Masked Hypertension in Low‐Income South African Adults

Abstract

While South Africa has one of the highest hypertension rates globally, there are few data on masked hypertension (MHT) and white‐coat hypertension (WCHT). This study measured the frequency of MHT and WCHT in low‐income (<$500 US per month) South African adults, evaluating cardiovascular risk by arterial stiffness. Participants (n=101, 50% male; mean age 39.4±9.7 years) were recruited from a large North‐West Province employer. Clinic and 24‐hour blood pressure (BP) and pulse wave analysis were recorded. Clinic BP identified 18% of patients as hypertensive, while 24‐hour BP showed that 63% of patients were hypertensive. The frequency of MHT was high (33 of 81, 41%) with only one case of WCHT. In comparison to those with normal clinic and 24‐hour BP, augmentation index and pulse wave velocity were significantly higher in those with hypertensive 24‐hour BP irrespective of clinic BP, indicating that, in this group, masked and sustained hypertension carry a similar elevated cardiovascular risk.

Cardiovascular disease (CVD) and stroke account for almost one third of all deaths globally, with over half of these deaths (more than 9 million per year) associated with hypertension (HT).1 South Africa currently has one of the highest reported rates of HT in adults older than 50 years (78%),2 although true prevalence of HT is unknown because clinic blood pressure (BP) may not reflect true BP.3, 4 Some authors suggest that when true HT prevalence in the population is around 50%, clinic BP would give an accurate diagnosis in only 61% of cases5 due to conditions such as white‐coat HT (WCHT) and masked HT (MHT).6, 7 The prevalence of WCHT is reported to be 15% to 25% in patients diagnosed with HT by clinic BP,8 while 15% to 30 % classified as normotensive by clinic BP may have MHT.9

Determining the prevalence of these conditions requires BP measurement outside of the clinic environment, ideally by 24‐hour ambulatory BP monitoring (ABPM). In South Africa, ABPM is uncommon, especially in public healthcare facilities used by the majority of the population (over 80%) with no medical aid coverage.10 As income and insurance status are generally related to HT diagnosis and treatment probability,11 we sought to determine the frequency of WCHT, MHT, and sustained HT (HT) in low‐income adults by comparing clinic BP and 24‐hour ABPM. We also measured 24‐hour pulse wave velocity (PWV) to assess arterial stiffness as an independent risk factor for cardiovascular morbidity and mortality.12, 13

Methods

Study Population

We conducted a cross‐sectional study (April 2013–October 2014), recruiting a convenience sample of 101 participants (50% male; 96% black, 4% mixed race) from a population of approximately 300 service workers (cleaning, gardening, security, or catering services) at an education facility in the North West Province of South Africa. Inclusion criteria were: full‐time employment, net income <R6000 (~$500 USD) per month, and working age (18–65 years). Exclusion criteria were: pregnancy; breast feeding; a current diagnosis of tuberculosis, malaria, chronic kidney disease, or cancer; or sick on the study day. The salary level for inclusion was selected to include participants who earned around the minimum wage (R15.47 or USD $1.22 per hour).14

Trained researchers who spoke the participants' home language (predominantly Setswana) explained the study and all participants provided written informed consent prior to taking part. The study complied with the ethical principles for medical research involving human subjects as stated in the Declaration of Helsinki15 and was approved by the Health Research Ethics Committee of the North‐West University. All measures were taken in a dedicated HT research facility during the participant's lunch break with lunch provided following the measures and smoking and caffeinated drinks prohibited during the period.

Cardiovascular Measures

Clinic BP was measured after 10 minutes of rest, with participants seated and their arm supported at heart level. We measured brachial systolic BP (SBP) and diastolic BP (DBP) on the right upper arm using the validated M10‐IT automated device (Omron Healthcare, Tokyo, Japan). With 5 minutes of rest between each measure, the measure was repeated twice on the left upper arm and once more on the right upper arm. The first measure on each side was disregarded and the mean of the second measures was used for analysis. All clinic BP measures were taken in a private room at the research facility, with the researcher present.

