Abstract
Over the past decade, blood pressure and sodium intake declined among children and adolescents (ie, youths) in the United States. We updated temporal trends and determined if secular changes in blood pressure might be partly associated with usual sodium intake. We included 12 249 youths aged 8 to 17 years who participated in the National Health and Nutrition Examination Survey from 2003 to 2016 and had blood pressure and dietary data. Logistic regression was used to describe secular trends and the association between usual sodium intake and blood pressure categorized according to 2017 Hypertension Guidelines. The prevalence of youths with combined elevated blood pressure/hypertension (ie, either elevated blood pressure or hypertension) significantly declined from 16.2% in 2003-2004 to 13.3% in 2015-2016 (P<0.001 for trend), as did hypertension from 6.6% to 4.9% (P=0.005 for trend). Across the same time period, mean usual sodium intake decreased from 3381 to 3208 mg/day (P<0.001 for trend). Holding constant survey cycle, sex, age, race and Hispanic origin, and weight status, the adjusted odds ratio per 1000 mg/day of usual sodium intake for elevated blood pressure/hypertension was 1.18 (95% CI, 1.03–1.35) and for hypertension was 1.20 (95% CI, 0.96–1.50). From 2003 to 2016, blood pressure and usual sodium intake declined among youths. Although 1000 mg/day higher usual sodium intake was associated with ≈20% higher odds of elevated blood pressure/hypertension and hypertension, the association with hypertension was not statistically significant.
Keywords: blood pressure, diet, hypertension, prevalence, sodium
Hypertension is a well-known risk factor for heart disease and stroke, which combined account for more deaths each year than any other major cause of death in the United States.1 Although hypertension in children and adolescents (hereafter referred to as youths) is less common than in adults, evidence suggests higher blood pressure (BP) levels in childhood are associated with higher BP levels later in life.2 Previous research has demonstrated a linear, dose-dependent relationship between sodium intake and BP in adults and that sodium intake is associated with BP in children.3 Over the period of 1999 to 2012, evidence has suggested both sodium intake4 and BP5,6 declined slightly among US youths. Whether current temporal trends in BP (up to 2016) might be associated with concurrent changes in population sodium intake is unclear.
Though several studies described temporal changes in BP among US youths,5-7 few examined concurrent trends in sodium intake.7,8 To our knowledge, only one examined temporal trends in BP before and after adjusting for sodium intake.7 In this study, sodium intake was adjusted for energy intake (both estimated from a single 24-hour dietary recall) and differences in BP were examined between 1988-1994 and 1999-2008.7 The use of a single 24-hour recall does not represent usual intake and may have attenuated the true association between sodium and BP because of the high, within-person, day-to-day variability in foods consumed.8 Furthermore, evidence suggests that although BP increased between 1988-1994 and 1999-2008, declines occurred from 2007 to 2012.5,6 We hypothesized that current temporal trends in BP among US youths might be partly associated with shifts in the distribution of usual sodium intake.
The objectives of this study were to describe among US youths aged 8 to 17 years (1) temporal trends in estimated usual sodium intake from 2003 through 2016, (2) concurrent temporal trends in elevated BP (EBP) or hypertension, and (3) to determine if temporal trends in EBP or hypertension were correlated with usual dietary sodium intake.
Methods
We used data from the National Health and Nutrition Examination Survey (NHANES), a nationally representative, multistage survey of the noninstitutionalized US civilian population.9 All participants provided written informed consent (parental consent was obtained for those <18 years) and NHANES was approved by the National Center for Health Statistics’ Ethics Review Board. Detailed information on NHANES data collection and survey procedures used in this analysis are publically available and can be found elsewhere.10,11 For our analyses, we used data collected from youths aged 8 to 17 years, publically released in 2-year cycles from 2003 to 2004 through 2015 to 2016. Exclusion criteria and additional methods are available in the online-only Data Supplement (Figure S1 and Methods in the online-only Data Supplement, respectively). For the purposes of reproducing this analysis, the analytic code used in this study is available from the corresponding author on request.
Measures
Dietary intake was assessed using up to two 24-hour dietary recalls collected using methods described previously.4 Nutrient values were assigned to foods and beverages using the United States Department of Agricultures’ Food and Nutrient Database for Dietary Studies corresponding with each 2-year cycle. Methods to estimate individual sodium intake changed over the time course of this study, but appropriate data files were used to ensure comparable methods for estimates of sodium intake from 2003 to 2016 (Methods in the online-only Data Supplement). Individual nutrient intake on each recall day was estimated by summing the amount of the specified nutrient consumed from each reported food and beverage for the previous 24-hour period. Total sodium intake excluded sodium from supplements, antacids, and salt added at the table. The National Cancer Institute’s method (Methods in the online-only Data Supplement), which accounts for within-person and day-to-day variability, was used to estimate usual daily nutrient intake for each individual in nonlinear mixed effects models. Using this method, we estimated usual daily sodium intake, sodium density, daily energy intake, daily potassium intake, and sodium-potassium ratio.
A certified examiner obtained up to 3 consecutive BP measurements from participants ≥8 years old using an appropriate cuff size and a mercury manometer, after the patient rested for 5 minutes in the seated position with feet flat on the floor.11 As described in the Methods (in the online-only Data Supplement), minor changes in BP measurement took place over the time course of the study and were not anticipated to affect estimates. Using the 2017 American Academy of Pediatrics guidelines, we classified participants according to their BP percentiles for age, sex, and height and then as having hypertension, EBP, or normal BP (Table S1).12 To coincide with previous reports, we combined participants classified as having EBP or hypertension into 1 group which we classified as having EBP/hypertension. We also characterized BP using the former 2004 National Heart Lung and Blood Institute guideline criteria13 to compare estimates with previous publications (Table S1).
