https://orcid.org/0000-0003-0192-3563
Figures
Abstract
Background
Parents or children’s primary caregivers are a key influence on child weight as both decision makers and role models for eating patterns, physical activity, and other social behaviors. It is unknown whether caregivers’ time preferences are associated with overweight or obesity in children. The primary objective was to estimate the association between parents’ or caregivers’ time preferences and children having overweight or obesity in Mexico.
Methods
A cross-sectional study was conducted using a representative survey of the Mexican population. A multinomial logistic model was used to examine the association between parents’ or caregivers’ time preferences (patience and time consistency) and child overweight or obesity, adjusting for potential confounders.
Results
The study included 9,102 children (mean age 10, 43% female) and 5,842 caregivers (mean age 37; 95% female). Intertemporal preference was strongly associated with increased odds of overweight or obesity in children. A medium patient caregiver had higher odds of having overweight (adjusted OR: 1.73; 95% CI: 1.19, 2.52). Similarly, having a caregiver with a present (OR: 2.52; 95% CI: 1.72, 3.70) or future bias (OR: 1.48; 95% CI: 1.11, 1.98) was associated with higher odds of obesity.
Citation: Molina M, Guindon GE, Anderson LN, Tarride J-E (2024) Association between children’s caregivers time preferences and childhood overweight and obesity in Mexico. PLoS ONE 19(3): e0283455. https://doi.org/10.1371/journal.pone.0283455
Editor: Guangyu Tong, Yale University, UNITED STATES
Received: May 27, 2022; Accepted: March 8, 2023; Published: March 7, 2024
Copyright: © 2024 Molina et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data relevant to this study belong to The Mexican Family Life Survey and can be downloaded at “http://www.ennvih-mxfls.org/ennhiv-3.html. ” The authors used Book V, Book S, and Book IIB for this study. The authors confirm they did not have any special access privileges that others would not have.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Due to its high prevalence, associated morbidity, and associated costs, obesity is a major public health issue. It is well understood that obesity is a multifactorial problem, combining biological, environmental, social, behavioral, and economic elements [1, 2] as well as lifestyle decisions [3–5]. While policy decisions and environmental factors influence healthy or unhealthy choices, individuals’ time preferences also shape decisions [3]. The term ‘time preferences’ (also sometimes referred to as time discounting) describes ‘the extent to which individuals value benefits and costs that arise in the future differently (usually less) than if they arise today’ [6]. In related terms, individuals discount future rewards as they are perceived as less valuable than the same reward now [7, 8]. Individuals with a higher discount rate tend to prefer short-term gratifications, while individuals with a low discount rate favor long-term and possibly larger rewards [7, 9, 10]. These time preferences (e.g., an inclination for the present or the future) have been recognized to play an important role in understanding individuals’ health habits or eating behavior and obesity risk [10–13]. Consuming highly palatable unhealthy food provides immediate pleasure but has negative consequences in the long term, whereas healthy eating or exercise has distant benefits but an immediate cost. For example, people could value more the immediate pleasure from eating over future healthy aging. This could explain why some individuals would prefer smaller immediate rewards over larger rewards available after a delay [14].
The results of a recent systematic review of 41 studies exploring the relationship between unhealthy diets, obesity, and time discounting found that high-time discount rates (i.e., individuals valuing the present more than the future) were significant risk factors for unhealthy diets [3]. However, most studies assessed in this systematic review focussed primarily on adults, and very few explored the impact of parents’ time preferences on childhood obesity. This is important because parents or caregivers are a key influence on child weight as decision-makers and role models for eating patterns, physical activity, and other social behaviors [15, 16]. The parental transmission of obesity has been explored as a direct effect through genetics and household environment, but also as an indirect influence given by the spread of unhealthy behaviors from parents to children [17–19]. Recently one study exploring the intergenerational effects of parents’ time preferences and children’s obesity status reported that parents with a preference for the present instead of the future had a higher likelihood of having children with obesity after adjusting for several covariates [17]. However, this study was conducted in the U.S., and the generalizability of results to other settings or countries is limited. For example, in Mexico, the prevalence of obesity in children is among the highest in the world [20], and the impact of time preferences is unknown. To fill a gap in the literature and to inform future health policies, our primary objective was to determine the association between parents’ or caregivers’ time preferences and overweight and obesity in their children in Mexico. Secondary objectives were to determine the association between caregiver’s time preferences and children’s overweight or obesity by child age and sex and to identify other factors independently associated with children’s obesity status.
2. Methods
2.1 Study design and data source
A cross-sectional study was conducted using the Mexican Family Life Survey (MxFLS). The MxFLS, a representative survey of the Mexican population at the national, urban, rural, and regional levels [21], is a face-to-face in-home survey that collects extensive information at the household and individual levels, including anthropometric measures of all household members and intertemporal preferences of adults, which is conducted every six years. Specifically, it obtained information about intertemporal preferences, the obesity status of parents or caregivers and children, and information about other individual and household characteristics and lifestyle habits. Data from the 2012 MxFLS was used in these analyses, as the data from the 2018 survey was not released at the time of writing.
2.2 Study population
The analysis included children between 2–17 years old and adults (≥18 years and older) taking care of at least one of the included children. The MxFLS captures who is the main caregiver for the children, which is either a parent or another family member (e.g., grandparent, uncles/aunts). If this information was not provided, or if the main reported caregiver did not live in the same house as the children (i.e., if the children’s caregiver is an employee), the information from one of the parents living in the household was used. Hereafter, we use the term "caregivers" to relate to parents or other family members helping with childcare. Participants with complete information on time preferences and other covariables described below were included.
2.3 Dependent variable
The primary dependent variable was based on children’s body mass index (BMI, [kg/m2]). Both height and weight were measured by trained personnel, using standardized instruments, at home when carrying out the survey [21]. Child overweight and obesity were defined using the World Health Organization (WHO) BMI cut-points [22]. Child BMI z-scores were calculated based on age and sex using the recommended WHO growth standards for children aged 2 to 4 years and the WHO reference data for children aged 5–17. Children were divided into three categories: normal weight (for ages two to five, BMI z-scores <2; and older than five, z-scores <1), overweight (for ages two and five, 2≤ BMI z-score <3; and older than five, 1≤ BMI z-score <2), and obesity (for age two to five, BMI z-score ≥3; and older than five, BMI z-score ≥2) [22]. Few children were underweight; thus, this group was combined with the normal-weight children for this study.