Participants were then fitted with a 24‐hour ABPM and pulse wave analysis (PWA) monitor (Mobil‐O‐Graph, I.E.M. GmbH, Stolberg, Germany) with an appropriate cuff for mid‐upper arm circumference (small, 20–24 cm; medium, 24–32 cm; large, 32–38 cm). The Mobil‐O‐Graph has been validated against aortic PWV from intra‐aortic catheterization16 and meets the European Society of Hypertension (ESH)17 and the Association for the Advancement of Medical Instrumentation criteria for BP measurement.18 Furthermore, the brachial oscillometric method used for PWA, PWV, and augmentation index is shown to produce similar results to the radial tonometric method of the SphygmoCor device (AtCor Medical, West Hyde, Australia).19

BP and PWA were measured every 15 minutes during the awake period and every 30 minutes during the sleep period. Awake/sleep periods (hereafter referred to as day/night) were determined by the participant's reported habitual sleep and wake times to account for earlier and later day shift patterns. No participants recorded 24‐hour ABPM while on an overnight shift. Participants were advised to avoid planned exercise while wearing the device and to report periods of physical activity that occurred during the period using a diary provided. Successful 24‐hour recordings were assessed using both the ESH20 and ARTEMIS (www.artemis.org) criteria, ie, interval between measurements not exceeding 30 minutes, at least 70% of expected number of readings obtained, at least 40 readings obtained over 24 hours, no more than 2 hours with missing readings, and no consecutive hours with missing readings. If ABPM recordings did not meet all of these criteria, the participant was asked to repeat the measure.

Hypertension definition followed ESH guidelines with one or more of the following: mean clinic SBP and/or DBP ≥140/90 mm Hg; 24‐hour ABPM SBP and/or DBP ≥130/80 mm Hg; day ABPM SBP and/or DBP≥135/85 mm Hg; and night ABPM SBP and/or DBP ≥120/70 mm Hg.20 Based on their readings, participants were then classified into one of the following categories: sustained NT (below the diagnostic threshold for HT by all measures); WCHT (hypertensive by clinic BP and normotensive by ABPM); MHT (hypertensive by ABPM and normotensive by clinic BP); and HT (hypertensive by both ABPM and clinic BP).

Of the participants currently taking antihypertensive medication (n=10, indicating a previously elevated clinic BP and diagnosis), 80% had uncontrolled BP by ABPM and were categorized as having sustained HT while those with successfully treated HT, ie, normotensive by both ABPM and clinic BP (n=2) were removed from the analysis (Figure 1). Human immunodeficiency virus (HIV) testing by trained counselors followed the South African Provider Initiated Counseling and Testing protocols.21

Figure 1.

Study flow diagram. ABPM indicates ambulatory blood pressure monitoring; BP, blood pressure.

Anthropometric Measures

Height was measured without shoes to the nearest 0.1 cm and weight was measured in light clothing to the nearest 0.01 kg (SECA stadiometer and portable electronic scales, Hamburg, Germany). Waist circumference was measured in triplicate midway between the lowest rib and the suprailiac crest (Lufkin steel anthropometric tape, Apex, NC) and the median value was used for analysis. Body mass index (BMI) and waist‐to‐height ratio (WHtR) were calculated using the formula: BMI=weight (kg)/height (m)² and WHtR=waist circumference (cm)/height (cm), respectively.

Questionnaires

The South African National Income Dynamics Survey (NIDS) Wave 1 questionnaire22 was used to obtain data on socioeconomic status, health, lifestyle (including smoking and alcohol consumption), and education and included the Center for Epidemiologic Studies Short Depression Scale (CES‐D 10).23 Data were collected on stress using the Perceived Stress Scale (10‐item PSS).24 Both the depression and stress questionnaires have previously been used in a low‐income South African population, with Cronbach‐alpha scores of 0.88 and 0.72, respectively.25 Self‐reported physical activity was assessed using the World Health Organization's (WHO's) recommended General Physical Activity Questionnaire (GPAQ), shown previously within South Africa to have good test‐retest reliability.26

Statistical Methods

We performed statistical analyses using SPSS version 21 (IBM, Armonk, NY). Differences between men and women for continuous variables were analyzed using an independent t test for normally distributed data and the Mann‐Whitney U‐test for non‐normally distributed data. Hypertensive groups were compared using analysis of variance and analysis of covariance, and post hoc analysis conducted using the Bonferroni correction. PWV was adjusted for mean arterial pressure (MAP) and age, and augmentation index was adjusted for heart rate and height. Chi‐square test and Fisher exact test (for small samples) were used for categorical variables. All analyses were considered statistically significant at P<.05.