Covariates
As potential covariates, we selected age, sex, race and Hispanic origin, and weight status a priori due to the association of these characteristics with BP. We examined trends among Mexican-Americans in accordance with analytic guidelines related to the sampling design.14 To assess weight status, we compared participants’ body mass index in kg/m2 to age- and sex-specific reference values from the 2000 Centers for Disease Control and Prevention growth charts.15 Under/normal weight was defined as body mass index for age/sex <85th percentile, overweight as ≥85th to <95th, and obesity as ≥95th. Several variables of interest (ie, self-reported frequency of salt added at the table, poverty-income ratio, and physical activity) were only available among a subset of included participants and were examined in sensitivity analyses (Tables S6 through S8).
Statistical Analyses
We examined frequencies and SE of selected covariates, mean (SE) of usual nutrient intake indicators, and the prevalence (SE) of EBP/hypertension and hypertension by 2-year survey cycle. Regression models are described in the online-only Data Supplement. Unadjusted and multiple logistic regression models with predicted marginals were used to examine frequencies/prevalence of sociodemographic variables and EBP/hypertension by 2-year survey cycle, whereas least squares linear regression models were used to examine mean usual nutrient intakes. We adjusted for age (years) when examining temporal trends in sociodemographic characteristics, weight status, EBP/hypertension, and hypertension, and for survey cycle, age (years), sex, race, and Hispanic origin, and weight status when examining temporal trends in usual sodium intake. Linear and quadratic trends were tested using orthogonal contrast matrices. Next, the associations between indicators of usual sodium intake and EBP/hypertension or hypertension were examined using logistic regression adjusted for survey cycle, sociodemographic characteristics, weight status, and estimated usual energy or potassium intake as applicable.16 Lastly, logistic regression models were used to examine temporal trends in EBP/hypertension or hypertension per survey cycle, successively adjusting for (1) sociodemographic characteristics, (2) weight status, and (3) indicators of usual sodium intake (including usual energy or potassium intake as applicable).
Statistical analyses were conducted using statistical software (SAS callable SUDAAN, version 11) to account for the complex survey design. Fourteen-year sample weights for NHANES 2003 to 2016 (constructed from day 1 dietary sample weights), were used to account for unequal sampling probabilities, nonresponse, noncoverage, and sample design. We calculated relative SE and according to recommendations: defined unstable estimates, which should be interpreted with caution, as 30%< relative SE ≤40% and suppressed estimates with relative SE >40%.14 Reported P values were not adjusted for multiple comparisons and are based on the Satterthwaite adjusted F test. All tests are 2 sided, and a P<0.05 was considered statistically significant.
Results
The final sample included 12 249 US youths with a mean age of 12.6 years. Compared with those excluded, higher proportions of included youths were 13 to 17 years, (51.5% versus 38.2%) white, non-Hispanic (27.9% versus 23.8%), and under/normal weight (61.0% versus 50.8%; P<0.001, overall difference by characteristic; Table S2). Among included youths, sex, age, and weight status distributions were similar from 2003 to 2004 through 2015 to 2016 (P>0.05 for temporal trend; Table 1, sample sizes in Table S3), but the proportion of white, non-Hispanic youths declined ≈11% (P=0.032 for trend).
Table 1.
Estimated Population Characteristics Among US Youths, NHANES 2003–2016 (N=12 249)
| Survey Cycles | 2003–2004 | 2005–2006 | 2007–2008 | 2009–2010 | 2011–2012 | 2013–2014 | 2015–2016 |
P Value Linear Trend* |
|---|---|---|---|---|---|---|---|---|
| Age, y, % (SE) | ||||||||
| 8–12 | 47.5 (2.3) | 48.3 (2.1) | 48.1 (2.4) | 49.6 (1.5) | 49.2 (1.6) | 49.5 (1.7) | 48.8 (1.7) | 0.494 |
| 13–17 | 52.5 (2.3) | 51.7 (2.1) | 51.9 (2.4) | 50.4 (1.5) | 50.8 (1.6) | 50.5 (1.7) | 51.2 (1.7) | |
| Sex, % (SE) | ||||||||
| Boys | 51.8 (1.8) | 51.2 (2.6) | 49.5 (2.0) | 48.5 (2.3) | 48.5 (2.4) | 51.6 (1.8) | 51.0 (1.8) | 0.798 |
| Girls | 48.2 (1.8) | 48.8 (2.6) | 50.5 (2.0) | 51.5 (2.3) | 51.5 (2.4) | 48.4 (1.8) | 49.0 (1.8) | |
| Race and Hispanic origin,† % (SE) | ||||||||
| White, NH | 63.9 (4.2) | 62.3 (4.2) | 60.4 (3.2) | 58.9 (4.3) | 54.3 (4.6) | 54.0 (5.8) | 52.5 (5.6) | 0.032 |
| Mexican-American | 12.1 (2.8) | 12.7 (1.2) | 13.2 (2.0) | 13.4 (2.8) | 15.0 (2.7) | 16.2 (3.7) | 15.3 (3.9) | 0.278 |
| Black, NH | 15.3 (2.4) | 14.2 (2.7) | 15.1 (1.8) | 13.5 (1.3) | 14.0 (2.9) | 13.3 (2.4) | 13.6 (3.4) | 0.600 |
| Weight status, % (SE) | ||||||||
| Under/normal | 64.1 (3.0) | 67.3 (3.0) | 62.9 (1.9) | 65.0 (1.4) | 62.5 (2.0) | 63.2 (2.5) | 61.9 (2.5) | 0.279 |
| Overweight | 16.9 (1.1) | 15.8 (1.2) | 17.3 (0.7) | 16.6 (0.7) | 17.4 (0.8) | 17.2 (0.9) | 17.6 (0.8) | 0.272 |
| Obese | 19.0 (1.9) | 16.9 (1.9) | 19.8 (1.5) | 18.4 (1.9) | 20.1 (1.4) | 19.6 (1.8) | 20.5 (1.8) | 0.283 |
The unweighted sample size for each subgroup can be found in Table S3. NH indicates non-Hispanic; NHANES, National Health and Nutrition Examination Survey.