2.4 Independent variables
The primary independent variables were the caregivers’ time preferences which were defined by two concepts: patience and time consistency. Patience refers to the extent an individual is willing to wait to receive a higher reward in the future, compared to a smaller immediate reward. Consistent with previous studies [3, 12, 23], patience was expressed in terms of patient, medium patient, and impatient. Time consistency refers to how individuals modify their willingness to wait when the time frame changes. Aligned with the literature, time consistency was expressed in terms of time consistent, present bias, and future bias [5, 17].
Fig 1 presents the trade-offs and rewards used to construct time preference measures in the MxFLS. Adult participants were asked two sets of questions. In the first set of questions, adults were asked to choose between 1,000 Mexican pesos (MXN) (approximately $50) today or MXN 1,500 (approximately USD) in a month; if they chose MXN 1,500 in a month over 1,000 today, they were asked if they would choose MXN 3,000 in a month or MXN 1,000 today (option B in Fig 1). If the answer was MXN 3,000, participants were asked whether they preferred MXN 2,000 in a month or MXN 1,000 today (option C in Fig 1). On the other hand, if participants chose MXN 1,000 today instead of MXN 1,500 in a month, participants were asked to choose between MXN 1,200 in a month or MXN 1,000 today (option D in Fig 1), and the survey ended. In addition to asking to choose a reward now or in a month, the second set of questions asked participants to decide between today and in a year similarly. These two questions were used to estimate the parameters of patience (represented by δ) and time consistency (β).
Therefore, using the survey information in the preference section and following previous literature on time preferences, as well as previously published thresholds to determine patience and time-consistency parameters [3, 10, 17], the first set of questions of the survey (select a hypothetical amount to receive today versus a possible larger sum in a month) was employed to calculate patience (i.e., δ) as:
Consistent with previous works [3, 10, 17], this information was used to categorize the children’s caregivers as patient if the children’s caregiver was willing to accept MXN 1,200 in a month instead of MXN 1,000 today (δ = 0.83 = $1000/1,200) and impatient when they preferred $1,000 today than $3,000 in a month (δ = 0.33; $1,000/$3,000). Those caregivers preferring to accept MXN 1,500 (δ = 0.67) or $2,000 in a month (δ = 0.50) instead of MXN 1,000 today were considered medium patient. Children’s caregivers who always chose to receive MXN 1,000 today (δ = 1) regardless of the hypothetical amount in a month were considered impatient, assuming there was no quantity higher enough to make them willing to defer their reward.
Similarly, we calculated the parameter for time consistency based on the responses from both sets of questions:
Children’s caregivers were classified as present biased if the amount of money they chose in a shorter term was higher than in the future (β>1, i.e., MXN 3,000 in a month and MXN 2,000 in a year), future biased if long term amount was higher (β<1, i.e., MXN 1,500 in a month and MXN 2,000 in a year), or time consistent if they chose the same amount in both 1-month and 1-year periods (β = 1). The methods used to derive the thresholds for δ and β used in this study were consistent with previous works [3, 10, 17].
Children were described in terms of age groups (i.e., 2–7, 8–12, and 13 years and older), sex, weekly hours spent in front of a screen for other proposes than schoolwork, and weekly hours allocated to physical activity. Caregivers were also described in terms of age groups (i.e., 18–39, 40 years or older), sex, education (i.e., less than high school, high school and more), marital status (single or married) and BMI categories (i.e., normal weight: BMI<25 kg/m2, overweight: 25≤BMI<30 kg/m2, and obesity BMI≥30 kg/m2). The households in which the children lived were described in terms of socioeconomic status (SES) stratified as low, medium, and high, constructed through principal component analysis with a combination of household assets and income data [24], and household location identified as rural/urban status.
2.5 Statistical analysis
Mean, standard deviations (S.D.s), and percentages for continuous and discrete variables, respectively, were first used to describe the populations. The children and the adult populations were stratified by BMI categories to identify any differences in baseline characteristics. To further describe the data, caregivers’ time preferences, patience, and time consistency were presented by parents’ BMI categories, education, household SES, and location. Multiple pairwise comparisons (Tukey’s test) for continuous variables [25] and an adaptation of Tukey’s test for discrete variables [26] were used to test for any differences between BMI categories.
To evaluate the association between caregivers’ time preferences and children’s overweight and obesity status, we used multivariable multinomial logistic regressions and reported odds ratios (OR) and 95% confidence intervals (CI) using a two-step strategy. First, to provide a baseline association between caregiver’s time preferences and children’s obesity, we adjusted for children’s age groups and sex and caregiver’s age group, sex, marital status and education, and household SES. In a second step, we ran another model in which we added the caregiver’s BMI and the weekly number of hours children spent in front of a screen or doing physical activity. Correlation and collinearity between covariates used in the model were tested using Pearson correlations and Variance Inflation Factors (VIF), respectively. To deal with potential measurement errors associated with caregivers’ BMI, we also ran a model in which caregivers’ BMI was analyzed as a continuous polynomial variable. Since the association between parental time preferences and children’s obesity status may also depend on children’s sex and age, separate stratified models were conducted by these two variables.
While multinomial logistic models are commonly used to analyze BMI categories [27, 28], a sensitivity analysis was conducted by analyzing the data using a logistic regression model with normal weight as the comparison group as it was done in the U.S. study on parental time preferences [17] as well as a generalized ordinal models (gologit). Gologit models are different from multinomial logistic models as they modify the comparison group allowing comparisons to all categories greater than the current to all those less than or equal to the current one (i.e., normal weight versus overweight and obesity; normal weight and overweight versus obesity). These models have been recently used to model ordinal data in the areas of multidrug resistance [29], tobacco or opioid consumption [30, 31], and depression or pain control [32, 33].