Results

Participant Characteristics

Of the 101 participants recruited, 15% (n=15) were excluded because of nonvalid ABPM measures, 5% were withdrawn, and 80% (n=81) completed the study and achieved successful ABPM recordings (Figure 1). Repeat ABPM measures were required in 21 participants. There was no difference between patients excluded for nonvalid ABPM and those included in the analysis in terms of sex, age, and HT medication use or HT status by clinic or ambulatory BP. However, the BMI of those included was lower than those excluded (27.3±7.1 compared with 33.3±8.2; P=.002), reflecting the difficulties in obtaining successful ABPM readings in individuals with increased upper arm adiposity even with appropriately sized cuffs.

Table 1 shows the baseline characteristics of the included participants (41 men, 40 women). Overall, 74% (n=60) of participants were HIV negative, 14% (n=11) were HIV positive, and 12% (n=10) chose not to be tested. HIV test outcome did not differ by sex, and BP (clinic and 24‐hour) was not different between the HIV‐negative and HIV‐positive groups. All obesity indices were significantly greater in women than men (all P≤.001), but there were no significant differences between men and women in clinic BP, ABPM, or subsequent BP categorization. Current smokers were exclusively men, and 49% of men and only 5% of women consumed alcohol 1 day per week or more (P≤.001). Women were primarily employed in cleaning services and men in gardening/landscaping services, with the monthly income greater for men than for women (P=.003). More than one third of both men and women had high stress scores (PSS score ≥20; 44% of men, 38% of women) and/or high depression scores (CES‐D 10 score ≥10; 41% of men, 38% of women).

Table 1.

Sociodemographic Characteristics of the Study Group (n=81) by Sex

	Men	Women	P Value
No. (%)	41 (51)	40 (49)
Age, y	38.2±11.0	39.8±8.3	.459
Population group, No. (%)
Black	39 (95)	39 (98)	.509
Mixed race	2 (5)	1 (2)
HIV‐infected	6 (15)	5 (13)	.761
Anthropometry
Weight, kg	65.3±12.3	81.1±16.0	≤.001
Height, cm	169.8±7.3	158.8±4.4	≤.001
BMI, kg/m2	22.6±3.7	32.2±6.4	≤.001
Obese (BMI ≥30), No. (%)	3 (7)	25 (63)	≤.001
Waist circumference cm, median (IQR)	79.2 (14.4)	99.4 (18.7)	≤.001
Waist‐to‐height ratio, median (IQR)	0.46 (0.08)	0.62 (0.13)	≤.001
Cardiovascular measures, mm Hg
Clinic SBP	122±11.9	117±13.1	.059
Clinic DBP	78±9.7	81±10.3	.186
24‐Hour SBP	127±11.0	125±11.5	.353
24‐Hour DBP	81±8.8	81±8.0	.973
Antihypertensive medication, No. (%)	3 (7)	7 (18)	.146
Socioeconomic indicators
Net monthly income, ZAR, median (IQR)	2400 (1305)	2100 (297)	.003
Employment type No. (%)a
Elementary occupations, eg, cleaner	8 (20)	36 (90)	≤.001
Service workers, eg, cook, shop worker	4 (10)	4 (10)
Agricultural workers, eg, landscaper	21 (51)	0 (0)
Craftsmen and machine operators	5 (12)	0 (0)
Technicians and clerks	3 (7)	0 (0)
Previously unemployed	17 (42)	17 (45)	.613
Own a motor vehicle	9 (22)	4 (10)	.143
Own a bicycle	28 (68)	3 (8)	≤.001
Finished high school (matric)	16 (40)	15 (50)	.405
Smokers	20 (49)	0 (0)	≤.001
Alcohol consumption ≥1 d/wk	20 (49)	2 (5)	≤.001
Moderate/vigorous PA, min/wk	2141±1189	1889±944	.295

Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; HIV, human immunodeficiency virus; IQR, interquartile range; PA, physical activity (self‐report); SBP, systolic blood pressure; ZAR, South African Rand. Data are presented as mean±standard deviation unless otherwise indicated. P values are for comparison between men and women. ^aEmployment coding based on South African Standard Classification of Occupations 2001.

