Associations in Physical Activity, Diet Quality, and Weight
Associations in Physical Activity, Diet Quality, and Weight
The major finding of this study was that participation in MVPA was very consistently inversely related to measures of weight status in US adults. This pattern was observed in both men and women in the analyses performed within age groups and in the analyses examining trends across the age groups. With great consistency, higher levels of physical activity were associated with lower BMI and lower waist circumference. These relations were largely unaffected by adjustments for race/ethnicity, socioeconomic status, smoking, age (within age categories), and diet quality. These findings, although limited by the cross-sectional nature of the NHANES data, provide support for efforts to prevent obesity through promotion of increased physical activity in American adults.
The findings for relations between diet quality and weight status were much less consistent than for physical activity. Better diet quality was significantly associated with lower BMI in three of six age groups for men and with lower waist circumference in two of six age groups for men. Better diet quality was significantly associated with lower BMI and waist circumference in only one age group for women (50–59 yr). The inconsistency of these findings may be explained, in part, by the limitations associated with measurement of dietary behavior in large surveys. In NHANES, diet was measured via interviewer-administered diet recall for two 24-h periods. We used this self-reported dietary data to calculate the HEI-2005, which we used as the measure of overall diet quality. Although this diet assessment approach is well established and widely applied, and the HEI is a validated approach to measuring dietary quality, having three or more dietary recalls would have been preferable. Measurement error research is still evolving for the study of dietary indices and ratio measures such as the HEI. What is known to date suggests that the energy adjustment built into the HEI may be countering some of the attenuating effects of nondifferential misclassification. However, although the statistical significance tests of the association may not be seriously biased, they most likely do not have adequate statistical power. Thus, the nonsignificant findings for the associations of diet quality and weight status should be treated with due caution.
The design and methodology of the present study are unique, but the findings are generally consistent with the existing literature examining relations among physical activity, diet quality, and measures of weight status in adults. We observed, on a very consistent basis, significant inverse relations between physical activity and the two measures of weight status considered in this study. This observation is consistent with previous studies that have examined similar relations using cross-sectional study designs in large samples. It is also consistent with the findings of studies using prospective, observational designs to examine the relation between physical activity and measures of weight status in adults. We found less consistency in terms of statistical significance of findings in our examination of the relation between diet quality and weight status measures. In the current study, higher diet quality was inversely associated with measures of weight status in three of six age groups for men and only one of six age groups for women. Previous studies have reported significant inverse relations between diet quality and weight change or body composition in adults. However, these studies pooled participants across age groups and reported marked gender differences.
A recently published study by Loprinzi et al. using NHANES data found that both physical activity and diet quality were associated with risk of overweight/obesity in US adults. However, the study design and analytic strategies used in the current study are very different from those used in the Loprinzi study. The current study examined these relations in age and gender subgroups, and the variables were treated continuously. We have examined our exposure variables in the full analysis sample (adjusting for age, gender, and other covariates), and like Loprinzi and colleagues, we found significant inverse relations between both physical activity and diet quality with weight status. However, the current study expands and complements that of Loprinzi et al. by providing detailed information about how these variables interact within different subgroups of the US population.
It is well known that US men and women gain weight and fat as they age. However, to our knowledge, this is the first study to comprehensively document age-related trends in weight status, physical activity, and diet quality using a representative sample of US men and women. The current investigation found that BMI and waist circumference tended to be higher in the older cohorts of adults, which is supported by previous studies of weight status in adult populations. In distinct contrast, MVPA was strikingly lower in the older age groups for both men and women. In both genders, MVPA in adults ≥70 yr was only 20%–25% of that observed in the 20- to 29-yr age groups. Previous studies have documented similar differences in physical activity by age groupings. Interestingly, men and women reported higher diet quality in the older age groups. This same trend was observed in an earlier round of NHANES These observations suggest that, at the population level, age-related increases in BMI and waist circumference are likely to be explained by the age-related decline in physical activity.
The present study stands alone as the first study to have examined the relations among physical activity, diet quality, and measures of weight status in nationally representative age/gender subgroups of US adults. Because of the cross-sectional study design, we were unable to determine directionality of these relations. However, the fact that these relations were studied across the adult lifespan in racially and ethnically diverse samples of men and women is a major strength of the study. Other strengths include use of an objective measure of physical activity and application of an accepted index of diet quality, although it is acknowledged that both measures have limitations. Accelerometry underdetects some forms of physical activity, and the index of diet quality is based on self-reported information. Furthermore, the physical activity measure used here does not measure energy expenditure, just as the diet quality index does not measure energy intake. We cannot exclude the possibility of selection bias given the number of participants excluded because of missingness in the objectively measured physical activity and 24-h dietary recall variables. Furthermore, the parameter estimates, although considered to be statistically significant, were small. Within the limitations of the study methodology, we conclude that higher levels of physical activity are very consistently associated with more favorable weight status in US adults. We believe that these findings provide support for public health efforts to prevent obesity by promoting increased physical activity in adult Americans.