Sampling weights were used for all descriptive and models estimations per the MxFLS survey instructions [21]. In addition, when calculating the 95% C.I.s associated with the O.R.s, the standard errors were clustered at the household level as some children have the same caregiver and share the same household. All analyses were conducted using Stata SE 16.1.
3. Results
3.1 Study population
From the original sample of 12,089 children with information about their caregiver, 1,833 did not have complete anthropometric information, 240 missed the time allocation variable, 342 did not have information about household characteristics, and 572 did not have information about the caregiver’s time preferences. S1 Appendix presents this information. The analysis included 9,102 children between two and 17 years of age, representing more than 13 million Mexican children and adolescents when the survey results were extrapolated to the Mexican population.
The mean age of the children was almost 10 (SD = 5), 49% were female, and children spent almost 14 hours (SD = 15) per week in front of a screen and 13 (SD = 13) hours per week of physical activity. The distribution of BMI categories indicated that 70% of the children had normal weight, 18% were classified as overweight, and 12% had obesity. Older children and males were statistically significantly more likely to have overweight or obesity relative to younger children and females. Similarly, children with overweight or obesity spent statistically significantly less time physically active and more time on a screen than children with normal weight. Table 1 provides the details.
The baseline characteristics of the children’s caregivers (i.e., n = 5,842 representing more than 9.5 million Mexican caregivers) are presented in Table 2 for the overall population and stratified by children’s BMI status. Approximately 94% of the caregivers were the children’s parents, while the remaining 6% were family members. The mean age of the caregivers was 37 (SD = 11) years; 95% were female, 81% had a lower education than high school, and 87% were married. Based on caregiver BMI, 38% were classified as overweight, and 37% had obesity. Almost two-thirds (61%) of the caregivers lived in an urban setting, and more than half were in a medium SES household (56%). The stratified analyses revealed several differences in the caregivers’ characteristics by children’s BMI category. For example, caregivers of normal-weight children were younger (68%) than children with overweight (64%) or obesity (59%) (p<0.01 for all comparisons between BMI categories). Children with overweight (45%) or obesity (58%) were also more likely to have caregivers with obesity compared to children with normal weight (31%; p<0.01 for all comparisons). Caregivers living in high SES neighborhoods or urban settings were also more likely to have children with obesity. Table 2 presents the details.
3.2 Caregiver’s time preferences
Table 3 presents the children caregivers’ time preferences for the overall population and when the caregivers were stratified by their BMI categories, household characteristics, and children’s physical activity and screen time. Overall, 67% of the caregivers’ population was impatient (i.e., preferred to have an immediate reward than a larger one in the future), and 14% were patient, the remaining 19% being medium patient. Between 6% (1-month assessment) and 7% (1 year) of caregivers chose the upper value of MXN 3,000 over MXN 1,000 today (Table A1 in S1 Appendix). More than half of the caregivers (58%) were time consistent (i.e., choices did not change when considering future regards in a month or one year), and 31% had a future bias. Approximately 82% of impatient caregivers were also time-consistent (Table A2 in S1 Appendix). The stratified analyses indicated differences in caregivers’ time preferences based on their characteristics. For example, patient caregivers were more likely to live in high SES neighborhoods (i.e., 43% versus 25%; p < .005) or urban areas (68% versus 60%, p<0.05) than impatient caregivers. Caregivers with a future bias were statistically significantly more likely to live in high SES neighborhoods (39%) or in urban areas (68%) than caregivers who were time consistent (25% for high SES and 60% for urban areas) or had a present bias (32% for high SES and 55% for urban area). Other statistically significant differences were also seen in caregivers’ BMI categories or children’s time spent doing physical activities or on a screen (Table 3).
3.3 Association between caregivers’ time preferences and children’s BMI status
The results of the multivariable multinomial logistic regressions are provided in Table 4 for the overall population of children and according to the two models (Model 1 without caregiver’s BMI and children’s time spent on T.V. or physical activity, and Model 2 including all variables). There was no evidence of correlation or multicollinearity between the covariates (all correlations below 0.35 and an overall VIF = 1.12). Having a caregiver who was classified as impatient, compared to patient, was not significantly associated with increased odds of childhood overweight or obesity. According to Model 1, having a medium patient caregiver, compared to patient, was significantly associated with increased odds of overweight in children (OR 1.47; 95% CI: 1.01, 2.14). In terms of time consistency, caregivers with a present (OR: 2.26; 95% CI: 1.36, 3.73) or future (OR: 1.45; 95% CI: 1.06, 1.99) bias, compared to caregivers who were time consistent in their choices, were also more likely to have children with obesity. As shown in Table 4, the overall time preference results did not change when the caregiver’s BMI and children’s time spent on T.V. or physical activity were included, although the magnitude of the O.R.s changed slightly. For example, for the association between having a medium patient caregiver and overweight, the OR changed from 1.47 (95% CI: 1.01, 2.14) to 1.74 (95% CI: 1.19, 2.53). For a present biased caregiver, the OR changed from 2.26 (95% CI: 1.36, 3.73) to 2.52 (95% CI: 1.74, 3.66). However, the caregiver’s BMI and time spent on T.V. or physical activity were independently associated with children’s obesity.
Overall and stepwise analysis.
Considering the low number of male caregivers (N = 386) and the fact that the caregiver’s sex was not associated with children’s obesity (Table 4), we excluded the caregiver’s sex in our stratified analyses. However, we included the caregiver’s BMI and children’s time spent watching T.V. or doing physical activity. Also shown in Table 5, our stratified analyses suggested that the association between time preferences and children’s obesity status may have been modified by sex or age. Our estimates, however, were not precisely estimated (the confidence intervals were wide and overlapped across almost all comparisons). For example, having a caregiver impatient increased the odds of having obesity in children 13–17 years of age by 2.48 (95% CI: 1.60, 3.84) compared to 1.11 (95% CI: 0.68, 1.82) in children 8–12 years of age. Similarly, having a caregiver with a future bias increased the odds of having obesity in male children (OR: 2.11; 95% CI: 1.21, 3.69) while it decreased the odds of having obesity in female children (OR: 0.96; 95% CI: 0.72, 1.29).
