Hypertension has been well recognized as a major independent risk factor for cardiovascular disease and stroke (1). Furthermore, hypertension has had great impact on health outcomes and disparities (2- 5). The National Health and Nutrition Examination Survey (NHANES) provided blood pressure (BP) data to track trends in the prevalence of hypertension in U.S. adults (6- 7). However, differences in BP measurement techniques made it impossible to precisely quantify trends in BP and hypertension prevalence in early NHANES cycles (NHANES I, II) (7). The continuous NHANES (from 1999 to 2010) standardized procedures for BP measurement (8), providing an opportunity for precise quantification of trends in the distribution of BP and prevalence of hypertension. Data from early national surveys revealed a decreasing trend in mean systolic blood pressure (SBP) and hypertension prevalence from 1960 to 1980, and from NHANES I (1971 to 1974) to NHANES III Phase 1 (1988 to 1991). Age-adjusted hypertension prevalence decreased by 15.9% from 36.3% to 20.4%, and hypertension awareness, treatment, and control showed an increasing trend during that period (7). However, hypertension prevalence increased from NHANES III (1988 to 1994) to NHANES 1999 to 2000, with no improvement of awareness, management, and control of hypertension (9). Between 1999 to 2000 and 2007 to 2008, prevalent hypertension remained constant (10). In this study, we assessed whether current clinical approaches have been effective in diagnosing and adequately treating hypertensive patients. We reported the trends in BP, and prevalence, awareness, management, and control of hypertension in the U.S. adult population based on NHANES data from 1999 to 2010.

The NHANES is a cross-sectional nationally representative health and nutrition examination survey conducted by the National Center for Health Statistics. The study was approved by the National Center for Health Statistics Institutional Ethics Review Board, and all adult participants provided written informed consent (11). NHANES maintains high standards to ensure minimal nonsampling and measurement errors during survey planning, data collection, and processing (11). The survey uses a complex, stratified, multistage probability sample to represent the civilian noninstitutionalized U.S. population. The NHANES sampling procedure consists of 4 stages: counties, segments, households, and individuals (12). The participant response rate ranged from 72.9% to 78.3% for interviews, and 68.6% to 72.7% for examinations at mobile examination centers (MECs) among adults (≥20 years of age) for each survey cycle through 1999 to 2010. In this report, only adult (≥20 years of age) participants with complete BP measurements were included for analysis. We assessed progress in hypertension prevention, diagnosis, and treatment using regression models with a 2-year survey cycle treated as a continuous variable.

BP was measured by mercury sphygmomanometer using a standardized protocol (8). Mean SBP and mean diastolic blood pressure (DBP) were calculated by averaging 3 to 4 BP measurements. Hypertension was defined as SBP ≥140 mm Hg or DBP ≥90 mm Hg, or on antihypertensive medication; pre-hypertension was determined as SBP ≥120 mm Hg or DBP ≥80 mm Hg, but not meeting the criteria for hypertension (13).

Hypertension awareness, management, and control were analyzed in hypertensive participants based on questionnaires and BP measurements. Awareness was defined as having been informed of hypertension diagnosis; management was defined as taking antihypertensive medication or adopting lifestyle modifications (increasing activity or controlling weight); hypertension control was defined as SBP <140 mm Hg and DBP <90 mm Hg, and was also analyzed in hypertensive participants who managed their hypertension (control in management).

Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared, and grouped into 3 categories: <25 kg/m², 25 to 29.9 kg/m² (overweight), and ≥30 kg/m² (obesity). Race/ethnicity was self-reported and was classified as non-Hispanic white, non-Hispanic black, Mexican American, other Hispanic, and other. Age was categorized as 20 to 39, 40 to 59, and 60+ years based on age at the interview. Education status was classed into not completing high school, high school only, and higher education. Poverty income ratio was used to reflect socioeconomic status, and <1 was determined to be low income and >1.85 was high income. Smoking status was classed into nonsmoker (smoked <100 cigarettes lifetime), past smoker, and current smoker.

