Sedentary behavior moderates the relationship between physical activity and cardiometabolic risk in young Latino children

Abstract

This study investigated the role of objectively measured moderate–vigorous physical activity (MVPA) and sedentary behavior on cardiometabolic risk factors of young Latino children. We hypothesized that MVPA would be associated with lower cardiometabolic risk when sedentary behavior is low. We studied 86 primarily low-income, Latino children using a cross-sectional study design. The study sample consisted of 51 girls and 35 boys, with mean age 5.6 (SD = .53) years. Physical activity was measured by accelerometry, anthropometric measures obtained, and fasting blood samples were used to measure cardiometabolic risk factors. Greater levels of sedentary behavior were associated with increased waist circumference (r_s = .24, p < .05) and metabolic risks. MVPA, however, had significant beneficial associations with all cardiometabolic risk factors (r_s-range = −.20 to −.45, p < .05) with the exception of plasma insulin. MVPA predicted latent variables representing anthropometric risk (β = −.57, p < .01), cardiac risk (β = −.74, p < .01), and metabolic risk (β = −.88, p < .01). Sedentary behavior significantly moderated the effect of MVPA on anthropometric (β-interaction = .49, p < .01), cardiac (β-interaction = .45, p < .01), and metabolic risk (β-interaction = .77, p < .01), such that more MVPA was associated with better health outcomes under conditions of lower sedentary behavior. The model explained 13%, 22%, and 45% variance in anthropometric, cardiac, and metabolic risk factors, respectively. Increased MVPA is associated with decreased cardiometabolic risk in young Latino children, particularly when sedentary behavior is low.

Lay Summary

This study investigated the role of objectively measured moderate-vigorous physical activity (MVPA) and sedentary behavior on cardiometabolic risk factors of young Latino children. We found that sedentary behavior significantly moderated the effect of MVPA on cardiometabolic risk, such that more MVPA was associated with better health outcomes under conditions of lower sedentary behavior.

Implications.

Practice: Young children should be encouraged not only to engage in recommended levels of moderate–vigorous physical activity, but also to reduce sedentary behavior in order to gain more health benefits from their physical activity.

Policy: These findings support policy guidelines concerning both physical activity and sedentary behavior in children and further suggest that policy efforts to reduce sedentary behavior be strengthened in consideration of the important health benefits.

Research: More research is needed using longitudinal study designs and larger multi-ethnic samples of children to explicate our understanding of the role of sedentary behavior as a moderator of the relationship between MVPA and cardiometabolic risk in children.

INTRODUCTION

Rates of pediatric obesity among ethnic minority children are estimated at 38.9% which is remarkably high [1]. Prevalence rates are especially high in young Latino children, with an estimated 46.2% considered to be overweight or obese compared to 29.4% of non-Latino white children [1]. Furthermore, studies have shown poorer health outcomes in Latino children: compared to obese non-Latino white children, obese Latino children have more metabolic problems such as insulin resistance and fatty liver [2, 3].

Considerable research has documented the beneficial role of physical activity in affecting health outcomes [4]. Objective measures of physical activity and sedentary behavior [5–9] have added precision in establishing these links [10–12]. Such studies including objective measures of physical activity have also helped in the determination of guidelines for healthy lifestyles and demonstrated numerous health benefits for children [13–16].

For example, a recent study in Latino youth showed associations of greater moderate–vigorous physical activity (MVPA) with lower triglycerides and less sedentary behavior with higher levels of HDL-cholesterol [10]. Another study with ethnic minority youth (74% Latino) showed that higher MVPA was related to a decrease in metabolic syndrome, with more MVPA associated with lower fasting glucose and systolic blood pressure [17]. This study also reported a positive relationship between sedentary behavior and metabolic syndrome and systolic blood pressure, whereas sedentary behavior was negatively correlated with HDL-cholesterol. Similarly, a study of Danish children showed greater MVPA and less sedentary behavior was associated with lower cardiometabolic risk [18].