Stratified analysis by children’s sex and age.
3.4 Other associations with children BMI status
As shown in Table 4, several other variables were also significantly associated with children’s BMI categories in the main analyses or the stratified analyses by children’s age groups and sex. Having a caregiver with overweight (except for children 8–12 years of age with obesity) or with obesity increased the odds of overweight or obesity in children compared to having a caregiver with normal weight in all analyses (i.e., females with overweight OR: 2.39; 95% CI: 1.35, 4.24 and males with overweight OR:1.96; 95% CI: 1.55, 2.49). Except for female children with obesity, living in rural areas also significantly increased the odds of overweight or obesity in children. Several other variables were associated with children’s BMI status, but the association depended on the studied populations rather than being seen across all analyses.
3.5 Sensitivity analyses
The results of the gologit model and the logistic model were mostly aligned with the multinomial logistic regression model regarding the association between time preferences and caregiver’s obesity status and children’s BMI status (see Table 6). For example, the multinomial model displayed an OR of 2.52 (95% CI:1.72, 3.70) for a present bias caregiver and a child with overweight, and the generalized ordinal model showed a coefficient of 2.42 (95% CI: 1.85, 3.17) and 1.64 (95% CI: 1.21, 2.23) in the logistic model. Differences were found in the association with an urban setting losing significance in the gologit model and in some magnitude of coefficients (but with overlapping C.I.s), i.e., having a caregiver with obesity with an OR of 4.03 (95% CI:2.90, 5.59) in the multinomial model, and 3.27 (95% CI:2.59, 4.13) and 3.23 (95% CI: 2.56, 4.07) in the generalized ordinal model, and the logistic model, respectively. The association of the other independent variables with children’s BMI status for the overall population or the stratified analyses was also consistent with the results from the multinomial logistic model (see S1 Appendix).
4. Discussion
To the best of our knowledge, this is only the second study evaluating the association between caregivers’ time preferences and overweight and obesity in children and the first study conducted in Latin America or exploring the association of caregivers’ time preferences and children’s BMI status by children age groups and sex. Based on the results of a survey representative of the Mexican population, our findings indicated that more than two-thirds of children’s caregivers were impatient, while approximately 60% were time consistent in their choices. Patient and time-consistent adults have been shown to make healthier choices [3, 10, 17, 34] than impatient or time-inconsistent adults.
The results of our analysis suggest that children caregivers’ time preferences were associated with children’s BMI status; in particular, present or future-biased time preferences were associated with increased odds of childhood overweight or obesity. Present-biased caregivers prefer to receive a larger reward in the short term than a higher reward in the future, which could lead to unhealthy behaviors for themselves and their children [17, 35], including the availability of unhealthy food at home and the lack of promotion of physical activity. While the relationship between future-biased caregivers and children’s obesity status is less intuitive, few studies have investigated the effects of future bias on healthy decisions [17, 36]. Some experiments have shown that the individual’s willingness to get a larger reward in the future evolves over time [13, 36, 37], which could explain some of the results as a future-biased individual at time t may become a present bias individually at time t+1. It is also possible that caregivers value having a bigger child, and child overweight or obesity may not be perceived as a concern. Future research should examine the reasons for a potential association between future biased time preferences and obesity.
Caregiver time preferences also showed increasing odds with children’s age which can be explained by imitation of unhealthy parental behavior–same diet and exercise routines [38, 39]. It should also be noted that in Mexico, sales of junk food are forbidden in elementary schools or nearby, while daily physical activity is only mandatory in elementary schools [40]. Compared to younger children attending elementary schools, Mexican children of middle school age are exposed to a more obesogenic environment (unhealthy food choices; lack of mandatory physician activity), which could also explain the age pattern in our study. Similarly, the obesity status of the caregivers was associated with the children’s BMI status, which may be explained by straightforward genetics, the imitation of food patterns by children, higher availability of high-dense energy food at home, or lower promotion for physical activity [41–44]. Living in an urban area was also associated with children’s BMI status [45, 46].
When comparing our results with the previous literature, BMI categories’ prevalence for children and their caregivers was consistent with official reports in Mexico for 2012 [47]. Similarly, our household characteristics (educational and SES status, rural status) were also aligned with nationally representative Mexican data [48]. It is not easy to compare our results with the only published study on parental time preferences and children’s BMI status due to important cultural and economic differences between the U.S. and Mexico. While the U.S. study found an association between patience and children’s obesity status [17], our results did not support this association. This difference could be explained because we found a higher proportion of impatient adults in Mexico than in the U.S. (66% in our study versus 27% in the U.S. [17]). Differences between countries in adults’ time preferences have been previously reported, showing a lower patience level in highly religious countries, middle-income, and with higher agricultural dependence [49]. Also, women were reported to be more impatient than men [49], which were the main caregivers in this study. Our results on the association of caregivers’ BMI status and children’s BMI were also consistent with several other studies [44, 50–53] conducted in Mexico or elsewhere. One Mexican study reported a strong association between the nutritional status of parents and obesity in children, which may be explained by a combination of genetics, eating patterns, and environment (i.e., sharing a household) [52].