P values for quadratic trends were nonsignificant and are not reported.
Data on other, multiracial is not presented.
Among youths, usual mean daily sodium intake declined from 3381 mg in 2003 to 2004 to 3208 mg in 2015 to 2016 (Table 2). Holding demographic characteristics and weight status constant, usual mean daily sodium intake declined, on average, by 35 mg per 2-year survey cycle (P<0.001 for linear trend; Table 2). Similar temporal linear declines in usual mean daily sodium intake occurred among all subpopulations explored (Table 2). In post hoc exploratory analyses, usual daily sodium density slightly increased from 2003 to 2016 (P<0.001 for trend), whereas sodium-to-potassium ratios, energy, and potassium intake declined (P<0.05 for trend; Table S4).
Table 2.
Mean Usual Daily Sodium Intake (mg) Among US Youths, 2003–2016, Overall and Among Selected Groups (N=12 249)
| Survey Cycles | 2003–2004 | 2005–2006 | 2007–2008 | 2009–2010 | 2011–2012 | 2013–2014 | 2015–2016 | β-Coefficient (95% CI)* |
P Value Linear Trend† |
|---|---|---|---|---|---|---|---|---|---|
| Population group, mean (SE) | |||||||||
| Overall | 3381 (22) | 3429 (17) | 3307 (15) | 3306 (34) | 3250 (25) | 3226 (26) | 3208 (22) | −35 (−43 to −27) | <0.001 |
| 8–12 y | 3290 (36) | 3309 (21) | 3215 (28) | 3167 (24) | 3157 (25) | 3140 (22) | 3123 (20) | −32 (−42 to −22) | <0.001 |
| 13–17 y | 3456 (41) | 3541 (32) | 3399 (20) | 3435 (49) | 3339 (31) | 3317 (32) | 3295 (28) | −35 (−48 to −23) | <0.001 |
| Boys | 3723 (32) | 3777 (29) | 3654 (25) | 3640 (39) | 3598 (31) | 3584 (37) | 3550 (34) | −34 (−47 to −22) | <0.001 |
| Girls | 3034 (17) | 3050 (21) | 2948 (26) | 2971 (25) | 2909 (18) | 2867 (26) | 2877 (16) | −31 (−38 to −24) | <0.001 |
| White, NH | 3438 (32) | 3526 (28) | 3411 (25) | 3368 (50) | 3289 (31) | 3274 (38) | 3267 (39) | −40 (−54 to −27) | <0.001 |
| Black, NH | 3266 (40) | 3243 (30) | 3097 (35) | 3164 (33) | 3100 (16) | 3053 (39) | 3068 (37) | −34 (−48 to −21) | <0.001 |
| Mexican-American | 3331 (22) | 3227 (26) | 3144 (24) | 3153 (39) | 3269 (64) | 3174 (31) | 3083 (26) | −26 (−37 to −14) | <0.001 |
| Under/normal | 3429 (28) | 3480 (23) | 3351 (20) | 3350 (34) | 3290 (30) | 3276 (35) | 3260 (26) | −35 (−46 to −25) | <0.001 |
| Overweight | 3279 (43) | 3292 (38) | 3178 (33) | 3210 (54) | 3211 (50) | 3216 (47) | 3107 (37) | −24 (−39 to −9) | 0.002 |
| Obese | 3298 (28) | 3380 (36) | 3271 (54) | 3251 (55) | 3142 (33) | 3091 (44) | 3143 (26) | −42 (−54 to −29) | <0.001 |
The unweighted sample size (n) for each subgroup can be found in Table S3. NH indicates non-Hispanic.
β-coefficients (adjusted change in sodium per 2-year survey cycle) were estimated using linear regression models adjusting for all covariates used to estimate the Best Linear Unbiased Predictors.
P values for quadratic trends were nonsignificant and are not reported.
Overall, the prevalence of youths with EBP/hypertension declined from 16.2% in 2003 to 2004 to 13.3% in 2015 to 2016 (P<0.001 for trend), as did hypertension, from 6.6% to 4.9% (P=0.005 for trend; Table 3). Similarly, using the former criteria, the EBP/hypertension prevalence and hypertension declined over this time period (Table S5). In exploratory analyses, significant temporal declines occurred in EBP/hypertension prevalence among most subgroups with a few exceptions: among Mexican-American youths or those who were overweight, temporal linear trends were not significant. Temporal linear trends in the prevalence of hypertension were inconsistent across other subgroups, and some prevalence estimates were statistically unreliable (Table 3).
Table 3.