Studies using objective measures of physical activity indicate consistent findings regarding the significance of MVPA in reducing cardiometabolic risk factors [15, 17, 19, 20], but studies examining these associations in low-income ethnic minority youth have either underestimated or ignored the role of sedentary behavior. Results of intervention studies indicate the effectiveness of PA interventions for low-income ethnic minorities [15], and increased MVPA is consistently associated with decreased cardiometabolic disease risk factors [16]. However, the role of sedentary behavior has been mixed with some studies showing no associations [21–24], while others show some associations with indicators of health [25, 26].

A study involving 120 overweight/obese children showed an inverse relationship between sedentary behavior and HDL: participants in the lowest quartile of 30-min bouts per day of sedentary behavior had 12% higher HDL than those in the highest quartile [27]. Questionnaire-based cross-sectional and longitudinal research in adults has consistently shown that sedentary behavior is related to higher cardiometabolic risk scores, body fat percentage, waist circumference, systolic blood pressure, and triglycerides [28–31]. However, empirical investigations using objective measures of physical activity and sedentary behaviors have not yet effectively determined the potential health risks of sedentary behavior especially in youth [21, 32].

A recent study from Norway used objective measures of physical activity and sedentary behavior with a sample of 700 10-year-old children and concluded that physical activity, but not sedentary time, was prospectively associated with lower cardiometabolic risk in healthy children [33]. A study using a nationally representative sample (National Health and Nutrition Examination Survey data) of 655 American children ages 12–17 years concluded that metabolic syndrome risk scores decreased by more than four times as physical activity increased from sedentary to vigorous levels [34]. Longitudinal studies with adults have shown a decrease in the relationship between sedentary behavior and cardiometabolic risk when physical fitness is taken into account [35], and the relationship between sedentary behavior and insulin resistance is reduced when controlling for MVPA [36].

In a systematic review examining physical activity and metabolic syndrome in children and adolescents, higher physical activity levels were associated with reduced risk for metabolic syndrome and an improved cardiometabolic profile [37]. In a study of 536 children in Canada, higher levels of MVPA were associated with lower waist circumference, fasting triglycerides, and diastolic blood pressure, and higher HDL-cholesterol, independent of sedentary time [38]. Results of a meta-analysis involving over 20,000 children and adolescents (75% White) showed that MVPA was significantly associated with all cardiometabolic outcomes, yet sedentary time was not associated with any outcome independent of time spent in MVPA [39].

Because most studies have been unable to identify links between sedentary behavior alone and cardiometabolic risk factors [38], it is possible that the influential role of MVPA may result in underestimation of the potential hazards of sedentary behavior [40, 41]. Examining sedentary behavior independent of MVPA is not likely to allow conclusions to be made regarding associations with cardiometabolic risk. Studies involving both MVPA and sedentary behavior have mainly attempted to investigate the associations of both factors with cardiometabolic risk factors [10, 17, 27]. This is particularly important to examine in young children as previous longitudinal research demonstrated that there is a decline in MVPA and increase in sedentary behavior as young children get older [42]. Additionally, the American Heart Association has noted the significance of examining cardiometabolic risk factors in Latino populations [43]. Therefore, the objective of this study was to investigate the role of sedentary behavior and MVPA in young Latino children beyond simple associations with cardiometabolic risk factors.

We hypothesized that the significant effect of MVPA would be associated with a decrease in cardiometabolic risk factors and sedentary behavior would be associated with an increase in cardiometabolic risk factors. We further hypothesized that sedentary behavior would moderate the relationship between MVPA and cardiometabolic risk factors. Due to the significance of MVPA, the independently predictive role of sedentary behavior may be masked [44]. However, under varying conditions of sedentary behavior, changing patterns of relationships [45, 46] between MVPA and cardiometabolic risk factors may help to elaborate the potential risk of sedentary behavior in cardiometabolic health risk indicators. More specifically, we hypothesized that greater levels of MVPA would result in lower cardiometabolic risk when sedentary behavior is low. In other words, we predicted that increased levels of sedentary behavior would be associated with a decline of the potential benefit of increased MVPA, particularly in a sample of young Latino children, in whom few studies have addressed this issue.