Even though our findings are novel, this study has some limitations. First, the study is cross-sectional; therefore, causality could not be established. Second, caregiver time preferences were based on intertemporal monetary trade-offs rather than using direct food or dietary decisions. However, this method using intertemporal monetary trade-offs has been broadly used for determining time preferences related to eating, smoking, and other health behaviors [3, 54–56]. Third, diet or intensity of physical activity was not considered in our analyses as this information was not documented in the MxFLS. Fourth, there is a potential loss of information using categorical variables instead of continuous variables (i.e., BMI, caregiver age, SES). However, we explored models in which caregivers’ BMI was analyzed as a continuous variable, which did not change the main results. We kept caregivers’ BMI as a categorical variable in our models. Fifth, the analyses were limited by the data captured in the survey. For example, although we had information on the socioeconomic status of the caregivers, we did not have access to income/spending balance or household debt variables which could be related to time preferences. However, studies that have explored the association between household (or personal) debt and obesity have not found an association between these variables [57–59]. Finally, while in theory, caregivers could have chosen an infinite value compared to MXN 1,000 today instead of the upper value of MXN 3,000 used in the survey, the impact of a potential ceiling effect should be minimal as less than 6% (30-day assessment) and 7% (one-year assessment) chose the upper value of MXN 3,000. Despite the limitations, this study provides relevant evidence about the association between caregiver time preferences and children’s BMI status in Mexico, and this study has several strengths. First, a nationally representative sample of the Mexican population was used, so our results should be useful to inform future policies in Mexico. Secondly, several approaches were used for our regression analyses, and the results’ consistency strengthened the study findings. Finally, the study used not only parents but children’s caregivers, who are often the person making decisions when parents are not available.
Worldwide there is increasing research and policymaking promoting healthy lifestyles to reduce the burden of obesity and overweight in children and adolescents [60, 61]. Designing effective policies is crucial to understand how to support individuals to make healthy choices, especially when policies target children. Several studies have shown the importance of incorporating time preferences when designing policies to change health behaviors such as smoking, physical activity, food consumption, and health investment [11, 62–64]. However, very few studies have evaluated the impact of intergenerational time preferences. Our study found that caregiver time preferences were associated with children’s BMI status. However, the association depended on the children’s BMI status (e.g., overweight or obesity), providing new insights which could be used to develop future policies in Mexico or elsewhere. For example, given that more than two-thirds of caregivers in Mexico were impatient (i.e., 67% as shown in this study), policies aimed at reducing childhood obesity should tailor communications by framing short-term instead of long-term outcomes [64] or creating short-term financial incentive-based interventions [5, 12, 63] An example of such a policy is the ’PROGRESA’ program in Mexico which provides monetary incentives in exchange of school attendance and routine G.P. visits for children [65]. These policies should also consider other important factors such as children’s caregiver BMI status or geography (i.e., urban settings), which were also strongly associated with children’s BMI status.
References
- 1. Popkin BM. The Nutrition Transition: An Overview of World Patterns of Change. Nutr Rev [Internet]. 2004 Jul [cited 2021 Mar 25];62:S140–3. Available from: https://academic.oup.com/nutritionreviews/article-lookup/doi/10.1111/j.1753-4887.2004.tb00084.x pmid:15387480
- 2. Lobstein T, Baur L, Uauy R. Obesity in children and young people: a crisis in public health. Obes Rev [Internet]. 2004 May [cited 2021 Mar 25];5(s1):4–85. Available from: http://doi.wiley.com/10.1111/j.1467-789X.2004.00133.x pmid:15096099
- 3. Barlow P, Reeves A, McKee M, Galea G, Stuckler D. Unhealthy diets, obesity and time discounting: a systematic literature review and network analysis: Time discounting, diets and obesity. Obes Rev [Internet]. 2016 Sep [cited 2020 Oct 20];17(9):810–9. Available from: http://doi.wiley.com/10.1111/obr.12431
- 4. Hruby A, Hu FB. The Epidemiology of Obesity: A Big Picture. PharmacoEconomics [Internet]. 2015 Jul [cited 2021 May 10];33(7):673–89. Available from: http://link.springer.com/10.1007/s40273-014-0243-x pmid:25471927
- 5. Loewenstein G, Asch DA, Friedman JY, Melichar LA, Volpp KG. Can behavioural economics make us healthier? BMJ [Internet]. 2012 May 23 [cited 2021 May 10];344(may23 1):e3482–e3482. Available from: https://www.bmj.com/lookup/doi/10.1136/bmj.e3482 pmid:22623635
- 6.
Hurley JE. Health economics. 1. ed. S.l.: McGraw-Hill Ryerson; 2010. 434 p.