HT Diagnosis and Classification by Clinic and Ambulatory BP

Among the 71 participants not currently taking antihypertensive medication, 18% (n=13) were considered hypertensive by clinic BP (Table 2). In comparison, 64% (n=45) were considered hypertensive by ABPM, significantly more than by the clinic BP measure (P=.017). The overall level of agreement between clinic and ambulatory BP was poor (κ=0.181, P=.017). Classification using clinic and 24‐hour BP, and participant reports of medication use, resulted in the following groups (Figure 2): NT (n=25); MHT (n=33); and HT (n=20). WCHT (n=1) and treated controlled HT (n=2) were not analyzed further because of the small group size. Among the MHT group, 58% (n=19) were diagnosed as hypertensive by all three ABPM criteria: 24‐hour, day, and night BP.

Table 2.

Hypertension Status by Clinic BP and ABPM in Participants With Valid 24‐Hour ABPM Measurement Excluding Patients on Treatment (n=71)

		ABPM Hypertension Status		Total
		Normotensive	Hypertensivea
Clinic BP hypertension status, No. (%)	Normotensive (≤140/90)	25 (35)	33 (47)	58 (82)
	Hypertensive (≥140/90)	1 (1)	12 (17)	13 (18)
Total, No. (%)		26 (36)	45 (64)	71

^aESH ambulatory blood pressure monitoring (ABPM) hypertension criteria used to determine hypertensive status: 24‐hour ≥130/80 mm Hg; and/or day ≥135/85 mm Hg; and/or night ≥120/75 mm Hg.

Figure 2.

Twenty‐four–hour ambulatory blood pressure measures in patients with normotension (NT, n=25), masked hypertension (MHT, n=33), and sustained hypertension (HT, n=20). The shaded area shows the nighttime period. Results of Bonferroni post‐hoc analysis: ^aboth hypertensive groups significantly different than the NT group; ^bsustained hypertensive group significantly different than the MHT and NT groups; ^csignificant difference between all three groups; ^dHT group significantly different than the NT group; and ^eMHT group significantly different than the NT group; all P<.05.

Comparison of Cardiovascular Risk Factors

Patients in the NT group were younger than patients in both hypertensive groups (P=.001; Table 3). Waist circumference, BMI, and WHtR were higher in patients in the HT group (P≤.025). No differences were observed between the groups in smoking, alcohol consumption, self‐reported physical activity levels, stress, or depression.

Table 3.

Characteristics of Participants Classified as Having NT, MHT, and HT

	NT (n=25)	MHT (n=33)	HT (n=20)	P Valuea
Age, y	32.8±6.6	41.4±10.8	41.6±7.3	.001b
Men, No. (%)	14 (56)	18 (55)	8 (40)	.501
Body mass index, kg/m2	26.3±6.5	25.7±5.8	30.9±8.7	.024c
Waist circumference, cm	86.0±14.9	86.6±13.9	97.3±15.5	.020d
Waist‐to‐height ratio	0.52±0.10	0.53±0.10	0.60±0.11	.025d
PSS score	19±9	17±8	18±9	.665
High stress levels (PSS score ≥20)	13 (52)	13 (37)	7 (35)	.405
CESD‐10 score	10±6	10±5	10±8	.877
High depression levels (CESD‐10 ≥11)	10 (40)	13 (40)	8 (41)	.998
Smokers, No. (%)	8 (32)	8 (24)	3 (15)	.418
Alcohol consumption ≥1 d/wk, No. (%)	6 (24)	10 (30)	5 (25)	.997
Estimated units of alcohol per week	17±10	23±9	13±8	.954
Work‐related MVPA, min/de	375±167	284±215	270±207	.139
Travel‐related MVPA, min/de	60±56	56±63	63±76	.920
Leisure‐related MVPA, min/de	14±34	8±19	4±10	.352
Meeting PA guidelines, No. (%)	25 (100)	33 (100)	20 (100)	–
Previous hypertension diagnosis, No. (%)	0 (0)	2 (6)	13 (65)	≤.001
Antihypertensive medication, No. (%)	0 (0)	0 (0)	8 (40)	≤.001