Temporal Trends in the EBP/HTN Prevalence of and HTN Among US Youths, 2003–2016, Overall and Among Selected Groups (N=12 249)*
| Survey Cycles | 2003–2004 | 2005–2006 | 2007–2008 | 2009–2010 | 2011–2012 | 2013–2014 | 2015–2016 |
P Value Linear Trend† |
|---|---|---|---|---|---|---|---|---|
| Population group, % (SE) | ||||||||
| Overall | ||||||||
| EBP/HTN | 16.2 (1.8) | 17.3 (2.2) | 15.6 (1.5) | 12.3 (1.0) | 12.5 (1.2) | 8.7 (1.0) | 13.3 (1.3) | <0.001 |
| HTN | 6.6 (1.3) | 7.8 (1.7) | 5.8 (0.7) | 5.0 (0.7) | 4.8 (1.0) | 2.7 (0.4) | 4.9 (0.7) | 0.005 |
| 8–12 y | ||||||||
| EBP/HTN | 13.1 (2.0) | 16.0 (3.1) | 13.0 (2.4) | 10.6 (1.4) | 9.3 (1.7) | 9.1 (1.4) | 10.8 (1.5) | 0.022 |
| HTN | 6.6 (1.5) | 7.6 (2.2) | 6.1 (1.3) | 4.8 (0.7) | 5.4 (1.6) | 3.8 (0.7) | 5.3 (1.0) | 0.094 |
| 13–17 y | ||||||||
| EBP/HTN | 19.2 (2.3) | 18.5 (1.9) | 18.2 (2.1) | 14.0 (1.7) | 15.7 (2.1) | 8.3 (1.7) | 15.6 (1.7) | 0.002 |
| HTN | 6.6 (1.6) | 8.1 (1.5) | 5.6 (1.3) | 5.2 (1.1) | 4.3 (1.2) | 1.6 (0.5) | 4.4 (0.7) | 0.001 |
| Boys | ||||||||
| EBP/HTN | 20.6 (2.3) | 19.4 (2.8) | 18.3 (2.3) | 16.0 (1.9) | 17.4 (2.0) | 10.4 (1.8) | 16.9 (1.6) | 0.009 |
| HTN | 7.2 (1.6) | 8.0 (2.3) | 6.2 (1.2) | 6.3 (1.1) | 6.8 (1.7) | 2.6 (0.6) | 5.9 (0.8) | 0.063 |
| Girls | ||||||||
| EBP/HTN | 11.3 (2.3) | 14.8 (2.3) | 12.9 (1.8) | 9.0 (1.3) | 8.1 (2.0) | 7.1 (1.2) | 9.8 (1.7) | 0.018 |
| HTN | 5.9 (1.6) | 7.6 (1.7) | 5.4 (0.9) | 3.8 (1.0) | - | 2.8 (0.8) | 3.8 (0.8) | 0.008 |
| White, NH | ||||||||
| EBP/HTN | 16.5 (2.4) | 17.4 (3.1) | 15.5 (1.7) | 11.7 (1.4) | 10.2 (1.8) | 7.3 (1.6) | 10.7 (1.7) | <0.001 |
| HTN | 7.0 (1.8) | 8.4 (2.4) | 6.3 (1.0) | 4.6 (1.2) | 3.8 (1.5)* | 1.5 (0.4) | 2.7 (0.7) | <0.001 |
| Black, NH | ||||||||
| EBP/HTN | 18.4 (2.5) | 20.1 (3.6) | 18.5 (3.0) | 15.5 (2.6) | 16.5 (2.0) | 9.4 (1.6) | 16.4 (1.4) | 0.010 |
| HTN | 6.8 (1.2) | 10.3 (2.1) | 7.1 (1.9) | 7.8 (1.4) | 4.9 (1.2) | 3.0 (1.2)* | 4.7 (1.4) | 0.003 |
| Mexican-American | ||||||||
| EBP/HTN | 15.1 (1.7) | 16.3 (2.0) | 15.5 (4.3) | 15.4 (2.8) | 13.5 (2.5) | 12.4 (1.6) | 16.9 (2.9) | 0.730 |
| HTN | 5.4 (1.5) | 4.3 (0.9) | - | 5.0 (1.5) | 5.8 (1.9) | 4.6 (1.8)* | 9.0 (2.2) | 0.186 |
| Under/normal | ||||||||
| EBP/HTN | 12.9 (1.6) | 16.0 (3.3) | 12.1 (1.5) | 7.8 (0.7) | 9.4 (1.1) | 6.4 (1.1) | 8.7 (1.7) | 0.001 |
| HTN | 4.9 (0.9) | 7.8 (2.4)* | 4.8 (0.8) | 2.8 (0.6) | 2.8 (0.6) | 1.7 (0.4) | 2.7 (1.0)* | 0.002 |
| Overweight | ||||||||
| EBP/HTN | 14.7 (3.2) | 10.0 (1.7) | 18.2 (3.6) | 15.7 (4.1) | 14.1 (2.4) | 8.0 (2.1) | 15.0 (3.1) | 0.624 |
| HTN | 6.9 (2.7)* | 3.7 (0.8) | 6.6 (1.9) | 4.8 (1.9)* | 6.0 (1.3) | - | 4.6 (1.2) | 0.300 |
| Obese | ||||||||
| EBP/HTN | 30.1 (5.0) | 28.6 (3.5) | 25.3 (3.5) | 24.9 (2.8) | 20.9 (3.9) | 16.2 (2.5) | 25.5 (2.4) | 0.029 |
| HTN | 12.4 (3.3) | 11.8 (2.0) | 8.6 (2.6) | 13.0 (2.2) | 10.1 (3.9)* | 5.9 (1.6) | 11.6 (2.1) | 0.347 |
The unweighted sample size (n) for each subgroup can be found in Table S3. EBP indicates elevated blood pressure; HTN, hypertension; NH, non-Hispanic; and RSE, relative SE.