- 7. Frederick S, Loewenstein G, O’donoghue T. Time Discounting and Time Preference: A Critical Review. J Econ Lit [Internet]. 2002 Jun [cited 2022 Dec 14];40(2):351–401. Available from: http://pubs.aeaweb.org/doi/abs/10.1257/002205102320161311
- 8. Kawakami N, Watanabe K, Nishi D, Takagi D, Hashimoto H, Tanaka SC. Time preference and personal value: a population-based cross-sectional study in Japan. BMC Psychol [Internet]. 2020 Dec [cited 2022 Dec 14];8(1):85. Available from: https://bmcpsychology.biomedcentral.com/articles/10.1186/s40359-020-00458-6 pmid:32807238
- 9. Laibson D. Golden Eggs and Hyperbolic Discounting. Q J Econ [Internet]. 1997 May 1 [cited 2022 Aug 5];112(2):443–78. Available from: https://academic.oup.com/qje/article-lookup/doi/10.1162/003355397555253
- 10. Hoch SJ, Loewenstein GF. Time-Inconsistent Preferences and Consumer Self-Control. J Consum Res [Internet]. 1991 Mar [cited 2021 Mar 26];17(4):492. Available from: https://academic.oup.com/jcr/article-lookup/doi/10.1086/208573
- 11. Hunter RF, Tang J, Hutchinson G, Chilton S, Holmes D, Kee F. Association between time preference, present-bias and physical activity: implications for designing behavior change interventions. BMC Public Health [Internet]. 2018 Dec [cited 2021 Dec 7];18(1):1388. Available from: https://bmcpublichealth.biomedcentral.com/articles/10.1186/s12889-018-6305-9 pmid:30567532
- 12. Zhang L, Rashad I. OBESITY AND TIME PREFERENCE: THE HEALTH CONSEQUENCES OF DISCOUNTING THE FUTURE. J Biosoc Sci [Internet]. 2008 Jan [cited 2020 Jan 29];40(1):97–113. Available from: https://www.cambridge.org/core/product/identifier/S0021932007002039/type/journal_article pmid:17517169
- 13. Takeuchi K. Non-parametric test of time consistency: Present bias and future bias. Games Econ Behav [Internet]. 2011 Mar [cited 2022 Aug 5];71(2):456–78. Available from: https://linkinghub.elsevier.com/retrieve/pii/S0899825610000904
- 14. Loewenstein G, Brennan T, Volpp KG. Asymmetric Paternalism to Improve Health Behaviors. JAMA [Internet]. 2007 Nov 28 [cited 2021 Mar 25];298(20):2415. Available from: http://jama.jamanetwork.com/article.aspx?doi=10.1001/jama.298.20.2415 pmid:18042920
- 15. Gicevic S, Aftosmes-Tobio A, Manganello JA, Ganter C, Simon CL, Newlan S, et al. Parenting and childhood obesity research: a quantitative content analysis of published research 2009–2015: Parenting and childhood obesity research. Obes Rev [Internet]. 2016 Aug [cited 2021 Mar 25];17(8):724–34. Available from: http://doi.wiley.com/10.1111/obr.12416
- 16. Gray LA, Hernandez Alava M, Kelly MP, Campbell MJ. Family lifestyle dynamics and childhood obesity: evidence from the millennium cohort study. BMC Public Health [Internet]. 2018 Dec [cited 2021 May 10];18(1):500. Available from: https://bmcpublichealth.biomedcentral.com/articles/10.1186/s12889-018-5398-5 pmid:29807535
- 17. Stoklosa M, Shuval K, Drope J, Tchernis R, Pachucki M, Yaroch A, et al. The intergenerational transmission of obesity: The role of time preferences and self-control. Econ Hum Biol [Internet]. 2018 Feb [cited 2020 Jan 29];28:92–106. Available from: https://linkinghub.elsevier.com/retrieve/pii/S1570677X17302678 pmid:29294461
- 18. Classen TJ, Thompson O. Genes and the intergenerational transmission of BMI and obesity. Econ Hum Biol [Internet]. 2016 Dec [cited 2022 Aug 5];23:121–33. Available from: https://linkinghub.elsevier.com/retrieve/pii/S1570677X1630106X pmid:27599025
- 19. Clark HR, Goyder E, Bissell P, Blank L, Peters J. How do parents’ child-feeding behaviours influence child weight? Implications for childhood obesity policy. J Public Health [Internet]. 2007 Jun 1 [cited 2021 May 31];29(2):132–41. Available from: http://academic.oup.com/jpubhealth/article/29/2/132/1505733/How-do-parents-childfeeding-behaviours-influence pmid:17442696
- 20.
Shamah-Levy T, Vielma-Orozco E, Heredia-Hernández O, Romero-Martínez M, Mojica-Cuevas J, Cuevas-Nasu L, et al. Encuesta Nacional de Salud y Nutrición 2018–19: Resultados Nacionales. [Internet]. s. Cuernavaca, México; 2020. Available from: https://ensanut.insp.mx/encuestas/ensanut2018/doctos/informes/ensanut_2018_informe_final.pdf
- 21. Rubalcava L, Teruel G. Encuesta Nacional sobre Niveles de Vida de los Hogares, Tercera Ronda [Internet]. 2013. Available from: www.ennvih-mxfls.org
- 22. WHO. Obesity and overweight: Key facts. [Internet]. 2020. Available from: https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight
- 23. Davis C, Patte K, Levitan R, Reid C, Tweed S, Curtis C. From motivation to behaviour: A model of reward sensitivity, overeating, and food preferences in the risk profile for obesity. Appetite [Internet]. 2007 Jan [cited 2020 Jan 29];48(1):12–9. Available from: https://linkinghub.elsevier.com/retrieve/pii/S019566630600420X pmid:16875757
- 24. Poirier MJP, Grépin KA, Grignon M. Approaches and Alternatives to the Wealth Index to Measure Socioeconomic Status Using Survey Data: A Critical Interpretive Synthesis. Soc Indic Res [Internet]. 2020 Feb [cited 2021 Jun 22];148(1):1–46. Available from: http://link.springer.com/10.1007/s11205-019-02187-9
- 25.
Haynes W. Tukey’s Test. In: Dubitzky W, Wolkenhauer O, Cho KH, Yokota H, editors. Encyclopedia of Systems Biology [Internet]. New York, NY: Springer New York; 2013 [cited 2021 Oct 20]. p. 2303–4. Available from: http://link.springer.com/10.1007/978-1-4419-9863-7_1212
- 26. Elliott AC, Reisch JS. Implementing a Multiple Comparison Test for Proportions in a 2 xc Crosstabulation in. In 2006.
