CHAPTER III JOURNAL ARTICLE
METHODS Study Population
Data from the National Health and Nutrition Examination Survey (NHANES) was used for this study. The NHANES is conducted by the National Centers for Health
Statistics and is designed to assess the health of individuals from different ethnicities living in the US. The NHANES is a cross-sectional health study continuously gathering data through home interviews and physical exams completed in their Mobile
26
Examination Centers.13,14 This study utilized data from the continuous NHANES 1999- 2016.
Subjects in this study included self-identified MAs and NH-Whites over the age of 20 years who participated in both home interviews and physical examinations from NHANES. Both male and female participants were included. Females who were pregnant or lactating were excluded from this study as well as those in the underweight body mass index (BMI) category (n=1033).
Design and Procedures
This study examined a total of nine NHANES data sets accounting for 18 years worth of data. Data was analyzed in 2-year increments to observe trends in the prevalence of HTN among MAs and NH-Whites. Associated risk factors of HTN were also
compared between MAs and NH-Whites. Risk factors included age, sex, BMI, and abdominal obesity (assessed using waist-to-height ratio). Hypertension was defined at a SBP measurement of ≥ 130mmHg or a DBP measurement of ≥ 80 mmHg, 2,15 or
participants that used anti-hypertensive medication.
Anthropometric measurements including height, weight, and waist circumference (WC) were taken following NHANES protocols.16 From those measurements, BMI and waist-to-height ratio (WHtR) were calculated to evaluate weight status and abdominal obesity, respectively. As a measure of abdominal obesity, the WHtR has been shown in some studies to better predict risk of diabetes and CVD risk factors than WC alone.4,5,17 This study determined the odds ratio or likelihood of HTN compared with a reference group for each HTN risk factor. Males were compared to females, age groups 40-59 years
old and 60+ years old were compared to the youngest age group (20-39 years old), overweight and obese BMI categories were compared to normal BMI category, and those with abdominal obesity (WHtR > 0.5)18 were compared to those without abdominal obesity between both ethnic groups.
The degree of acculturation was based on the subject’s country of birth, number of years the subject had lived in the US, and language spoken.11 Acculturation was only assessed among MAs, as the acculturation questions were generally not given to NH- Whites. Answers to these questions were assigned points as follows: if country of birth was within the US, one point was given, if born outside the US, zero points were given; if participants had lived in the US for greater than 20 years, one point was given, less than 20 years resulted in zero points; and participants who spoke only/mostly English or spoke both languages (English and Spanish) equally were given one point, and those who spoke only/mostly Spanish were given zero points. Addition of these points resulted in an acculturation score ranging from zero (lowest) to three (highest).11 Acculturation score was analyzed to assess the likelihood of MAs developing HTN as acculturation status increased.
Statistical Analysis
Statistical Analysis Software (SAS System version 9.2, SAS Institute Inc., Cary, North Carolina, USA) was used to analyze data from NHANES. All data analyses including variance estimates used examination sample weights, stratification, and clustering to account for the NHANES complex sample design. Linear regression analyses were conducted on the prevalence of HTN and related risk factors over time to
28
determine significant trends. Multivariate logistic regression analysis was used to examine likelihood of HTN and the associated risk factors of HTN: age, sex, BMI, and WHtR between MAs and NH-Whites. T-tests and chi-squared analyses were conducted to determine statistical significance for continuous and categorical variables, respectively. The significance level was set at <0.05. This study was deemed exempt from review by the Central Washington University Institutional Review Board.
RESULTS
Descriptive statistics of the sampled population are shown in Table 1. A total of 28,462 non-pregnant and non-lactating adults with accurate blood pressure measurements were analyzed from the continuous NHANES 1999-2016 data. Included in this sample were 8,002 self-identified MAs and 20,460 NH-Whites. The MA population was younger and had a greater percentage of males in comparison to NH-Whites. The average SBP and DBP were higher among NH-Whites than MAs. The percentage of those taking HTN medication was greater among NH-Whites (23.1%) than MAs (11.8%). The average BMI for both ethnic groups were in the overweight category (25-29.9 kg/m2) with the mean BMI was significantly greater in MAs (29.5kg/m2) than NH-Whites (28.4kg/m2). The average WHtR was significantly greater among MAs (0.60) than in NH-Whites (0.58).