Abbreviations: CESD‐10, Center for Epidemiologic Studies Short Depression Scale; PA, physical activity; PSS, Perceived Stress Scale. Data are presented as mean±standard deviation unless otherwise indicated. ^aAnalysis of variance and chi‐square tests. ^bSustained hypertension (HT) and masked hypertension (MHT) significantly different than normotension (NT) according to results of Bonferroni post hoc analysis. ^cSignificant difference between MHT and HT according to results of Bonferroni post hoc analysis. ^dHT significantly different to MHT and NT according to results of Bonferroni post hoc analysis. ^eModerate vigorous physical activity (MVPA) by Global Physical Activity Questionnaire, with work based on a 5‐day week and travel and leisure on a 7‐day week.

Clinic BP was highest in the patients with HT, followed by those with MHT, and lowest in patients with NT (P≤.001; Table 4). Comparing the 24‐hour measures, SBP, DBP, cSBP, and MAP followed the same pattern as clinic BP, ie, highest in patients with sustained HT, then those with MHT then those with NT (P<.001), although there was no difference between the groups in pulse pressure, heart rate, or dipping status.

Table 4.

Comparison of 24‐Hour BP and Arterial Stiffness Between Participants Classified as Having NT, MHT, and HT

	NT (n=25)	MHT (n=33)	HT (n=20)	P Valuea
Clinic BP, mm Hg
Systolic	111±9	118±10	131±13	≤.001b
Diastolic	72±8	78±6	90±11	≤.001c
Ambulatory BP, mm Hg
SBP, 24‐hour	118±6	128±9	136±10	<.001c
SBP, day	120±6	131±10	139±11	<.001c
SBP, night	108±5	118±9	124±11	<.001c
DBP, 24‐hour	74±4	83±6	89±8	<.001c
DBP, day	77±4	86±7	93±8	<.001c
DBP, night	63±5	72±5	77±10	<.001c
PP, 24‐hour	44±5	45±10	47±8	.464
PP, day	43±6	45±10	46±9	.484
PP, night	45±5	46±8	47±8	.403
MAP, 24‐hour	94±4	104±6	111±8	<.001c
MAP, day	97±4	107±6	114±8	<.001c
MAP, night	83±4	93±6	99±10	<.001c
cSBP, 24‐hour	108±5	120±8	127±9	<.001c
cSBP, day	110±5	122±9	129±9	<.001c
cSBP, night	102±6	113±9	118±10	<.001c
Heart rate, 24‐hour, beats per min	79±9	80±9	82±9	.668
Heart rate, day	83±10	83±10	85±10	.837
Heart rate, night	66±7	68±9	71±10	.082
AIx, 24‐hour, %d	23.6±6.0	26.5±6.0	27.9±6.0	.047
AIx, day	25.7±6.0	27.8±6.0	28.7±6.0	.222
AIx, night	16.2±9.9	22.7±9.9	25.0±9.9	.009f
PWV, 24‐hour, m/se	5.5±0.1	6.6±0.2	6.8±0.2	<.001f
PWV, day	5.6±0.1	6.7±0.3	6.8±0.2	<.001c
PWV, night	5.3±0.2	6.4±0.3	6.4±0.2	<.001f
Nondippers, SBP, No. (%)g	11 (44)	16 (49)	8 (40)	.653
Nondippers, DBP, No. (%)g	0	6 (18)	6 (30)	.055

Abbreviations: BP, blood pressure; cSBP, central systolic blood pressure; DBP, diastolic blood pressure; PP, pulse pressure; SBP, systolic blood pressure. Data are presented as mean±standard deviation unless otherwise indicated. ^aAnalysis of variance, analysis of covariance, and chi‐square tests. ^bSustained hypertension (HT) significantly different than masked hypertension (MHT) and normotension (NT) according to results of Bonferroni post hoc analysis. ^cSignificant difference between all three groups according to results of Bonferroni post hoc analysis. ^dAugmentation index (AIx) adjusted for heart rate and height. ^ePulse wave analysis (PWV) adjusted for mean arterial pressure (MAP) and age. ^fHT and MHT significantly different than NT according to results of Bonferroni post hoc analysis. ^gNondipping classified as change from day to night BP of ≤10%.