- 27. Lohse T, Rohrmann S, Faeh D, Hothorn T. Continuous outcome logistic regression for analyzing body mass index distributions. F1000Research [Internet]. 2017 Nov 1 [cited 2021 Oct 20];6:1933. Available from: https://f1000research.com/articles/6-1933/v1 pmid:29259768
- 28. Juvanhol LL, Lana RM, Cabrelli R, Bastos LS, Nobre AA, Rotenberg L, et al. Factors associated with overweight: are the conclusions influenced by choice of the regression method? BMC Public Health [Internet]. 2016 Dec [cited 2021 Oct 20];16(1):642. Available from: https://bmcpublichealth.biomedcentral.com/articles/10.1186/s12889-016-3340-2 pmid:27461119
- 29. Agga GE, Scott HM. Use of generalized ordered logistic regression for the analysis of multidrug resistance data. Prev Vet Med [Internet]. 2015 Oct [cited 2021 Oct 4];121(3–4):374–9. Available from: https://linkinghub.elsevier.com/retrieve/pii/S0167587715300040 pmid:26342790
- 30. Romberg AR, Miller Lo EJ, Barton AA, Xiao H, Vallone DM, Hair EC. Cigarette smoking, prescription opioid use and misuse among young adults: An exploratory analysis. Prev Med [Internet]. 2019 Dec [cited 2021 Oct 17];129:105845. Available from: https://linkinghub.elsevier.com/retrieve/pii/S0091743519303214 pmid:31518628
- 31. Romberg AR, Rath JM, Miller Lo EJ, Mayo A, Liu M, Vallone DM, et al. Young Adults’ Opioid Prescription History and Opioid Misuse Perceptions. Am J Health Behav [Internet]. 2019 Mar 1 [cited 2021 Oct 17];43(2):361–72. Available from: https://www.ingentaconnect.com/content/10.5993/AJHB.43.2.12 pmid:30808475
- 32. Lin L yi, Sidani JE, Shensa A, Radovic A, Miller E, Colditz J.B., et al. ASSOCIATION BETWEEN SOCIAL MEDIA USE AND DEPRESSION AMONG U.S. YOUNG ADULTS: Research Article: Social Media and Depression. Depress Anxiety [Internet]. 2016 Apr [cited 2021 Oct 17];33(4):323–31. Available from: https://onlinelibrary.wiley.com/doi/10.1002/da.22466
- 33. Cai B, Oderda GM. The Association Between Pain and Depression and Some Determinants of Depression for the General Population of the United States. J Pain Palliat Care Pharmacother [Internet]. 2012 Sep 4 [cited 2021 Oct 17];26(3):257–65. Available from: http://www.tandfonline.com/doi/full/10.3109/15360288.2012.703292 pmid:22973915
- 34. Chen S, Li G. Time-inconsistent preferences, consumption, investment and life insurance decisions. Appl Econ Lett [Internet]. 2020 Mar 11 [cited 2021 Mar 26];27(5):392–9. Available from: https://www.tandfonline.com/doi/full/10.1080/13504851.2019.1617395
- 35. Santiago-Torres M, Adams AK, Carrel AL, LaRowe TL, Schoeller DA. Home Food Availability, Parental Dietary Intake, and Familial Eating Habits Influence the Diet Quality of Urban Hispanic Children. Child Obes [Internet]. 2014 Oct [cited 2022 Aug 5];10(5):408–15. Available from: http://www.liebertpub.com/doi/10.1089/chi.2014.0051 pmid:25259675
- 36. Attema AE, Brouwer WBF, Claxton K. Discounting in Economic Evaluations. PharmacoEconomics [Internet]. 2018 Jul [cited 2021 Aug 17];36(7):745–58. Available from: http://link.springer.com/10.1007/s40273-018-0672-z pmid:29779120
- 37. Scholten M, Read D. Discounting by Intervals: A Generalized Model of Intertemporal Choice. Manag Sci [Internet]. 2006 Sep [cited 2022 Dec 14];52(9):1424–36. Available from: http://pubsonline.informs.org/doi/10.1287/mnsc.1060.0534
- 38. García García E, Vázquez López MÁ, Galera Martínez R, Alias I, Martín González M, Bonillo Perales A, et al. Prevalencia de sobrepeso y obesidad en niños y adolescentes de 2 a 16 años. Endocrinol Nutr [Internet]. 2013 Mar [cited 2022 Aug 5];60(3):121–6. Available from: https://linkinghub.elsevier.com/retrieve/pii/S1575092212003270
- 39. Li X, Xiang ST, Dong J, Zhong Y, Zhao S, Xiao Z, et al. Association between Physical Activity and Age among Children with Overweight and Obesity: Evidence from the 2016–2017 National Survey of Children’s Health. BioMed Res Int [Internet]. 2020 Sep 24 [cited 2022 Aug 5];2020:1–8. Available from: https://www.hindawi.com/journals/bmri/2020/9259742/ pmid:33029532
- 40. Secretaría de Gobernación. POLÍTICAS PÚBLICAS PARA UNA EDUCACIÓN FÍSICA DE CALIDAD UNESCO/MÉXICO [Internet]. 2018. Available from: https://www.gob.mx/salud/documentos/politicas-publicas-para-una-educacion-fisica-de-calidad-unesco-mexico-167271
- 41. Pearson N, Biddle SJ, Gorely T. Family correlates of fruit and vegetable consumption in children and adolescents: a systematic review. Public Health Nutr [Internet]. 2009 Feb [cited 2021 Mar 26];12(2):267–83. Available from: https://www.cambridge.org/core/product/identifier/S1368980008002589/type/journal_article pmid:18559129
- 42. Nepper MJ. Associations of the Home Food Environment with Eating Behaviors and Weight Status among Children and Adolescents. J Nutr Food Sci [Internet]. 2015 [cited 2021 Mar 26];s12. Available from: https://www.omicsonline.org/open-access/associations-of-the-home-food-environment-with-eating-behaviors-and-weight-status-among-children-and-adolescents-2155-9600-S12-004.php?aid=57387
- 43. Jago R, Davison KK, Thompson JL, Page AS, Brockman R, Fox KR. Parental Sedentary Restriction, Maternal Parenting Style, and Television Viewing Among 10- to 11-Year-Olds. PEDIATRICS [Internet]. 2011 Aug 22 [cited 2021 Mar 26];peds.2010–3664. Available from: http://pediatrics.aappublications.org/cgi/doi/10.1542/peds.2010-3664 pmid:21859910
- 44. Keane E, Layte R, Harrington J, Kearney PM, Perry IJ. Measured Parental Weight Status and Familial Socioeconomic Status Correlates with Childhood Overweight and Obesity at Age 9. Kazembe L, editor. PLoS ONE [Internet]. 2012 Aug 17 [cited 2022 Aug 5];7(8):e43503. Available from: https://dx.plos.org/10.1371/journal.pone.0043503 pmid:22912886
- 45. Jia P, Xue H, Cheng X, Wang Y, Wang Y. Association of neighborhood built environments with childhood obesity: Evidence from a 9-year longitudinal, nationally representative survey in the U.S. Environ Int. 2019 Jul;128:158–64.
- 46. Nurwanti E, Hadi H, Chang JS, Chao JCJ, Paramashanti BA, Gittelsohn J, et al. Rural–Urban Differences in Dietary Behavior and Obesity: Results of the Riskesdas Study in 10–18-Year-Old Indonesian Children and Adolescents. Nutrients. 2019 Nov 18;11(11):2813. pmid:31752101
- 47.