Statistical analyses were carried out with SAS for Windows version 9.2 (SAS Institute, Cary, North Carolina). According to NHANES Analytic and Reporting Guidelines (11,14), all analyses took into account differential probabilities of selection and the complex sample design, and nonresponse and noncoverage by using sample weights and SAS survey analysis procedures. Standard errors were calculated using Taylor series linearization. Age-adjusted values were adjusted to the standard population (the 2000 Census population) by the direct method using the age groups 20 to 39, 40 to 59, and 60+ years. Linear trends in the distributions of mean SBP and DBP, and prevalence, awareness, management, and control of hypertension between 1999 and 2010 were assessed with regression models with a 2-year survey cycle treated as a continuous variable. Statistical significance was determined as 2-sided p < 0.05. To further examine trends in SBP and DBP, selected percentiles of SBP and DBP by sex in NHANES 1999 to 2010 were graphed.

In NHANES 2009 to 2010, 6,218 adults were interviewed, 6,059 were examined at MECs, and 5,764 with complete BP values were included for analysis. From 1999 to 2010, 32,464 adults were interviewed, 30,752 were examined, and 28,995 were included for analysis. The demographic characteristics of adult participants are shown in (Table 1). The age-adjusted higher education rate was 58.4% in 2009 to 2010, and there was an overall increase from 1999 to 2010 (p = 0.004 for trend). Low income prevalence was 15.0% in 1999 to 2000, and decreased to 10.6% in 2005 to 2006, and then increased to 14.6% in 2009 to 2010. Participants (78.8%) were covered by health insurance in 2009 to 2010, with a 3% decrease from 1999 to 2010 (p = 0.019 for trend). Obesity prevalence increased from 30.0% to 35.5% from 1999 to 2010 (p < 0.001). Current smokers decreased from 23.8% to 20.5% from 1999 to 2010 (p = 0.017 for trend), whereas past smokers remained stable at around 24%.

Mean SBP and DBP from 1999 to 2010 are shown in (Table 2). Age-adjusted mean SBP and DBP were 120 (95% CI: 120 to 121) mm Hg and 70 (95% CI: 69 to 71) mm Hg, respectively, in 2009 to 2010, with a 4 mm Hg decrease in SBP and 3 mm Hg decrease in DBP since 1999 to 2000 (p < 0.001 for both trends). There was a significant decrease for both SBP and DBP among both men and women from 1999 to 2010. Selected percentiles of mean SBP and DBP are presented in (Figure 1). High percentiles (75th to 95th) of mean SBP >130 mm Hg showed a notable decreasing trend, especially in women. In 1999 to 2010, age explained 21.7% of the variance in mean SBP (R² = 0.217; p < 0.001). Sex, ethnicity, age group, body weight status, education level, income level, smoking status, and marital status were strongly associated with mean SBP. Similar patterns were observed for mean DBP, except that income level became nonsignificant.

Prevalence of hypertension and pre-hypertension is presented in (Table 3). The age-adjusted hypertension prevalence and pre-hypertension + hypertension prevalence were 29.5% (95% CI: 27.7% to 31.4%) and 52.6% (95% CI: 50.1% to 55.0%), respectively, in 2009 to 2010, with no significant improvement since 1999 to 2000. The prevalence of pre-hypertension + hypertension only showed a slight decrease in the 60+ years group and women. No improvement in hypertension prevalence or pre-hypertension + hypertension prevalence was observed among other groups. Throughout 1999 to 2010, the 60+ years group had an almost 9-fold higher hypertension prevalence than the 20- to 39-year-old group (about 60% vs 6%). Hypertension prevalence did not differ by sex; however, women had a much lower prevalence of pre-hypertension than men from 1999 to 2010. Ethnicity, age group, body weight status, education level, health insurance coverage, and income level were strongly associated with hypertension prevalence. There was a higher prevalence for non-Hispanic blacks, the overweight and obese, the elderly, and those covered by health insurance; the prevalence was lower for those with higher education.

Hypertension awareness, management, and control are presented in (Table 4). The age-adjusted awareness rate was 74.0% (95% CI: 68.6% to 79.4%) in 2009 to 2010, 69.7% (95% CI: 62.0% to 77.4%) among men, and 80.7% (95% CI: 74.5% to 86.8%) among women. From 1999 to 2010, Mexican Americans had lower awareness compared with non-Hispanic blacks and non-Hispanic whites; the 20- to 39-year-old group had the lowest awareness across sex and ethnic groups. In 2009 to 2010, 71.6% (95% CI: 65.7% to 77.5%) of participants managed their hypertension, and the age-adjusted management rate was 65.5% (95% CI: 57.5% to 73.5%) for men and 81.0% (95% CI: 74.2% to 87.9%) for women. Similar to awareness, Mexican Americans and the 20- to 39-year-old group also had lower management rates. The age-adjusted control rate was 36.8% (95% CI: 34.4% to 39.1%) in 1999 to 2010 and 46.5% (95% CI: 41.1% to 51.9%) in 2009 to 2010. Mexican Americans had the lowest control rate, and non-Hispanic whites had the highest control rate. Older participants tended to have better hypertension control than the younger group, except that non-Hispanic white older women had a lower control rate compared with their younger counterparts. In 1999 to 2010, 64.4% (95% CI: 59.9% to 68.9%) of participants attained BP control when managing their hypertension. Among participants who managed their hypertension from 1999 to 2010, men had lower control rate than women; Mexican Americans and non-Hispanic blacks had lower control rates than non-Hispanic whites; and older people had a lower control rate than middle-aged people.