Table 1. Weighted descriptive statistics of sampled population, NHANES 1999-2016
Mexican Americans Non-Hispanic Whites P-value*
N 8,002 20,460
Percent Males 54.3 49.2
<0.0001
Percent Females 45.7 50.8
Age (years) 40.3±0.34 49±0.22 <0.0001
Systolic Blood Pressure (mmHg)
120.5±0.31 122.7±0.21 <0.0001
Diastolic Blood Pressure (mmHg) 69.9±0.21 70.8±0.18 0.0005 Percent on Hypertension Medication ^ 11.8 23.1 <0.0001 Weight (kg) 79.7±0.35 82.4±0.21 <0.0001 Height (cm) 164.2±0.15 170±0.08 <0.0001 BMI (kg/m2) 29.5±0.13 28.4±0.08 <0.0001 Waist-to-Height Ratio 0.60±0.0022 0.58±0.0012 <0.0001
* P-value determined by t-tests between Mexican Americans and Non-Hispanic Whites
^ Percent based off entire population within each ethnic group, not of those who are hypertensive
Overall, the actual prevalence of HTN from 1999-2016 was lower among MAs than in NH-Whites as shown in Figure 1. A linear regression analysis of the prevalence of HTN from 1999-2016 among NH-Whites (P = 0.81) and MAs (P = 0.09) showed no significant change. Conversely, from 2005 to 2016 the prevalence of HTN among MAs was significant (P = 0.005). Because the mean age of the MA population was much lower than in NH-Whites (40.3 vs. 49.0 years), the age-adjusted prevalence of HTN among MA was determined for each NHANES cycle years by using the NH-White NHANES sample population as the standard population. The age-specific prevalence of HTN in each 10-
30
year age group of the MA sample population was multiplied by the proportion of NH- White sample population in that age group resulting in the estimated age-adjusted
prevalence of that MA age group. The age-adjusted HTN prevalence for MAs is shown in Figure 1 (dashed line) and was similar to the prevalence of HTN in NH-Whites. A linear regression analysis for the age-adjusted HTN prevalence showed that the HTN
prevalence trend from 1999 to 2016 did not significantly increase (P = 0.229). However, from 2005 to 2016 there was a significant increasing trend in the HTN prevalence for MAs (P = 0.014) but no significant trend upwards for NH-Whites (P = 0.84). In the most current NHANES cycle (2015-2016) the age-adjusted HTN prevalence was higher among MAs (51%) than in NH-Whites (48.6%). In addition, the age-adjusted HTN prevalence of male MAs (52.6%) was greater than in NH-White males (50.3%), MA females (49.7%) also had a higher age-adjusted HTN prevalence than NH-White females (47%).
Figure 1. Prevalence of hypertension among adult Mexican Americans and Non-Hispanic Whites by years.
Shows the prevalence of hypertension (± standard error bars) for non-Hispanic whites and Mexican Americans. 0 10 20 30 40 50 60 '99-'00 '01-'02 '03-'04 '05-'06 '07-'08 '09-'10 '11-'12 '13-'14 '15-'16 P rev alen ce o f Hy p er ten sio n ( %) Years
Table 2. Odds of hypertension prevalence among Mexican Americans and Non-Hispanic Whites, NHANES
1999-2016
Mexican Americans
Odds Ratio (95% CI)
Non-Hispanic Whites
Odds Ratio (95% CI)
Gender (ref = Females)
Male 2 (1.8-2.3) 1.4 (1.3- 1.5)
Age Group (ref = 20-39 years)
40-59 years 3.5 (2.9- 4.1) 3.3 (3- 3.7)
60+ years 16.3 (13.7- 19.4) 10.9 (9.8- 12.2)
BMI (ref= Normal (18.5-24.5 kg/m2))
Overweight (25-29.9 kg/m2) 1.6 (1.3- 2) 1.4 (1.3- 1.5)
Obese (30+ kg/m2) 3.2 (2.6- 3.9) 2.6 (2.4-2.9) Abdominal Obesity (ref= Normal WHtR)
WHtR >0.5 1 (0.8- 1.3) 1.4 (1.2- 1.5)
Acculturation Score * (ref= lowest (0))
Low (1) 1 (0.8- 1.2) -
High (2) 1.3 (1- 1.6) -
Highest (3) 1.5 (1.2- 1.7) -
* Acculturation Score is the sum of the following questions ranging from lowest (0) to highest
(3): country of birth (1 point if born in United States, 0 if born outside United States), years living in the United States (1 point if lived in United States >20 years, 0 points if lived <20 years in United States), and language spoken at home (1 point if only or mostly English or both Spanish and English equally were spoken at home, 0 points for Spanish).