METHODS

Study sample and procedures

Participants were recruited for a pilot obesity prevention program from two local urban elementary schools with a largely Latino population in low-income urban neighborhoods of South Florida. The research was approved by the appropriate university Institutional Review Board. Teachers in all kindergarten and first grade classrooms at the two schools distributed study flyers describing the project to approximately 600 parents; 234 parents returned flyers indicating their interest in participating in the study. Interested and eligible families then were scheduled for a meeting at the school and written consent was obtained from parents for their children’s participation. Study measures included: demographics (provided by parents); direct measurement of children’s anthropometrics and blood pressure; a fasting blood sample that provided measures of glucose, insulin, proinsulin, and a lipid panel (including total cholesterol, LDL-cholesterol, HDL-cholesterol, and triglycerides); and measurement of physical activity as described below. These measures constituted the baseline assessment for the pilot prevention study (reported separately). Thus, the current study employed a cross-sectional study design.

The study sample consisted of 86 children including 51 girls (59.3%) and 35 boys; other children were potentially available, but they and their parents who were interested either were not able to be reached to schedule an appointment or did not keep their study appointment. Given the study timeline, enrollment was discontinued for this pilot study after the sample of 86 children had been enrolled. Participants were primarily low income and most (95.3%) were Hispanic. The mean age of the children was 5.6 (SD = .53) years, and 56% received free or reduced lunch. Regarding parent demographics, 88% of participating parents were female, and 91% were born outside the US, with a mean duration of living in the US of 16 (SD = 10) years; 67% of parents reported having a high school degree or less; mean family income was $30,400; 48% had Medicaid; and 29% received food stamps.

Parents and children were instructed about the use of the device for measurement of physical activity. During the following week, children participated in physical activity monitoring wearing a uniaxial accelerometer (Lifecorder Plus, Suzuken Co. Ltd, Japan) [47] on a Velcro waist belt that was calibrated by entering the child’s age, weight, height, and current time and date. Participants and their caregivers were instructed to have the child wear the waist belt for seven days. Caregivers were told that the child should wear the waist belt from the time of awakening and remove it before going to bed or if they were bathing or swimming. After seven days, a research assistant collected the accelerometer from the child at school and later uploaded the data to a web interface. Seventy-three devices were returned with valid data (85% response rate); only these data were included in the statistical analyses (without imputation of missing data).

Measures

Physical Activity

Lifecorder Plus was used to measure physical activity and sedentary behavior; this device has been shown to be valid in children [47]. The device provided activity data in four-second epochs over 2-min periods as levels of intensity. Intensity levels range from 0 to 9 with 0 being no activity. Once uploaded to the web interface, each data point represents the measured intensity over the 2-min period following the marked time. The intensity of a 2-min period is measured as the most frequent intensity recorded during the 30 four-second epochs in that 2 min. Activity ranging from 1.0 to 1.5 METs was defined as sedentary behavior and activity ranging 4.0 to 9.0 METs was defined as moderate to vigorous activity [48]. The four best days (i.e., days with the most data recorded) of the seven days worn were used in the analyses. Non-wear time was defined as 40 consecutive minutes of zero activity recording. Daily average sedentary behavior was estimated as average time of minimal activity (intensity level ranging from 1 to 1.5) divided by intensity of activity; daily average MVPA was calculated by the average intensity of activity in a day (intensity level > 4) dividing minutes of moderate to vigorous activity. National guidelines (i.e., at least 60 min per day) were used to determine recommended minutes of daily MVPA.

Anthropometric

Several anthropometric measures were used including height (by stadiometer) and weight (by balance scale), standardized body mass index (using CDC national norms for gender and age to calculate zBMI), percent body fat (measured by a Tanika bioelectric impedance scale), and waist and hip circumference (in cm measured three times and using the mean of the second and third measurements).

Cardiometabolic

Cardiometabolic measures included: systolic blood pressure (SBP; measured by an automated Dinamap blood pressure monitor); high-density lipoprotein (HDL), low-density lipoprotein (LDL), and triglycerides; fasting blood glucose, plasma insulin and proinsulin, and estimated insulin resistance based on fasting glucose and insulin (Homeostatic Modeling Assessment; HOMA-IR) [49]. American Heart Association guidelines were used for measurement of blood pressure, using the mean of the second and third resting blood pressure measurements over a 5-min period. These average scores were converted to percentiles scores based on norms for children (considering their age, sex, and height). Blood samples were obtained by venipuncture in the morning after an overnight fast and analyzed at the Clinical Chemistry Laboratory of the University of Miami Diabetes Research Institute using standardized assays and procedures.