In 1999 to 2010, sex, ethnicity, age group, body weight status, smoking, and health insurance coverage were all strongly associated with hypertension awareness; the awareness rate was lower in men, Mexican Americans, the younger group, those with normal body weight, and no health insurance, but higher in past smokers. The same pattern was observed in hypertension management and control. Additionally, marital status was also strongly associated with hypertension control, with currently married persons having better control rates. Overall, awareness, management, and control of hypertension were significantly increased from 1999 to 2010. However, awareness, management, and control rates remained poor in 2009 to 2010, with rates of 74.0% for awareness, 71.6% for management, 46.5% for control, and 64.4% for control in management. Compared with 2007 to 2008, hypertension prevalence and pre-hypertension + hypertension prevalence remained constant (p = 0.36 for both) in 2009 to 2010, and both men and women showed no improvement. Hypertension awareness, management, control, and control in management also showed no improvement (p = 0.89, 0.21, 0.92, and 0.095, respectively), which was evident among both men and women.

From 1999 to 2010, age-adjusted mean SBP decreased by 4 mm Hg (p < 0.001 for trend), especially for women (5 mm Hg; p < 0.001 for trend). Mean DBP displayed a similar pattern, with a significant decrease of 3 mm Hg (p < 0.001 for trend), and was significant for both men and women (p < 0.001 for both). Prevalence of hypertension among U.S. adults (≥20 years old) remained at a high level of around 30%, whereas pre-hypertension prevalence decreased in non-Hispanic black men (5.7%; p = 0.023 for trend). Awareness, management, and control of hypertension were significantly improved in almost all sex/ethnic groups, but remained poor, and did not improve from 2007 to 2010. Therefore, effective prevention, detection, management, and control of hypertension should continue to be important goals for health policy, public health, and medical care decision makers, as well as advocates and individuals at risk for hypertension (15).

Our findings on trends in mean SBP and mean DBP, and prevalence of hypertension were consistent with other studies (10,15). From 1999 to 2010, mean SBP and mean DBP both decreased significantly by 4 and 3 mm Hg, respectively (both trends, p < 0.001). It was reported (10) from 1988 to 2008 that SBP decreased in individuals with hypertension, but increased among individuals without hypertension (p = 0.02), which might have been caused by the adverse diet and lifestyle of nonhypertensive people. In our study, we found that mean SBP did not decrease among nonhypertensive participants (p = 0.173 for all, p = 0.268 for men, and p = 0.274 for women), but mean DBP showed a 2 mm Hg decrease (p < 0.001 for all, and men and women) from 1999 to 2010. The high percentiles (75th to 95th) of mean SBP >130 mm Hg showed a notable decreasing trend, especially in women, whereas lower percentiles did not decrease, reflecting the improvement in hypertension control from 1999 to 2010, which brought down BP among hypertensive people. The relatively stable trends in lower percentiles also suggested that prevalence of hypertension had no improvement, and prevention of hypertension had little success.

Despite the downward trends in mean SBP and mean DBP, hypertension prevalence remained constantly high, affecting 30% of the U.S. adult (≥20 years of age) population in 1999 to 2010. This stable trend was consistent with other studies (10,15- 16). We also found that ethnicity, age group, body weight status, education level, and marital status were strongly associated with hypertension prevalence. Increased hypertension prevalence was observed in non-Hispanic blacks, the overweight and obese group, those with lower education, and participants not currently married. Flegal et al. (17- 19) showed that obesity prevalence remained stable, with a possible slight increase from 1999 to 2010. From 1999 to 2010, among participants with complete BP measurements, obesity prevalence increased by 5.5% (p < 0.001 for trend) from 30.0% (95% CI: 26.9% to 33.1%) to 35.5% (95% CI: 33.7% to 37.3%), which might partially explain the lack of improvement in hypertension prevalence. Factors other than obesity were also likely to contribute to the adverse trend in hypertension prevalence, such as increasing consumption of dietary sodium, the increasingly sedentary lifestyle, and the suboptimal levels of health literacy among the general U.S. population (20- 21). Therefore, improving diet and lifestyle (10,22) for the entire population might have some impact on improving hypertension prevention (23) and help to reverse the adverse trend in hypertension prevalence.

