA recommendations and use passive MC.

Limitations
The main limitation of the study was the cross-sectional design and, therefore, no cause and effect relationship can be established in the associations found. A longitudinal study would be required to determine the direction of the relationship. There was a relevant loss data of sample regarding the initial data collection, because the questionnaires were incomplete. Also, a non-randomised sample was included, therefore it is not possible to generalize results to the population. In addition, a self-reported questionnaire, that has a lower objectivity to determine the PA than devices such as accelerometers, was used.
Finally, in relation to the mode of commuting, we did not have the distance to and from school (continuous value) and it was not used to establish their impact on AC.

Conclusions
The results have demonstrated a low association between the parents-offspring's MVPA. However, a strong association was found between the MC to work of parents and MC to school of their offspring. It was not possible to establish a clear association of the fathers' AC over the mother or vice versa about ACS of their boys and girls offsprings. Increasing the parents' AC to work could mean an improvement not only in their own levels of MVPA, but also in the MVPA levels and ACS of their offspring. Otherwise, attention should be given to the use of passive modes of parents, which are largely associated with the passive modes in children. The results suggest the importance of involving families in the design, implementation and evaluation of interventions in the PA and ACS in children and adolescents.

Declarations
Ethics approval and consent to participate The study in accordance with the Declaration of Helsinki was performed. In addition, was reviewed and accepted by both the Ethical Committee of the University of Granada, Spain (No.162/CEIH/2016) and the Ethical Committee of the Ponti cia Universidad Católica de Valparaíso, Chile (CCF02052017).

Consent for publication
Not applicable