Oropeza Abúndez C, Instituto Nacional de Salud Pública (Mexico), Mexico, editors. Encuesta nacional de salud y nutrición 2012: resultados nacionales. Primera edición. Cuernavaca, Morelos, México: Instituto Nacional de Salud Pública : Secretaría de Salud; 2012. 195 p.
- 48. INEGI. Encuesta Nacional de Ingresos y Gastos de los Hogares 2020, Nueva serie [Internet]. 2020. Available from: https://www.inegi.org.mx/rnm/index.php/catalog/685/variable/F13/V1758?name=est_socio
- 49. Falk A, Becker A, Dohmen T, Enke B, Huffman D, Sunde U. Global Evidence on Economic Preferences*. Q J Econ [Internet]. 2018 Nov 1 [cited 2021 Mar 26];133(4):1645–92. Available from: https://academic.oup.com/qje/article/133/4/1645/5025666
- 50. Wang Y, Min J, Khuri J, Li M. A Systematic Examination of the Association between Parental and Child Obesity across Countries. Adv Nutr Int Rev J [Internet]. 2017 May [cited 2021 Mar 25];8(3):436–48. Available from: https://academic.oup.com/advances/article/8/3/436-448/4558063 pmid:28507009
- 51. Martínez-Villanueva J, González-Leal R, Argente J, Martos-Moreno GÁ. Parental obesity is associated with the severity of childhood obesity and its comorbidities. An Pediatría Engl Ed [Internet]. 2019 Apr [cited 2021 Mar 26];90(4):224–31. Available from: https://linkinghub.elsevier.com/retrieve/pii/S2341287919300365
- 52. Klünder-Klünder M, Cruz M, Medina-Bravo P, Flores-Huerta S. Do children of parents with overweight and obesity have an increased risk of developing obesity and changes in the blood pressure? Boletín médico del Hospital Infantil de México. 2011;68(6).
- 53. Fuemmeler BF, Lovelady CA, Zucker NL, Østbye T. Parental obesity moderates the relationship between childhood appetitive traits and weight: Childhood Appetitive Traits and Obesity. Obesity [Internet]. 2013 Apr [cited 2021 Mar 26];21(4):815–23. Available from: http://doi.wiley.com/10.1002/oby.20144
- 54. Madsen KP, Kjær T, Skinner T, Willaing I. Time preferences, diabetes self‐management behaviours and outcomes: a systematic review. Diabet Med [Internet]. 2019 Nov [cited 2021 Mar 26];36(11):1336–48. Available from: https://onlinelibrary.wiley.com/doi/abs/10.1111/dme.14102 pmid:31392757
- 55. Peretti-Watel P, L’Haridon O, Seror V. Time preferences, socioeconomic status and smokers’ behaviour, attitudes and risk awareness. Eur J Public Health [Internet]. 2013 Oct 1 [cited 2021 Mar 26];23(5):783–8. Available from: https://academic.oup.com/eurpub/article-lookup/doi/10.1093/eurpub/cks189 pmid:23345320
- 56. Brown H, Pol M. THE ROLE OF TIME PREFERENCES IN THE INTERGENERATIONAL TRANSFER OF SMOKING. Health Econ [Internet]. 2014 Dec [cited 2021 Mar 26];23(12):1493–501. Available from: https://onlinelibrary.wiley.com/doi/10.1002/hec.2987 pmid:23959580
- 57. Averett SL, Smith JK. Financial hardship and obesity. Econ Hum Biol [Internet]. 2014 Dec [cited 2022 Aug 3];15:201–12. Available from: https://linkinghub.elsevier.com/retrieve/pii/S1570677X13001263 pmid:24411309
- 58. Keese M, Schmitz H. Broke, Ill, and Obese: Is There an Effect of Household Debt on Health?: Broke, Ill, and Obese. Rev Income Wealth [Internet]. 2014 Sep [cited 2022 Aug 3];60(3):525–41. Available from: https://onlinelibrary.wiley.com/doi/10.1111/roiw.12002
- 59. Averett SL, Smith JK. Indebted and overweight: The link between weight and household debt. IZA DP [Internet]. (6898). Available from: https://www.econstor.eu/bitstream/10419/67186/1/727546546.pdf
- 60. PAHO. Plan of Action for the Prevention of Obesity in Children and Adolescents [Internet]. 2014. Available from: https://iris.paho.org/bitstream/handle/10665.2/49138/obesity-plan-of-action-child_eng_2015.pdf?sequence=1&isAllowed=y
- 61. Molina M, Anderson LN, Guindon GE, Tarride J. A review of implementation and evaluation of Pan American Health Organization’s policies to prevent childhood obesity in Latin America. Obes Sci Pract. 2021 Oct 30;osp4.572.
- 62. Zauberman G, Urminsky O. Consumer intertemporal preferences. Curr Opin Psychol [Internet]. 2016 Aug [cited 2022 Jan 17];10:136–41. Available from: https://linkinghub.elsevier.com/retrieve/pii/S2352250X16000075
- 63. Koffarnus MN, Jarmolowicz DP, Mueller ET, Bickel WK. CHANGING DELAY DISCOUNTING IN THE LIGHT OF THE COMPETING NEUROBEHAVIORAL DECISION SYSTEMS THEORY: A REVIEW: CHANGING DELAY DISCOUNTING. J Exp Anal Behav [Internet]. 2013 Jan [cited 2021 Dec 7];99(1):32–57. Available from: https://onlinelibrary.wiley.com/doi/10.1002/jeab.2
- 64. Ortendahl M, Fries JF. Framing health messages based on anomalies in time preference. Med Sci Monit Int Med J Exp Clin Res. 2005 Aug;11(8):RA253–256. pmid:16049392
- 65. Campos-Alanis J. Análisis del Programa de Educación, Salud y Alimentación (Progresa) en el municipio de San Felipe del Progreso, Estado de México, 1998–2000. Papeles de población. 2003;9(35):181–207.