- Acculturation score was not determined for NH-Whites due to limited answers to acculturation question
Using multivariate logistic regression analysis, Table 2 shows the odds or likelihood of having HTN stratified by risk factors associated with HTN. Overall, males were two times more likely than females to have HTN in MAs and 1.4 times more likely
32
in NH-Whites. The likelihood of HTN substantially increased with increasing age. Those MAs in the > 60 year age group were over 16 times more likely to have HTN, while NH- Whites were about 11 times more likely to have HTN compared to younger adults. The likelihood of HTN was 3.2 and 2.6 times greater in obese compared to normal weight MAs and NH-Whites, respectively. NH-Whites who were abdominally obese had a 1.4 times higher risk of HTN than those without abdominal obesity. However, abdominal obesity was not associated with greater risk for HTN in MAs. A higher acculturation score resulted in a higher likelihood for MAs to have HTN with an odds ratio (OR) of 1.5 among the highest acculturation group compared to those in the lowest acculturation group.
Table 3 and 4 shows the prevalence of risk factors of HTN from 1999-2016. These tables were created to identify possible contributing causes for the increase in HTN prevalence in MAs. There was no significant change in the gender proportions in either MAs or NH-Whites. The prevalence within the young adult group was decreasing while the older adult age group was increasing in both ethnic groups. However, the prevalence within the middle-age adult group was increasing among MAs but not among NH-whites. The prevalence within the obese BMI category significantly increased over time for both ethnic groups, but the prevalence of obesity was higher among MAs (40%) than among NH-Whites (34%). The prevalence within the overweight BMI category did not
significantly change for either MAs or NH-whites. For both MAs and NH-Whites, the prevalence of abdominal obesity increased over time, but the prevalence was higher in MAs. The percentages of those MAs in the lowest (0 score) acculturation score
34
Table 3. Prevalence of risk factors of hypertension among Mexican Americans, NHANES 1999-2016
’99-‘00 ‘01-‘02 ‘03-‘04 ‘05-‘06 ‘07-‘08 ‘09-‘10 ‘11-‘12 ‘13-‘14 ‘15-‘16 Average P-value* Gender (% ± Std. Error) Males 54±1.7 55.3±1.4 55.5±2 56.4±1.1 56±2.6 54.5±0.9 53.1±1.9 53.1±2 51.7±1.8 54.3±0.6 0.054 Females 46±1.7 44.7±1.4 44.5±2 43.6±1.1 44±2.6 45.5±0.9 46.9±1.9 46.9±2 48.3±1.8 45.7±0.6 0.054 Age (% ± Std. Error) 20-39 years old 57.7±3.1 62.8±2.3 58.8±4.5 56.2±2.6 55.8±2.9 50.5±2.6 55±2.4 50.1±3.2 