RESULTS

Descriptive results for the study sample for the main variables are provided in Table 1. Preliminary analysis showed that the major predictors, i.e., sedentary behavior and MVPA, were negatively skewed (>−3), whereas fasting blood glucose was positively skewed (>4). Due to substantial deviation from normality, we used nonparametric estimations in testing relationship and hypothesis. Spearman rho correlations were computed to assess relationships between MVPA, sedentary behavior, and the anthropometric and health outcome measures.

Table 1 |.

Descriptive results for study sample

	Mean	SD	Min	Max	Skewness	Kurtosis
Age (years)	5.60	0.56	4.00	7.00	−0.22	−0.53
Body fat (%)	22.48	9.92	7.20	56.90	1.05	1.78
zBMI	0.94	1.11	−1.61	3.33	0.14	−0.61
Waist circumference (mm)	60.15	8.96	48.35	93.33	1.52	2.62
Hip circumference (mm)	69.10	8.22	56.30	98.30	1.43	2.56
SBP percentiles (%)	66.36	20.43	50.00	99.00	0.46	1.79
HDL (mg/dL)	49.59	12.61	23.00	96.00	0.90	1.64
LDL (mg/dL)	103.71	21.71	57.00	169.00	0.60	0.90
Blood glucose (mg/dL)	85.74	10.41	68.00	162.00	4.64	33.83
Plasma insulin (mU/L)	8.09	5.49	0.45	27.80	1.37	2.30
Proinsulin (pmol/L)	4.21	2.91	1.90	18.40	2.67	8.43
HOMA	1.73	1.19	0.08	5.92	1.21	1.47
Triglyceride (mg/dL)	81.73	34.53	38.00	190.00	1.22	1.11
Sedentary behavior	3.03	0.58	0.50	3.94	−3.45	13.08
MVPA	172.29	26.15	0.00	212.16	−3.92	25.73

zBMI, body mass index (Z-scores); SBP, systolic blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein; HOMA, homeostatic model assessment.

Results presented in Table 2 showed that being female was associated with decreased sedentary behavior (r_s = .21, p < .05) and increased MVPA (r_s = −.31, p < .01), and being younger was associated with increased MVPA (r_s = −.21, p < .05). Older age of children was associated with a lower fasting blood glucose (r_s = −.22, p < .05), while higher blood glucose was associated with being male (r_s = .31, p < .01). Though being male was associated with lower percent body fat (r_s = −.19, p < .05), boys were still higher on zBMI (r_s = .26, p < .01), had higher SBP (r_s = .19, p < .05), and had higher values of HOMA-IR (r_s = .19, p < .05) compared to girls.

Table 2 |.

Bivariate correlations among study variables

		1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16
1	Age	–	.005	.135	−.040	.095	.264**	.049	−.076	−.136	−.219*	−.095	.166	−.104	−.149	−.031	−.206*
2	Gender		–	.189*	.264**	.308**	.282**	.073	−.134	.051	.312**	.155	.097	.187*	.015	.208*	−.306**
3	Percent body fat			–	.785**	.741**	.777**	.403**	−.177	.202*	−.050	.415**	.338**	.382**	.298**	.109	−.215*
4	zBMI				–	.775**	.762**	.390**	−.293**	.213*	.156	.462**	.298**	.452**	.310**	.137	−.243*
5	Waist circumference					–	.831**	.261**	−.248*	.194*	.159	.433**	.363**	.420**	.282**	.244*	−.295**
6	Hip circumference						–	.416**	−.244*	.176	.156	.421**	.426**	.419**	.246*	.163	−.454**
7	Systolic blood pressure							–	−.082	.130	.084	.209*	.081	.204*	.156	.046	−.331**
8	HDL								–	.100	.053	−.137	−.174	−.116	−.505**	.017	.198*
9	LDL									–	.264**	.099	.004	.109	.253**	−.003	−.207*
10	Blood glucose										–	.296**	.026	.375**	.125	.271**	−.285**
11	Plasma insulin											–	.406**	.988**	.372**	.273**	−.154
12	Proinsulin												–	.395**	.238*	.086	−.235*
13	HOMA													–	.367**	.286**	−.172
14	Triglyceride														–	−.045	−.055
15	Sedentary behavior															–	−.410**
16	MVPA																–

Gender (0, female, 1 male).

zBMI, body mass index (Z-scores); SBP, systolic blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein; HOMA, homeostatic model assessment.