30%< RSE ≤40%. RSE >40% are suppressed.
P values for quadratic trends were nonsignificant and are not reported.
EBP/hypertension was significantly associated with usual sodium intake among youths (Table 4). Holding constant survey cycle, sex, age, race and Hispanic origin, and weight status, an additional 1000 mg of sodium consumed per day was associated with 18% higher adjusted odds ratio (AOR) of EBP/hypertension (1.18; 95% CI, 1.03–1.35). The AOR for hypertension was 1.20 (0.96–1.50). Additionally adjusting for energy intake, the AORs per 1000 mg sodium consumed per day were 1.39 (1.08–1.80) for EBP/hypertension and 1.33 (0.89–1.98) for hypertension (Table 4). The AOR for EBP/hypertension per mg of sodium per 1000 kcal consumed was 2.53 (1.14–5.60) and per unit difference in the molar ratio of sodium-to-potassium intake was 1.35 (1.12–1.64; Table 4). The AOR for hypertension per unit difference in the molar ratio of sodium-to-potassium intake was 1.58 (1.13–2.21).
Table 4.
AORs* for EBP/HTN and HTN Per Unit Difference in Usual Intake Measures Among US Youths, 2003–2016 (N=12 249)
| Usual intake measure (unit difference) | EBP/HTN | HTN | ||
|---|---|---|---|---|
| AOR (95% CI) | P Value | AOR (95% CI) | P Value | |
| sodium (1000 mg) | 1.18 (1.03–1.35) | 0.019 | 1.20 (0.96–1.50) | 0.102 |
| sodium (1000 mg) adjusted for energy | 1.39 (1.08–1.80) | 0.012 | 1.33 (0.89–1.98) | 0.167 |
| sodium density (mg per 1000 kcal) adjusted for energy | 2.53 (1.14–5.60) | 0.023 | 2.46 (0.71–8.59) | 0.156 |
| sodium/potassium (mmol/mmol) | 1.35 (1.12–1.64) | 0.002 | 1.58 (1.13–2.21) | 0.008 |
| sodium/potassium (mmol/mmol) adjusted for K | 1.49 (1.19–1.88) | 0.001 | 1.76 (1.18–2.62) | 0.006 |
AOR indicates adjusted odds ratio; EBP, elevated blood pressure or hypertension; and HTN, hypertension.
All models adjusted for survey cycle, age, sex, race and Hispanic origin, and weight status groups
The AORs for EBP/hypertension per 2-year survey cycle from 2003 to 2016 were 0.91 (0.86–0.96) when adjusting for sociodemographic characteristics and weight status (model 2) and 0.91 (0.87–0.96) additionally adjusting for usual sodium intake (model 3) and were similar for other sodium intake measures (models 4–7; Table 5). The AOR for hypertension per 2-year survey cycle also was similar across models.
Table 5.
AORs for the Prevalence of EBP/HTN and HTN Per 2-Year Survey Cycle, Among US Youths, 2003–2016 (N=12 249)
| Model | EBP/HTN | HTN | ||||
|---|---|---|---|---|---|---|
| AOR (95% CI) |
β-Coefficient | P Value | AOR (95% CI) |
β-Coefficient | P Value | |
| 1* | 0.92 (0.87–0.96) | −0.086 | <0.001 | 0.89 (0.83–0.95) | −0.118 | 0.001 |
| 2† | 0.91 (0.86–0.96) | −0.095 | <0.001 | 0.88 (0.82–0.95) | −0.126 | 0.001 |
| 3‡ | 0.91 (0.87–0.96) | −0.089 | 0.001 | 0.89 (0.82–0.96) | −0.119 | 0.002 |
| 4§ | 0.91 (0.86–0.96) | −0.096 | <0.001 | 0.88 (0.82–0.95) | −0.123 | 0.001 |
| 5∥ | 0.90 (0.85–0.95) | −0.103 | <0.001 | 0.88 (0.81–0.95) | −0.130 | 0.001 |
| 6¶ | 0.91 (0.87–0.96) | −0.093 | <0.001 | 0.88 (0.82–0.95) | −0.123 | 0.001 |
| 7# | 0.91 (0.87–0.96) | −0.089 | 0.001 | 0.89 (0.82–0.96) | −0.118 | 0.002 |
AOR indicates adjusted odds ratio; EBP, elevated blood pressure or hypertension; and HTN, hypertension.
Model 1 is adjusted for survey cycle, age (y), sex, and race and Hispanic origin.
Model 2 is model 1 plus weight status
Model 3 is model 2 plus usual daily sodium intake (mg).
Model 4 is model 3 plus usual daily energy intake (kcal).
Model 5 is model 2 plus usual sodium density (mg/1000 kcal) and usual daily energy intake (kcal).
Model 6 is model 2 plus usual sodium-to-potassium ratio (mmol/mmol).
Model 7 is model 6 plus usual daily potassium intake (mmol).
Sensitivity Analyses
The frequency of youths adding salt at the table changed over this time period. Among the subset of participants with data, temporal trends in poverty-income ratio were not significant, and inactivity significantly increased (Table S6).
