Intake of Disease-Related Nutrients Among Baby Boomers
Intake of Disease-Related Nutrients Among Baby Boomers
We analyzed data from the National Health and Nutrition Examination Survey (NHANES), including the NHANES for 1988–1994 (NHANES III) and the NHANES for 2007–2010 (NHANES 2010), focusing on respondents who were in the baby boom age group during either period, and the two cohorts were compared by dietary intake of key nutritional components.
The NHANES is a national survey series that began more than 40 years ago and is sponsored by the Centers for Disease Control and Prevention (CDC). The NHANES is conducted continuously across the nation in select locations using a series of multistage surveys in a complex sampling design to assess the health and nutritional status of the civilian noninstitutionalized population of the United States. It includes demographic, socioeconomic, dietary, and health-related information of survey respondents. Although both the NHANES III and the NHANES 2010 were designed to address the same aspects of the US population, they were conducted more than a decade apart and some of the survey questions differed. For the purposes of the present study, the variables chosen for analysis were as similar as possible in each cohort, with some recoding of variables in the surveys as necessary.
Demographic characteristics including age, sex, race, socioeconomic status (SES), education level, and ratio of family income to poverty level were collected. Race was defined in four categories: non-Hispanic white, non-Hispanic black, Mexican American, and other. Education was transformed into a dichotomous variable of ≤11 years and >11 years of education completed. Body measurement variables such as height, weight, and body mass index (BMI) also were obtained. BMI categories were underweight (<18.5), normal weight (≥18.5 and ≤24.9), overweight (≥25 and ≤29.9), and obese (≥30), based on the CDC categories.
Of two 24-hour dietary recalls collected in the NHANES 2010, only the first 24-hour recall interview was used for the purpose of comparison with the NHANES III, which collected only 1 day of dietary recall for respondents. Most of the dietary intake variables, including total energy, total fat, vitamin C, potassium, magnesium, carbohydrate, protein, calcium, potassium, sodium, cholesterol, and fiber, were extracted directly from the total nutrient intake of 1-day recall datasets (not including supplement intake). Variables such as vegetables, fruit, milk, and milk products were calculated by using the first digit of the NHANES DRXFDCD food code (eg, "1" for milk and milk products, "6" for fruits, "7" for vegetables) and adding the amount of the same kinds of food together in the datasets of individual foods of 1-day recall for each respondent. For water intake, different questions were asked in the NHANES III and the NHANES 2010; only the questions regarding plain water intake were used in the present study.
Statistical analysis software (SAS version 9.3, SAS Institute, Cary, NC) was used for analysis in this study. To determine the appropriate sample weight to account for the complex survey design (including oversampling), survey nonresponse, and poststratification, the program specialist at the National Center for Health Statistics was consulted. All of the estimates for NHANES III were weighted using 6-year weights from the mobile examination center examination weights. For NHANES 2010, the dietary day 1 weights were used without constructing 4-year dietary weights because the same population was used to poststratify the sample weights. To correct for the complex sampling between the two NHANES periods, the CDC program specialists recommended the creation of a new strata variable by adding 49 to the original strata variable "sdmvstra" in the NHANES 2010 sample.
Descriptive analysis was conducted for demographic characteristics of the respondents. Differences in dietary intake, including total energy, total fat, protein, fiber, carbohydrate, vitamin C, calcium, cholesterol, sodium, potassium, water, fruit, vegetables, milk, and milk products, between the two NHANES eras were determined using procedure SURVEYMEANS and procedure SURVEYREG. Based on the finding of existing differences, a regression model was constructed to predict whether current baby boomers had an intake of key nutrients different from the previous generation of middle-aged adults, controlling for possible confounding factors, including age, race, sex, and SES. Poverty ratio, a proportion of the population that meets the definition of poverty, was used to represent SES.
One day of a 24-hour dietary recall was used for data collection and was considered important to address the possibility of overreporting or underreporting dietary intake by respondents. The revised Goldberg method was applied to evaluate overreporting or underreporting of dietary intake in both NHANES cohorts based on physical activity level (PAL) and compared with the ratio of self-reported energy intake to basal metabolic rate (rEI:BMR). BMR was calculated using the Mifflin equation, based on age, sex, height, and weight. PAL in the NHANES 2010 was determined based on answers to three questions regarding recreational activities in the past week: walking or bicycling, vigorous recreational activities, and moderate recreational activities.
For the NHANES III, PAL was determined in terms of the sum of all of the answers to the questions regarding respondents' recreational activities in the past month divided by 4.3 (average times per week, 22). Overreporters and underreporters were calculated based on correspondence of the reported rEI:BMR ratios with each subject's reported PAL, taking into account coefficients of variation in intakes and other components of energy balance. Lower and upper confidence interval (CI) limits were calculated for rEI:BMR ratios using plausible upper and lower limits based on the World Health Organization standard rEI:BMR ratios of 1.55 for light physical activity (women and men), 1.64/1.78 for moderate activity (women and men), and 1.82/2.10 for individuals with high physical activity (women and men); standard deviation levels used were −2/+2 for 95% CI limits.
After the lower and upper limit calculations were obtained for individuals in the study population, we compared their rEI/BMR ratios to the World Health Organization standard lower and upper confidence limits based on sex and PAL level. We reported as a underreporter if an individual rEI/BMR ratio was less than the lower limit and as an overreporter if an individual rEI/BMR ratio was greater than the upper limit. The mean for lower limit for rEI:BMR was 0.99 (95% CI 0.98–0.99) and for upper limit was 3.11 (95% CI 3.10–3.12). Individual respondents were classified based on their own lower and upper limits. We found that 38.8% of respondents were underreporters and <1% (0.87%) were overreporters.