Discussion
The major finding of this study was that participation in MVPA was very consistently inversely related to measures of weight status in US adults. This pattern was observed in both men and women in the analyses performed within age groups and in the analyses examining trends across the age groups. With great consistency, higher levels of physical activity were associated with lower BMI and lower waist circumference. These relations were largely unaffected by adjustments for race/ethnicity, socioeconomic status, smoking, age (within age categories), and diet quality. These findings, although limited by the cross-sectional nature of the NHANES data, provide support for efforts to prevent obesity through promotion of increased physical activity in American adults.
The findings for relations between diet quality and weight status were much less consistent than for physical activity. Better diet quality was significantly associated with lower BMI in three of six age groups for men and with lower waist circumference in two of six age groups for men. Better diet quality was significantly associated with lower BMI and waist circumference in only one age group for women (50–59 yr). The inconsistency of these findings may be explained, in part, by the limitations associated with measurement of dietary behavior in large surveys. In NHANES, diet was measured via interviewer-administered diet recall for two 24-h periods. We used this self-reported dietary data to calculate the HEI-2005, which we used as the measure of overall diet quality. Although this diet assessment approach is well established and widely applied, and the HEI is a validated approach to measuring dietary quality, having three or more dietary recalls would have been preferable. Measurement error research is still evolving for the study of dietary indices and ratio measures such as the HEI. What is known to date suggests that the energy adjustment built into the HEI may be countering some of the attenuating effects of nondifferential misclassification. However, although the statistical significance tests of the association may not be seriously biased, they most likely do not have adequate statistical power. Thus, the nonsignificant findings for the associations of diet quality and weight status should be treated with due caution.
The design and methodology of the present study are unique, but the findings are generally consistent with the existing literature examining relations among physical activity, diet quality, and measures of weight status in adults. We observed, on a very consistent basis, significant inverse relations between physical activity and the two measures of weight status considered in this study. This observation is consistent with previous studies that have examined similar relations using cross-sectional study designs in large samples. It is also consistent with the findings of studies using prospective, observational designs to examine the relation between physical activity and measures of weight status in adults. We found less consistency in terms of statistical significance of findings in our examination of the relation between diet quality and weight status measures. In the current study, higher diet quality was inversely associated with measures of weight status in three of six age groups for men and only one of six age groups for women. Previous studies have reported significant inverse relations between diet quality and weight change or body composition in adults. However, these studies pooled participants across age groups and reported marked gender differences.
A recently published study by Loprinzi et al. using NHANES data found that both physical activity and diet quality were associated with risk of overweight/obesity in US adults. However, the study design and analytic strategies used in the current study are very different from those used in the Loprinzi study. The current study examined these relations in age and gender subgroups, and the variables were treated continuously. We have examined our exposure variables in the full analysis sample (adjusting for age, gender, and other covariates), and like Loprinzi and colleagues, we found significant inverse relations between both physical activity and diet quality with weight status. However, the current study expands and complements that of Loprinzi et al. by providing detailed information about how these variables interact within different subgroups of the US population.
It is well known that US men and women gain weight and fat as they age. However, to our knowledge, this is the first study to comprehensively document age-related trends in weight status, physical activity, and diet quality using a representative sample of US men and women. The current investigation found that BMI and waist circumference tended to be higher in the older cohorts of adults, which is supported by previous studies of weight status in adult populations. In distinct contrast, MVPA was strikingly lower in the older age groups for both men and women. In both genders, MVPA in adults ≥70 yr was only 20%–25% of that observed in the 20- to 29-yr age groups. Previous studies have documented similar differences in physical activity by age groupings. Interestingly, men and women reported higher diet quality in the older age groups. This same trend was observed in an earlier round of NHANES These observations suggest that, at the population level, age-related increases in BMI and waist circumference are likely to be explained by the age-related decline in physical activity.
The present study stands alone as the first study to have examined the relations among physical activity, diet quality, and measures of weight status in nationally representative age/gender subgroups of US adults. Because of the cross-sectional study design, we were unable to determine directionality of these relations. However, the fact that these relations were studied across the adult lifespan in racially and ethnically diverse samples of men and women is a major strength of the study. Other strengths include use of an objective measure of physical activity and application of an accepted index of diet quality, although it is acknowledged that both measures have limitations. Accelerometry underdetects some forms of physical activity, and the index of diet quality is based on self-reported information. Furthermore, the physical activity measure used here does not measure energy expenditure, just as the diet quality index does not measure energy intake. We cannot exclude the possibility of selection bias given the number of participants excluded because of missingness in the objectively measured physical activity and 24-h dietary recall variables. Furthermore, the parameter estimates, although considered to be statistically significant, were small. Within the limitations of the study methodology, we conclude that higher levels of physical activity are very consistently associated with more favorable weight status in US adults. We believe that these findings provide support for public health efforts to prevent obesity by promoting increased physical activity in adult Americans.