*p < .05.

**p < .01.

Increased sedentary behavior was associated with greater waist circumference (r_s = .24, p < .05) and greater metabolic risks, i.e., higher fasting blood glucose (r_s = .27, p < .01), plasma insulin (r_s = .27, p < .01), and HOMA-IR (r_s = .29, p < .01). No significant associations were observed for the relationship between sedentary behavior and cardiac risk factors including SBP and lipids. Levels of MVPA, however, had significant associations with several anthropometric, cardiac and metabolic risk factors. Increased levels of MVPA were associated with decreased percent body fat (r_s = −.22, p < .05), zBMI (r_s = −.24, p < .05), waist circumference (r_s = −.30, p < .01), and hip circumference (r_s = −.45, p < .01). MVPA also showed significant associations with lower cardiac risk factors: increased levels of MVPA were associated with lower SBP (r_s = −.35, p < .01) and LDL (r_s = −.21, p < .05), and with higher HDL (r_s = .20, p < .05). Further, higher MVPA was associated with lower metabolic risk factors including fasting blood glucose (r_s = −.29, p < .01), proinsulin (r_s = −.24, p < .05), and HOMA-IR (r_s = −.41, p < .01).

The main hypotheses of the study were tested using structural equation modeling in MPlus version 8.0 [50]. Three latent variables were derived from assessments of anthropometric, cardiac, and metabolic measures. The anthropometric health risk dimension included percent body fat, zBMI, waist circumference, and hip circumference. All indicators of the latent variable for anthropometric health risk had excellent loadings (λ ranging = .83 to .95). The latent variable for cardiac health risk was derived from SBP, HDL, and LDL. HDL, being a positive indicator, had a negative loading (λ = −.28) on the latent cardiac health risk, while LDL (λ = .38) and SBP (λ = .47) positively contributed to the latent variable. Finally, the latent variable for metabolic health risk constituted fasting blood glucose (λ = .30), proinsulin (λ = .53), and HOMA-IR (λ = .69). Plasma insulin had a very high correlation with HOMA (r_s = .99, p < .01) and so was excluded from the latent variable model to avoid multicollinearity. As preliminary analysis showed that triglycerides appeared to have no correlation with either MVPA or sedentary behavior in our data, it was also excluded from the latent variable moderation model.

To test moderation by sedentary behavior, both physical activity variables, i.e., MVPA and sedentary behavior, were mean-centered and multiplied to generate the interaction term. Along with the predictor MVPA and moderator sedentary behavior, the interaction term was also used as an independent variable to predict the three latent health risk dimensions. Given the high values of skewness and kurtosis, the model was estimated using MLM estimator. MLM–maximum likelihood parameter estimates with standard errors and a mean-adjusted chi-square test statistic were used that are robust to non-normality.

The results showed promising findings for the prediction of the three latent health risk dimensions by MVPA and moderation by sedentary behavior, but the latent variable moderation model showed a poor fit to the data with Satorra–Bentler Chi-square = 117.763, df = 53, CFI = .886, TLI = .839, RMSEA = .135 (p < .01), and SRMR = .072. The poor model fit was likely due to covariance among the anthropometric measures of health risk. Exploration of the modification index showed that the residual of zBMI had covariation with percent body fat, HDL, and SBP percentile, which further correlated with waist circumference. Similarly, the residual of fasting blood glucose covaried with percent body fat, HDL, and HOMA-IR. Covariation also existed between the residuals of HDL and LDL. Addition of these covariances resulted in significant improvement of the latent variable moderation model with ∆χ²(df) = 49.742(8), p < .01. The final model showed a very good fit of the model to the data with Satorra–Bentler Chi-square = 68.021, df = 45, CFI = .959, TLI = .932, RMSEA = .087 (p = .087), and SRMR = .053.