Methods
We analyzed data from the National Health and Nutrition Examination Survey (NHANES), including the NHANES for 1988–1994 (NHANES III) and the NHANES for 2007–2010 (NHANES 2010), focusing on respondents who were in the baby boom age group during either period, and the two cohorts were compared by dietary intake of key nutritional components.
The NHANES is a national survey series that began more than 40 years ago and is sponsored by the Centers for Disease Control and Prevention (CDC). The NHANES is conducted continuously across the nation in select locations using a series of multistage surveys in a complex sampling design to assess the health and nutritional status of the civilian noninstitutionalized population of the United States. It includes demographic, socioeconomic, dietary, and health-related information of survey respondents. Although both the NHANES III and the NHANES 2010 were designed to address the same aspects of the US population, they were conducted more than a decade apart and some of the survey questions differed. For the purposes of the present study, the variables chosen for analysis were as similar as possible in each cohort, with some recoding of variables in the surveys as necessary.
Demographics and Body Measurement Variables
Demographic characteristics including age, sex, race, socioeconomic status (SES), education level, and ratio of family income to poverty level were collected. Race was defined in four categories: non-Hispanic white, non-Hispanic black, Mexican American, and other. Education was transformed into a dichotomous variable of ≤11 years and >11 years of education completed. Body measurement variables such as height, weight, and body mass index (BMI) also were obtained. BMI categories were underweight (<18.5), normal weight (≥18.5 and ≤24.9), overweight (≥25 and ≤29.9), and obese (≥30), based on the CDC categories.
Dietary Intake Variables
Of two 24-hour dietary recalls collected in the NHANES 2010, only the first 24-hour recall interview was used for the purpose of comparison with the NHANES III, which collected only 1 day of dietary recall for respondents. Most of the dietary intake variables, including total energy, total fat, vitamin C, potassium, magnesium, carbohydrate, protein, calcium, potassium, sodium, cholesterol, and fiber, were extracted directly from the total nutrient intake of 1-day recall datasets (not including supplement intake). Variables such as vegetables, fruit, milk, and milk products were calculated by using the first digit of the NHANES DRXFDCD food code (eg, "1" for milk and milk products, "6" for fruits, "7" for vegetables) and adding the amount of the same kinds of food together in the datasets of individual foods of 1-day recall for each respondent. For water intake, different questions were asked in the NHANES III and the NHANES 2010; only the questions regarding plain water intake were used in the present study.
Statistical Analysis
Statistical analysis software (SAS version 9.3, SAS Institute, Cary, NC) was used for analysis in this study. To determine the appropriate sample weight to account for the complex survey design (including oversampling), survey nonresponse, and poststratification, the program specialist at the National Center for Health Statistics was consulted. All of the estimates for NHANES III were weighted using 6-year weights from the mobile examination center examination weights. For NHANES 2010, the dietary day 1 weights were used without constructing 4-year dietary weights because the same population was used to poststratify the sample weights. To correct for the complex sampling between the two NHANES periods, the CDC program specialists recommended the creation of a new strata variable by adding 49 to the original strata variable "sdmvstra" in the NHANES 2010 sample.
Descriptive analysis was conducted for demographic characteristics of the respondents. Differences in dietary intake, including total energy, total fat, protein, fiber, carbohydrate, vitamin C, calcium, cholesterol, sodium, potassium, water, fruit, vegetables, milk, and milk products, between the two NHANES eras were determined using procedure SURVEYMEANS and procedure SURVEYREG. Based on the finding of existing differences, a regression model was constructed to predict whether current baby boomers had an intake of key nutrients different from the previous generation of middle-aged adults, controlling for possible confounding factors, including age, race, sex, and SES. Poverty ratio, a proportion of the population that meets the definition of poverty, was used to represent SES.
One day of a 24-hour dietary recall was used for data collection and was considered important to address the possibility of overreporting or underreporting dietary intake by respondents. The revised Goldberg method was applied to evaluate overreporting or underreporting of dietary intake in both NHANES cohorts based on physical activity level (PAL) and compared with the ratio of self-reported energy intake to basal metabolic rate (rEI:BMR). BMR was calculated using the Mifflin equation, based on age, sex, height, and weight. PAL in the NHANES 2010 was determined based on answers to three questions regarding recreational activities in the past week: walking or bicycling, vigorous recreational activities, and moderate recreational activities.
For the NHANES III, PAL was determined in terms of the sum of all of the answers to the questions regarding respondents' recreational activities in the past month divided by 4.3 (average times per week, 22). Overreporters and underreporters were calculated based on correspondence of the reported rEI:BMR ratios with each subject's reported PAL, taking into account coefficients of variation in intakes and other components of energy balance. Lower and upper confidence interval (CI) limits were calculated for rEI:BMR ratios using plausible upper and lower limits based on the World Health Organization standard rEI:BMR ratios of 1.55 for light physical activity (women and men), 1.64/1.78 for moderate activity (women and men), and 1.82/2.10 for individuals with high physical activity (women and men); standard deviation levels used were −2/+2 for 95% CI limits.
After the lower and upper limit calculations were obtained for individuals in the study population, we compared their rEI/BMR ratios to the World Health Organization standard lower and upper confidence limits based on sex and PAL level. We reported as a underreporter if an individual rEI/BMR ratio was less than the lower limit and as an overreporter if an individual rEI/BMR ratio was greater than the upper limit. The mean for lower limit for rEI:BMR was 0.99 (95% CI 0.98–0.99) and for upper limit was 3.11 (95% CI 3.10–3.12). Individual respondents were classified based on their own lower and upper limits. We found that 38.8% of respondents were underreporters and <1% (0.87%) were overreporters.
Source...