Augmentation index adjusted for height and heart rate was higher in patients in the MHT and HT groups compared with those in the NT group, especially during the night period (P=0.009). PWV adjusted for MAP and age was also significantly higher in patients in the MHT and HT groups than in the patients in the NT group (P<.001). Sensitivity analysis showed that removal of the 4% mixed‐race group did not alter BP or arterial stiffness results.

BP profile over the 24 hours (Figure 2) showed that patients in the MHT and NT groups were more similar during the afternoon and evening, but BP in patients in the MHT group increased during the night and morning to values more similar to the patients in the HT group. However, the PWV profile of the groups over the 24‐hour period adjusted for MAP and age showed that the MHT group had a PWV profile consistently more similar to those in the HT group (Figure 3).

Figure 3.

Twenty‐four–hour pulse wave velocity (PWV) measures in patients with normotension (NT, n=24), masked hypertension (MHT, n=33), and sustained hypertension (HT, n=20). PWV adjusted for age and group mean arterial pressure. Shaded area shows the nighttime period. Results of Bonferroni post‐hoc analysis: ^aboth hypertensive groups significantly different than the NT group, and ^dsignificant difference between all three groups; all P<.05.

Discussion

This study sought to determine the frequency of MHT, sustained HT, and WCHT in a sample of South African adults employed in low‐income occupations. The finding that more than 40% of the group (33 of 81) had MHT was surprising and concerning, especially because our arterial stiffness data indicate that MHT carries a similar CVD risk to overt HT. However, as clinic BP is within the normal range, adults with MHT are unlikely to receive clinical diagnosis and treatment.

A similar prevalence of MHT (45%) has been reported in African Americans,27 and in African Americans with hypertensive kidney disease (43%).28 Among this renal‐impaired MHT group, more cases showed uncontrolled nighttime ABPM (94%) than uncontrolled daytime ABPM (57%). In contrast, we observed similar numbers of uncontrolled day and nighttime ABPM in the MHT group, with most MHT cases having uncontrolled day, night, and 24‐hour ABPM, although renal function was not assessed. The overall prevalence of MHT has previously been shown to be twice as high in black Americans compared with white Americans,29 although it is unclear to what extent biological, psychosocial, and economic factors play a role. While MHT and HT have been reported to share several risk factors (age, male sex, BMI), within individuals these factors do not change between the in‐clinic and out‐of‐clinic BP measurements, suggesting factors outside the clinic such as physical activity, smoking, alcohol use, and psychological stress associated with work may play a role.9 While we did find a significant difference in BMI between the HT and MHT group, we found no difference between the groups in these behavioral or psychological measures. Ogedegbe and colleagues30 found lower state anxiety in their MHT patients compared with WCHT patients, suggesting that MHT may present with a calmer state of mind during clinic BP measurement.

In more general comparisons, others investigators have found that BP at work is generally higher than that at home, with HT occurring more frequently in those with low occupational status, low levels of autonomy combined with a high workload (job strain), and effort‐reward imbalance (high effort and low reward).31, 32 A large study (n=2370) among white‐collar employees in a high‐income country found a lower prevalence of MHT (15%).33 Although, it is unclear whether this can be compared with manual labor low‐income employees in an upper‐middle–income country since poverty is generally associated with multiple stressors, elevated psychosocial stress, and increased BP already observable in adolescence.34 While this low‐income group earned around the minimum wage, they did remain above the upper bound poverty line (R779 per person per month, or around $5 per day).14, 35 The high levels of stress and depression observed here and by others in a similar population25 may contribute to the high levels of MHT observed.