Results presented in Fig. 1 showed that MVPA negatively predicted anthropometric health risk (β = −.58, p < .01), cardiac health risk (β = −.89, p < .01), and metabolic risk (β = −.88, p < .05). Although sedentary behavior did not significantly predict any of the health risk dimensions (p > .05), it positively moderated the effect of MVPA on all three dimensions of health risk (β interactions: .51, .67, and .76; p < .05, respectively, for anthropometric, cardiac, and metabolic health risks). The moderating effect of sedentary behavior is visually presented in Figs 2–4. The figures show a similar pattern of moderation by sedentary behavior on all three latent dimensions of health risks. A steeper slope for the effect of MVPA on all three dimensions of health risks was observed when sedentary behavior was low compared to when it was high.

Fig 1 |.

Latent variable moderation model to test moderating effect of sedentary behavior in predicting anthropometric, cardiac, and metabolic risk by MVPA.

Fig 2 |.

Interaction graph showing moderating role of sedentary behavior for the relationship between MVPA and anthropometric health.

Fig 4 |.

Interaction graph showing moderating role of sedentary behavior for the relationship between MVPA and metabolic health risk.

Fig 3 |.

Interaction graph showing moderating role of sedentary behavior for the relationship between MVPA and cardiac health risk.

DISCUSSION

The findings from this study extend our understanding of the role of MVPA and sedentary behavior in relation to cardiometabolic health risk factors, and demonstrates these effects to exist in a high-risk, under-studied sample of young Latino children. Examining these factors in young Latino children is important as the rates of overweight and obesity in Latino children are substantially higher than non-Latino white children [1]. Objective measurement of MVPA and sedentary behaviors has been increasingly used in health behavior research in recent years and has provided evidence of the effectiveness of MVPA for health promotion [13, 51].

Consistent with previous research findings, the results of the present study also demonstrated the significant relationships between MVPA and measures of anthropometric, cardiac, and metabolic health risk indicators, and extended these findings to young Latino children in the age range of 4–7 years. Significant associations of MVPA with the anthropometric health indicator showed lower zBMI, percent body fat, and both waist and hip circumference with greater MVPA. Similar findings have been shown in the literature for these health indicators in pediatric and adult samples for BMI [14, 34, 35, 42, 52], percent body fat [26, 35, 53], metabolic syndrome [54], and waist circumference [14, 22, 23, 35, 38, 52, 55].

Our results demonstrating the effects of MVPA in reducing cardiac risk factors are also supported by previous studies in children [33, 34, 39]. In contrast to a recent study conducted with adults with prediabetes [55], our results confirmed that increased MVPA is associated with significantly lower SBP [14, 22, 23, 32, 55–57]; however, one study with adults did not find a significant relationship between MVPA and both HDL and LDL [14]. Our results are also consistent with others showing that MVPA is associated with higher levels of HDL [26, 28, 30, 38, 55, 57, 58] and lower LDL [26] in both adults and children. Furthermore, our findings are consistent with research indicating that greater MVPA is associated with significant reductions in metabolic risk factors including fasting blood glucose [10, 14, 22, 28, 32, 55, 57], insulin and HOMA [36, 44, 55].

There are inconsistent findings with respect to the role of sedentary behavior in increasing health risk factors [22, 23, 53, 59, 60]. Our results are novel given that few studies have examined this relationship in young ethnic minority children. Importantly, our findings are also consistent with those showing positive relationships between sedentary behavior and waist circumference in adults [28] and children [38], as well as body mass index in children [42]. Compliance with behavioral guidelines to decrease sedentary behavior has favorable effects on decreasing metabolic risk in overweight and obese youth [4]. Our results also showed that among the three dimensions of health risk, sedentary behavior significantly increased all three indicators of metabolic risk including fasting blood glucose, proinsulin, and HOMA-IR. Similar findings were reported in a study of adults at high risk for type 2 diabetes showing associations of sedentary time with 2-hr plasma glucose, triacylglycerol and HDL-cholesterol, after adjustment for MVPA and BMI [52]. Our findings are also consistent with a recent study showing that deleterious levels of HDL-cholesterol, triglycerides, insulin resistance, and plasminogen activator inhibitor-1 were all associated with higher levels of sedentary behavior in Latino youth in the age range of 8–16 years [10].