The AORs for EBP/hypertension and hypertension per 1000 mg of usual sodium intake were similar before and after adjusting for poverty-income ratio and physical activity status. Adjusting for table salt use, the AOR for hypertension with higher usual sodium intake was similar, whereas EBP/hypertension became nonsignificant (P=0.344; Table S7). The AOR for EBP/hypertension per survey cycle remained similar before and after adjusting for table salt use, poverty-income ratio, or physical activity status (Table S8).
Discussion
From 2003 to 2016, we observed significant linear declines in both EBP/hypertension and hypertension among a nationally representative sample of US youths. A slight, but statistically significant, temporal linear decline also occurred in usual sodium intake. Although usual sodium intake was positively associated with EBP/hypertension, the temporal linear decline in BP was similar before and after adjusting for usual sodium intake. These results suggest recent temporal linear declines in EBP/hypertension were not correlated with the small linear decline in usual population sodium intake.
This study updates temporal trends in sodium intake and EBP/hypertension prevalence up to 2015-2016. Although there are differences in the estimation of sodium intake, definitions of EBP/hypertension, years included, and population subgroups examined, in general, our study results seem consistent with published trends through 2012, with 1 exception.4-7 Unlike our study, Xi et al6 suggested mean population sodium intake did not decline between 2003 and 2012; however, these results might be explained by not accounting for adjustments for salt added during cooking, which could have led to an artificial increase in sodium intake before 2009. Our results are similar to a previous study which also used methods to ensure consistency in sodium estimation over time.4 Though mean sodium intake slightly declined through 2016, the decline seemed to be related to a lower energy intake rather than less consumption of sodium dense foods.4 Although the sodium content of some US foods declined, reductions are inconsistent and improvements in diet quality of youths may be primarily driven by decreased empty calorie, and increased whole fruit, consumption.17,18
Our results are consistent with recent prevalence estimates among US youth,19,20 with 1 exception. In 2011 to 2012, EBP/hypertension prevalence was slightly higher in our study compared with previous estimates, perhaps attributed to the use of different definitions for EBP/hypertension.5,6 The lack of significant decline among Mexican-American youths is consistent with previous research through 2012 despite differences in samples, race categorization (ie, Hispanic), and EBP/hypertension definitions6 This could be due to a slightly higher proportion of Mexican-American overweight youths (19% versus 16% in non-Hispanic white youths), as EBP/hypertension also did not decline over time among overweight participants. Furthermore, the temporal declines in usual sodium intake were smallest among Mexican-American youths and those who were overweight, which might explain the lack of a secular trend in EBP/hypertension in these groups. Additionally, the EBP/hypertension prevalence and hypertension were significantly lower in 2013 to 2014 compared with other survey cycles, potentially due to the higher level of physical activity seen in 2013 to 2014, chance, or other factors.
To our knowledge, the correlation between temporal trends in BP and sodium intake among US youths has only been investigated in one other study, using NHANES data preceding 2009.7 Similar to our results, this study indicated after adjusting for demographic characteristics, body mass index Z score, and waist circumference Z score, further adjusting for sodium intake did not change temporal trends in BP, despite a positive association between BP and sodium intake. Potential reasons for the lack of correlation in the current study are unexplained measurement error, the small magnitude of the decline (<250 mg/day) in usual sodium intake over the time period, or a true null association.
In the current study, usual sodium intake was modestly associated with EBP/hypertension. The odds for EBP/hypertension adjusting for energy and other characteristics were consistent with the 36% higher odds seen in a previous study, despite differences in methods.7 Additionally, the associations of EBP/hypertension with the molar ratio of sodium-to-potassium and when additionally holding potassium intake constant suggests sodium intake is associated with EBP/hypertension independent of potassium intake.
Our study has some limitations. Although data were weighted to account for nonresponse, NHANES participants may differ from nonparticipants. In our study, hypertension was defined using up to 3 BP readings on a single visit; whereas, a clinical diagnosis requires confirmed readings on 3 separate occasions.12 This could result in misclassification of some participants, potentially attenuating observed associations with sodium intake. Dietary recalls can be subject to errors in reporting and nutrient estimation.21 Bias from self-reported intake rather than true dietary sodium intake may attenuate the true association between sodium and BP.8 Usual sodium intake estimation did not account for the amount of sodium consumed from salt added at the table, estimated to contribute ≈5% of total sodium intake.22 We could not account for several covariates among the full sample of participants, including income, physical activity, and use of antihypertensive medication. Given the low prevalence (<1%) of antihypertensive use among adolescents aged 16 to 17 years, it is doubtful that increases in medication use could explain the decline in BP. The sample sizes in certain subgroups were small diminishing the reliability of hypertension prevalence estimates for some 2-year survey cycles. Additionally, low statistical power or prevalence may explain the lack of statistically significant associations for some outcomes. However, lack of correction for multiple comparisons could increase the likelihood of statistically significant results due to chance.23
Lifestyle modifications, such as maintaining a healthy weight, consuming a Dietary Approaches to Stop Hypertension-type eating pattern, limiting sodium, and increasing physical activity, are reasonable strategies for hypertension prevention.12 As indicated by the small decline in sodium observed in our study, additional efforts may be required to achieve substantive reductions in sodium intake. Reformulation efforts could support sodium reduction activities as the majority of sodium intake among US youths is estimated to come from salt added to commercially processed foods.24,25 Additionally, clinician advice can support sodium reduction efforts through increased consumer awareness and motivation to make healthier choices.12,24,26 This study, along with previous studies,3,7 indicate higher sodium intake is associated with EBP/hypertension, supporting efforts to reduce population sodium intake.