Our finding that the HT group had higher levels of obesity compared with the other groups is similar to evidence from young adults where the prevalence of HT (but not MHT) increased as obesity increased from overweight to moderately obese to severely obese.36 Obesity may worsen or maintain HT so that it is evident throughout the day, including during clinic BP measures, possibly through upregulated sympathetic nervous system activity.37 The similarity between the HT and MHT BP profile over the 24‐hour period, with similar nondipping levels in the MHT and HT groups, may also suggest a role for elevated sympathetic activation during the night and early morning period. Increased nocturnal sympathetic activation is associated with an attenuated fall in nighttime epinephrine and norepinephrine38 and an increased CVD risk compared with those with a normal dipping pattern.39

A prominent finding from our study is that 24‐hour arterial stiffness as reflected by PWV was similar between the MHT and HT groups, indicating similar levels of CVD risk. While it is difficult to draw conclusions on CVD risk comparing these groups cross‐sectionally, longitudinal research also supports our findings, showing that MHT was associated with a similar CVD mortality and stroke morbidity to sustained HT over a 10‐year period.40 Furthermore, other authors have suggested that MHT carries a greater risk of mortality, as a result of the potential for poor BP control.41

Our findings are in contrast with other research indicating that PWV is lower in patients with MHT than those with HT.42 In our study, PWV was similar between these two groups, suggesting similar cardiovascular risk in patients with MHT and those with HT. Overall, the PWV values measured with the Mobil‐O‐Graph device were lower than may be expected when compared with the reference ranges for carotid femoral PWV as obtained with the SphygmoCor device.43 Luzardo and colleagues19 also reported lower PWV values obtained with this technique compared with radial tonometry, and further work is needed to determine the reason for this.

Study Strengths and Limitations

The limitations of this study include the cross‐sectional design, the convenience sampling approach used, and the small sample size. A strength of this study is the homogeneity of the population in terms of income and employment status. Data obtained from the nationally representative South African NIDS Survey (n=3852)44 show an average net income of R2397 per month, with 66% (n=2554) of participants earning less than R6000 per month in full‐time employment. Of this group, most (88%) were employed in similar occupations as those described in this article and 28% had a clinic BP of ≥140/90 mm Hg. As such, our study sample appears aligned with a large proportion of the South African adult population. With recent data showing that 25% of the adult population are unemployed,45 and unemployment in itself reported to be an independent predictor of HT,46 the situation in South Africa may well be even worse than our data would suggest.

Although this study was small, our findings suggest that HT prevalence estimates for South Africa may underestimate the magnitude of the problem. Our clinic BP measures recorded in the research facility could be lower than those recorded in primary care, although the similarity between our HT prevalence by clinic BP or HT medication use (31.6%, 31 of 98) and the WHO‐reported prevalence for South African adults (33.7%)47 would suggest this is not the case. Therefore, given the observed 24‐hour BP findings, our results may support an underestimation of HT prevalence when relying on clinic BP measures alone. This not only has serious clinical implications for HT diagnosis and treatment, but also for future CVD and stroke rates. A larger longitudinal study in the country is needed in order to verify these findings.

Conclusions

Our results suggest that current hypertension prevalence estimates may be an underestimate because of the potential high incidence of masked hypertension in the region. This is of particular concern as arterial stiffness, as indicated by PWV, suggests a similar elevated CVD risk in the masked hypertension group to adults with hypertension observable by clinic BP. Further research is needed to determine the prevalence and impact of masked hypertension in sub‐Saharan Africa.

Author Contributions

LW, AES, and KR designed the research; LW, TN, and LG conducted the research; LW, KR, and JT performed the statistical analysis; LW, KR, AES, JVR, and RS wrote the paper; and AES had primary responsibility for final content of this paper. All authors read and approved the final manuscript.

Disclosure

This work was funded by the North‐West University and the National Research Foundation (NRF) of South Africa. The funding source had no involvement in the conduct of the research or preparation of the article. None of the authors have any conflicts of interest to declare.

Disclaimer

The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of the National Research Foundation (NRF) or other funding bodies. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and therefore the NRF does not accept any liability in regard thereto.

J Clin Hypertens (Greenwich). 2016;18:396–404. 10.1111/jch.12768. © 2016 Wiley Periodicals, Inc.