48.4±3.1 54.4±1.1 0.002 40-59 years old 31.8±2.3 27.8±1.3 30.8±2.3 32.5±2 32.6±1.8 35.6±1.3 35±2 35.9±2.5 35.8±1.6 33.4±0.7 0.003 60+ years old 10.5±1.2 9.4±1.9 10.3±2.7 11.3±1.1 11.6±1.6 13.9±1.7 10±1.6 14±1.4 15.8±3.1 12.2±0.7 0.015 BMI (% ± Std. Error) Normal (18-24.9 kg/m2) 28.7±2.4 31.8±2.4 25.8±3.2 26.1±1.9 22.3±2.3 20±1.7 20.6±2.9 17.4±1.4 16.8±1.7 22.5±0.8 <0.001 Overweight (25-29.9 kg/m2) 38.1±1.9 39.6±1.5 37.9±1.4 41.1±1.8 38.6±2.5 40.5±1.5 33.8±2.1 36.9±1.8 34.2±2 37.7±0.6 0.100 Obese (30+ kg/m2) 33.2±2.6 28.7±2 36.3±2.6 32.7±1.2 39.1±3.9 39.5±1.9 45.6±2.4 45.6±2.4 49±1.7 39.8±0.7 <0.001 Abdominal Obesity (% ± Std. Error) Waist-to-Height Ratio >0.5 83.1±1.8 77.4±1.8 82.6±2.4 83.7±1.8 85±1.6 87.3±1.2 85.4±2.4 85.6±1.9 87.4±1.2 84.5±0.6 0.016 Acculturation Score (0-3) Lowest (0) (%± Std. Error) 34.5±3.3 44.5±4.2 36.5±6.4 41.4±4.3 35.6±4.5 31.6±4 34.5±3.6 25.3±4.7 25.6±3.4 33.7±1.6 0.016 Low (1) (%± Std. Error) 17.5±1.5 13.5±2 17±2 18±1.8 16.7±1.7 23.5±1.8 15.2±1.6 21.2±1.2 20.7±1.3 18.4±0.6 0.125 High (2) (%± Std. Error) 10.7±1.3 7.1±0.9 11.7±2 7.2±1.1 10.6±1.4 12.6±2.1 8.1±1.2 13.6±1.9 12.6±2.6 10.7±0.7 0.201
Table 3. Prevalence of risk factors of hypertension among Mexican Americans, NHANES 1999-2016 (Continued) Acculturation Score (0-3)
Highest (3) (%± Std. Error) 37.4±3.2 34.9±4.3 34.9±4.3 33.5±3 37.2±3 32.3±3 42.2±4.3 39.8±5.2 41.1±3.2 33.2±1.3 0.110 * P-value determined by linear regression analysis over the years
Table 4. Prevalence of risk factors of hypertension among Non-Hispanic Whites, NHANES 1999-2016
’99-‘00 ‘01-‘02 ‘03-‘04 ‘05-‘06 ‘07-‘08 ‘09-‘10 ‘11-‘12 ‘13-‘14 ‘15-‘16 Average P-value* Gender (% ± Std. Error) Males 49.5±1.1 49.9±0.6 48.8±1 49.7±0.7 48.4±0.7 49.1±0.6 49±1.2 49.1±0.9 49.2±0.9 49.2±0.3 0.317 Females 50.5±1.1 50.1±0.6 51.2±1 50.3±0.7 51.6±0.7 50.9±0.6 51±1.2 50.9±0.9 50.8±0.9 50.8±0.3 0.317 Age (% ± Std. Error) 20-39 years old 38.9±1 34.5±1.7 33.5±1.5 31.9±1.3 32.6±1.5 31.4±1.3 30.4±2.6 30.3±1.2 30.3±1.8 32.5±0.5 0.002 40-59 years old 35.8±1 41.7±1.5 40.3±1.3 40.9±1.7 39.9±1.3 39.5±0.7 39.1±1.6 37.5±1.1 36.5±1.6 39.1±0.5 0.169 60+ years old 25.3±1.4 23.8±1.2 26.2±1.2 27.2±2.3 27.5±1.2 29.1±1.2 30.5±1.4 32.2±1.3 33.1±2.1 28.4±0.5 <0.001 BMI (% ± Std. Error) Normal (18-24.9 kg/m2) 36.4±2.2 34.5±0.8 33.9±1.5 32.7±1.7 31.5±1.3 30.8±1.6 31.1±1.8 29±0.9 28±1.5 31.9±0.5 <0.001 Overweight (25-29.9 kg/m2) 34.4±1.3 35.1±1.3 34.7±1 33.6±1.2 35.3±1.1 33.8±1.5 35.5±1.6 33.6±1 33.4±0.4 34.4±0.4 0.316 Obese (30+ kg/m2) 29.2±1.8 30.5±1.2 31.4±1.3 33.7±1.7 33.2±1.7 35.5±1.2 33.4±1.8 37.4±1.1 38.6±0.5 33.8±0.5 <0.001 Abdominal Obesity (% ± Std. Error)
36