Perspectives
To our knowledge, this is the first study to assess concurrent temporal trends through 2016 in usual sodium intake and EBP/hypertension defined using the 2017 BP guidelines and to determine if trends were correlated. The decline in EBP/hypertension from 2003 to 2016 was not correlated with usual sodium intake, potentially due to the relatively small decline in sodium intake (≈35 mg/survey cycle). As population distributions of covariates examined were similar over this time period, they also did not account for BP declines. As a greater proportion of youths are classified with EBP/hypertension under the current new guidelines, understanding the contributions of lifestyle factors to temporal decreases could help achieve further declines. Expanding assessment of physical activity and potentially medication use among US youths in nationally representative surveys might help us better understand the contribution of these factors and how they interact with diet in the prevention of EBP/hypertension.
Supplementary Material
Novelty and Significance.
What Is New?
Temporal trends in elevated blood pressure (EBP)/hypertension (ie, elevated blood pressure or hypertension) and usual sodium intake were updated through 2016 among a nationally representative sample of US youths.
Significant temporal declines in EBP/hypertension were concurrent with slight temporal reductions in mean usual sodium intake.
Higher usual sodium intake was associated with ≈20% higher EBP/hypertension. Declines in EBP/hypertension from 2003 to 2016 were not correlated with the concurrent, but slight reductions (mean <250 mg/day over 14 years) in usual population sodium intake.
What Is Relevant?
About 1 in 8 youth have EBP/hypertension. Evidence suggests that EBP in childhood tracks with age and can increase the risk for high blood pressure in adults, a major risk factor for heart disease and stroke.
As American youth consume ≈3208 mg of sodium/day, continued efforts are needed to lower intake. Clinicians can provide dietary counseling to reduce sodium consumption to enhance patients’ efforts to prevent EBP.
Summary
Modest declines in BP and usual sodium intake were observed from 2003 to 2016 among US youths. The results from this study indicate that usual sodium intake is associated with BP. The small decline in sodium observed over this time period suggests that additional efforts are required to reduce sodium intake among this population. Future studies examining other lifestyle factors are warranted to help us understand how these factors may contribute to prevention efforts.
Acknowledgments
The findings and conclusions presented in this article are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention (CDC).
Sources of Funding
This work (staff time and National Health and Nutrition Examination Survey [NHANES]) was supported by the Centers for Disease Control and Prevention (CDC).
Footnotes
Disclosures
None.
Contributor Information
Katherine J. Overwyk, Division for Heart Disease and Stroke Prevention, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, GA; IHRC, Inc, Atlanta, GA
Lixia Zhao, Division for Heart Disease and Stroke Prevention, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, GA; IHRC, Inc, Atlanta, GA.
Zefeng Zhang, Division for Heart Disease and Stroke Prevention, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, GA.
Jennifer L. Wiltz, Division for Heart Disease and Stroke Prevention, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, GA; United States Public Health Service, Atlanta
Elizabeth K. Dunford, Food Policy Division, The George Institute for Global Health, University of New South Wales, Sydney, Australia
Mary E. Cogswell, Division for Heart Disease and Stroke Prevention, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, GA
References
- 1.Benjamin EJ, Virani SS, Callaway CW, et al. ; American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics-2018 update: a report from the American Heart Association. Circulation. 2018;137:e67–e492. doi: 10.1161/CIR.0000000000000558 [DOI] [PubMed] [Google Scholar]
- 2.Shen W, Zhang T, Li S, Zhang H, Xi B, Shen H, Fernandez C, Bazzano L, He J, Chen W. Race and sex differences of long-term blood pressure profiles from childhood and adult hypertension: the bogalusa heart study. Hypertension. 2017;70:66–74. doi: 10.1161/HYPERTENSIONAHA.117.09537 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.National Academies of Sciences, Engineering, and Medicine. Dietary Reference Intakes for Sodium and Potassium. Washington, DC: The National Academies Press; 2019. [PubMed] [Google Scholar]
- 4.Centers for Disease Control and Prevention. Trends in the prevalence of excess dietary sodium intake–United States, 2003–2010. MMWR Morb Mortal Wkly Rep. 2013;62:1022–1025. [PMC free article] [PubMed] [Google Scholar]
- 5.Kit BK, Kuklina E, Carroll MD, Ostchega Y, Freedman DS, Ogden CL. Prevalence of and trends in dyslipidemia and blood pressure among US children and adolescents, 1999–2012. JAMA Pediatr. 2015;169:272–279. doi: 10.1001/jamapediatrics.2014.3216 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Xi B, Zhang T, Zhang M, Liu F, Zong X,Zhao M, Wang Y. Trends in elevated blood pressure among US children and adolescents: 1999–2012. Am J Hypertens. 2016;29:217–225. doi: 10.1093/ajh/hpv091 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Rosner B, Cook NR, Daniels S, Falkner B. Childhood blood pressure trends and risk factors for high blood pressure: the NHANES experience 1988–2008. Hypertension. 2013;62:247–254. doi: 10.1161/HYPERTENSIONAHA.111.00831 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Freedman LS, Commins JM, Moler JE, Willett W, Tinker LF, Subar AF, Spiegelman D, Rhodes D, Potischman N, Neuhouser ML, Moshfegh AJ, Kipnis V, Arab L, Prentice RL. Pooled results from 5 validation studies of dietary self-report instruments using recovery biomarkers for potassium and sodium intake. Am J Epidemiol. 2015;181:473–487. doi: 10.1093/aje/kwu325 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Johnson CL, Dohrmann SM, Burt VL, Mohadjer LK. National health and nutrition examination survey: sample design, 2011–2014. National Center for Health Statistics. Vital Health Stat. 2014;2:1–25. [PubMed] [Google Scholar]
- 10.Zipf G, Chiappa M, Porter KS, et al. National Health and Nutrition Examination Survey: plan and operations, 1999–2010. National Center for Health Statistics. Vital Health Stat. 2013;1:1–27. [PubMed] [Google Scholar]
- 11.National Center for Health Statistics. National Health and Nutrition Examination Survey. 2013–2014 Survey Operations Manual: Physician Examination Procedures Manual. https://wwwn.cdc.gov/nchs/data/nhanes/2013-2014/manuals/Phys_Exam_Manual_2013.pdf. Accessed March 1, 2018.