There were significant improvements in awareness, management, and control of hypertension from 1999 to 2010 in almost all gender/ethnicity groups. However, non-Hispanic black and Mexican-American women did not show any significant increase in awareness (p = 0.182 and p = 0.076, respectively), and non-Hispanic black men showed no significant increase in control in management (p = 0.707). Nonetheless, the awareness, management, and control rates remained poor in 2009 to 2010 (74.7% for awareness, 72.3% for management, 45.1% for control, and 61.9% for control in management); what is worse, these rates showed no improvement from 2007 to 2010. These subtle improvements can be ascribed to the heavy campaign of programs, guidelines, and policies to facilitate hypertension prevention, detection and/or awareness, treatment, and control by several national initiatives (24- 27), although it also reflected the relatively small success achieved by these initiatives. Additionally, disparities were evident among sex, age groups, and ethnicity. Men, young people, and Mexican Americans had lower hypertension awareness, management, and control rates. For men, the middle-aged group had a better control rate. However, in non-Hispanic whites and Mexican Americans, control rates decreased with age. For participants who managed their hypertension, Mexican Americans and non-Hispanic blacks had lower control rates compared with non-Hispanic whites, and older people had lower control rates than younger and middle-aged people. Age-related differences in control might be explained by increased prevalence of treatment-resistant hypertension in older people (28). Because hypertension is a major risk factor for cardiovascular events in the elderly, improving control in this population would be extremely beneficial (29- 30). These findings suggest that public health efforts should be directed toward increasing awareness, management, and control among men, Mexican Americans, and young people, while increasing control and treatment of hypertension in minority groups and older people. Behavioral telephone intervention and home BP monitoring were found to be effective to promote BP control in a clinical trial by Bosworth et al. (31), and more large randomized trials would further validate the efficacy of such intervention.

Our findings had several clinical implications. First, there was no improvement in hypertension prevalence. To decrease hypertension prevalence, effective population-based strategies should be taken for hypertension prevention through advocating healthful eating and lifestyle in the entire population. Preventive population-level interventions tailored to the built environment and the food environment might lead to health benefits for the entire population (18). Second, awareness, management, and control of hypertension remained poor, with no improvement from 2007 to 2010. Therefore, strategies should be initiated to further improve hypertension detection, treatment, and control, especially in minority groups. Additionally, men, Mexican Americans, and 20 to 39 year olds had lower hypertension awareness, management, and control rates. Non-Hispanic blacks, Mexican Americans, and older people were less likely to achieve BP control when treating hypertension, suggesting that more efforts should be made to increase hypertension detection, management, and control in men, Mexican Americans, and the younger group, and to increase hypertension control among minority and elderly patients receiving hypertension treatment.

Our report had the following limitations: low hypertension prevalence and small sample size in the younger group (20 to 39 years old) in each 2-year survey cycle made it impossible to perform trend analysis in any 20- to 39-year-old gender/ethnicity subgroups. Hispanics other than Mexican Americans were not over sampled in 1999 to 2006, and any analysis on Hispanic groups before 2007 to 2008 was unreliable; thus, we only reported trends in Mexican Americans.

In 2009 to 2010, the prevalence of hypertension was 30.5% among men and 28.5% among women. From 1999 to 2010, hypertension prevalence remained constantly high; hypertension awareness, management, and control were significantly improved, but remained poor, and did not improve from 2007 to 2010. Strategies should be taken to improve hypertension prevention in the whole population, to increase hypertension detection, management, and control among men, Mexican Americans, and young people, and to increase control of hypertension among older people and minority groups receiving hypertension treatment.

The authors would like to thank Brandon H. Coleman, Department of Biostatistics, School of Public Health, University of Alabama at Birmingham, Birmingham, Alabama, for his review and revision of the paper.