A major innovation of the present study is that it provided empirical evidence for the role of MVPA in decreasing cardiometabolic risks of young Latino children using a latent variable model. In general, there is consensus on the positive role of MVPA with regard to cardiometabolic risk, but some studies have shown improvement in some indicators while others have shown improvements in other indicators. A latent variable analysis, as we have used, has the potential to identify generalized stable improvements rather than being specific for each indicator which may not be consistent across studies mainly due to the variant nature and type of activities considered as MVPA. Variation in physical activities may affect variant indicators resulting in inconsistency of findings, yet a stable effect may be visible in a more generic measure (i.e., a latent dimension or construct). Our results from the latent variable moderation model showing that MVPA has a significant role in decreasing anthropometric, cardiac, and metabolic risks, particularly when sedentary behavior is low, are supported by the findings of the studies reporting the effects of MVPA on either few or most indicators.

These findings extend the research literature particularly on the role of sedentary behavior. We assumed that underestimation of the role of sedentary behavior may be due to the overwhelming significance of MVPA. These assumptions were supported by showing a change in the relationship between MVPA and cardiometabolic risk factors under varying levels of sedentary behaviors [39]. A study using isotemporal substitution analysis to test the theoretical replacement of sedentary behavior with MVPA showed that substitution of sedentary time with MVPA resulted in decreased cardiometabolic risk [31]. Hence, in the current study we hypothesized that the role of sedentary behavior may be better explained through its interaction with MVPA.

Our results from the latent variable moderation model confirmed our prediction showing that sedentary behavior acts as a moderator for the relationship between MVPA and anthropometric, cardiac, and metabolic risk factors. The moderating effect of sedentary behavior on MVPA showed a similar pattern on all three latent dimensions of health risks. An increase in the effect of MVPA on decreasing anthropometric, cardiac, and metabolic risks is observed when sedentary behavior is low compared to when sedentary behavior is high. These findings support the concept that both children and adults should be encouraged to decrease sedentary behavior in order to benefit more from their physical activity [61], and also have important policy implications: more efforts should be made to reduce sedentary behavior in order to improve public health.

This study was a correlational cross-sectional study design and therefore future research is needed to determine the generalizability of these findings. Study limitations also include the relatively small sample, with missing data from 15% of the sample (who did not return the accelerometry device or did not provide usable accelerometry data). Despite the small sample and limited statistical power, we were able to demonstrate significant relationships among physical activity, sedentary behavior, and cardiometabolic risk that supported our hypothesis of the moderating role of sedentary behavior; the strong measurement model provides confidence in the study findings. The fact that these health behaviors had negative health effects in young children is also significant, particularly given that there is limited research examining these factors in Latino children [62]. Further, our study sample was community-based and represented a small proportion of potentially available children at the schools; therefore, there is some possibility of the sample being biased in some ways. Research using longitudinal study designs and larger multi-ethnic samples of children will help to explicate our understanding of the role of sedentary behavior as a moderator of the relationship between MVPA and cardiometabolic risk in young children.

CONCLUSIONS

In young Latino children in the age range of 4–7 years, increased MVPA is associated with decreased cardiometabolic risk, particularly when sedentary behavior is low. These findings support the idea that children should not only engage in MVPA, but also be encouraged to decrease sedentary behavior in order to benefit more from their physical activity. Health policies for children should continue to emphasize both increasing MVPA and decreasing sedentary behavior.

Funding: This research was supported by grant #1R34 DK074552 from the National Institutes of Health to the senior author. The first author was supported by grant # 21-6/HEC/R&D/PPCR/2017.

Compliance With Ethical Standards

Authors’ Statement of Conflict of Interest and Adherence to Ethical Standards: The authors declare that they have no conflicts of interest.

Ethical Approval: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. These procedures were approved by the Senior Author’s University IRB before any data collection took place. This article does not contain any studies with animals performed by any of the authors.

Informed Consent: Informed consent was obtained from all individual participants included in the study.

Transparency Statements: This study was not formally registered. The analysis plan was not formally pre-registered. De-identified data from this study are not available in a public archive.

Data Availability: De-identified data from this study will be made available (as allowable according to institutional IRB standards) by emailing the corresponding author. Analytic code used to conduct the analyses presented in this study are not available in a public archive. They may be available by emailing the corresponding author. Materials used to conduct the study are not publically available.