- 12.Flynn JT, Kaelber DC, Baker-Smith CM, et al. ; Subcommittee on Screening and Management of High Blood Pressure in Children. Clinical practice guideline for screening and management of high blood pressure in children and adolescents. Pediatrics. 2017;140:e20171904. [DOI] [PubMed] [Google Scholar]
- 13.Expert Panel on Integrated Guidelines for Cardiovascular Health and Risk Reduction in Children and Adolescents; National Heart, Lung, and Blood Institute. Expert panel on integrated guidelines for cardiovascular health and risk reduction in children and adolescents: summary report. Pediatrics. 2011;128(suppl 5):S213–S256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Johnson CL, Paulose-Ram R, Ogden CL, National Health and Nutrition Examination Survey: analytic guidelines, 1999–2010. National center for health statistics. Vital Health Stat. 2013;2:CS241906. [PubMed] [Google Scholar]
- 15.Ogden CL, Kuczmarski RJ, Flegal KM, Mei Z, Guo S, Wei R, Grummer-Strawn LM, Curtin LR, Roche AF, Johnson CL. Centers for Disease Control and Prevention 2000 growth charts for the United States: improvements to the 1977 national center for health statistics version. Pediatrics. 2002;109:45–60. [DOI] [PubMed] [Google Scholar]
- 16.Willet W. Chapter 11: Implications of total energy intake for epidemiologic analyses In: Nutritional Epidemiology. New York, NY: Oxford University Press; 2013:260–286. [Google Scholar]
- 17.Wolfson JA, Moran AJ, Jarlenski MP, Bleich SN. Trends in sodium content of menu items in large chain restaurants in the US. Am J Prev Med. 2018;54:28–36. doi: 10.1016/j.amepre.2017.08.018 [DOI] [PubMed] [Google Scholar]
- 18.Gu X, Tucker KL. Dietary quality of the US child and adolescent population: trends from 1999 to 2012 and associations with the use of federal nutrition assistance programs. Am J Clin Nutr. 2017;105:194–202. doi: 10.3945/ajcn.116.135095 [DOI] [PubMed] [Google Scholar]
- 19.Sharma AK, Metzger DL, Rodd CJ. Prevalence and severity of high blood pressure among children based on the 2017 American Academy of Pediatrics guidelines. JAMA Pediatr. 2018;172:557–565. doi: 10.1001/jamapediatrics.2018.0223 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Jackson SL, Zhang Z, Wiltz JL, Loustalot F, Ritchie MD, Goodman AB, Yang Q. Hypertension among youths–United States, 2001–2016. Am J Transplant. 2018;67:758–762. doi: 10.1111/ajt.15050 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Espeland MA, Kumanyika S, Wilson AC, Reboussin DM, Easter L, Self M, Robertson J, Brown WM, McFarlane M; TONE Cooperative Research Group. Statistical issues in analyzing 24-hour dietary recall and 24-hour urine collection data for sodium and potassium intakes. Am J Epidemiol. 2001;153:996–1006. doi: 10.1093/aje/153.10.996 [DOI] [PubMed] [Google Scholar]
- 22.Harnack LJ, Cogswell ME, Shikany JM, Gardner CD, Gillespie C, Loria CM, Zhou X, Yuan K, Steffen LM. Sources of sodium in US adults from 3 geographic regions. Circulation. 2017;135:1775–1783. doi: 10.1161/CIRCULATIONAHA.116.024446 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Streiner DL. Best (but oft-forgotten) practices: the multiple problems of multiplicity – whether and how to correct for many statistical tests. Am J Clin Nutr. 2015; 102:721–728. doi: 10.3945/ajcn.115.113548 [DOI] [PubMed] [Google Scholar]
- 24.Institute of Medicine. Strategies to Reduce Sodium Intake in the United States. Washington, DC: National Academies Press; 2010. [Google Scholar]
- 25.Quader ZS, Gillespie C, Sliwa SA, Ahuja JKC, Burdg JP, Moshfegh A, Pehrsson PR, Gunn JP, Mugavero K, Cogswell ME. Sodium intake among US school-aged children: national health and nutrition examination survey, 2011–2012. J Acad Nutr Diet. 2017;117:39–47. doi: 10.1016/j.jand.2016.09.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Eckel RH, Jakicic JM, Ard JD, de Jesus JM, Miller NH, Hubbard VS, Lee I-M, Lichtenstein AH, Loria CM, Millen BE, Nonas CA, Sacks FM, Smith SC, Svetkey LP, Wadden TA, Yanovski. 2013 AHA/ACC Guideline on lifestyle management to reduce cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines. Ciruclation. 2014; 129(suppl 2):S76–S99. doi: 10.1161/01.cir.0000437740.48606.d1 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.