Prevalence and Therapeutic Control of Hypertension in 30 000 Subjects in the Workplace
—To assess blood pressure (BP) control in a French working population through the use of a careful assessment of BP based on 2 different visits in 1 month, 17 359 men and 12 267 women were evaluated from January 1997 to April 1998. The initial phase was a cross-sectional analysis of a cohort study designed to assess the incidence of arterial hypertension in a French working population. Information was collected by the work-site physician during the annual examination. BP was measured with a validated automatic device. Among subjects with BP ≥ mm Hg, patients not treated with antihypertensive drugs were invited to have an additional BP measurement taken 1 month later. The prevalence of hypertension (BP ≥ mm Hg) based on 2 visits was 16.2% in men and 9.4% in women. When the diagnosis of hypertension was based on 2 visits, its prevalence was 41% lower in men and 36% lower in women compared with that of a diagnosis based on a single visit. Accordingly, the awareness of hypertension was 49% higher in men and 40% higher in women. Overall, 12.5% of hypertensive men and 33.2% of hypertensive women taking antihypertensive medication had their BP levels lowered to < mm Hg by treatment. Although the percentage of hypertensive men and women under current treatment who were aware of their hypertension increased with age, BP control among treated subjects decreased with age. Ineffective BP control with treatment accounted for 33% of BP levels ≥ mm Hg in men and 40% of those observed in women. In this large French working population, estimates of hypertension therapeutic control depend heavily on the number of BP measurements. Despite these methodological precautions, insufficient awareness of BP and insufficient BP control through treatment remain 2 major public health problems.
Awareness, treatment, and control of high blood pressure (BP) have been major public health concerns over the past 20 years. In randomized trials, treatment of arterial hypertension has been shown to be quite efficacious in reducing cardiovascular mortality and morbidity.1 In contrast, hypertension control at the population level remains disappointing. In the United Kingdom, only 6% of hypertensive subjects had their BP levels lowered to < mm Hg.2 Among treated hypertensive subjects, BP control was achieved in 27% of the subjects in the United States3,4 and 24% of the subjects in France.5
Measurement of BP has also been identified as a key factor not only for deciding to treat hypertensive patients but also for assessing both efficacy and effectiveness of BP control. More than 30 years ago, Armitage and Rose6 underlined that “multiple readings on the same occasion are easier to obtain, but they yield a smaller improvement in precision than readings obtained on different occasions.” Accordingly, most international guidelines now emphasize that BP should be measured on several different occasions before the diagnosis of hypertension is established or treatment is initiated. However, despite the universal consensus on this subject, estimates of prevalence, awareness, and control of hypertension have largely been based not on several measurements but on a single occasion. National Health and Nutrition Examination Survey (NHANES) III is an exception because 80% of the subjects had their BP measured on 2 separate occasions.4 Logistical difficulties in measuring BP on 2 separate occasions on large samples of the population have explained this limitation. Indeed, previous studies have assessed hypertension control on the basis of 1 visit.2–5,7–14
Thus, the goal of this study was to assess BP control in a French population of workers and to analyze to what extent the results of BP control at the population level are improved if the diagnosis of hypertension is defined on the basis of 2 visits.
This study is based on data from the initial cross-sectional phase of a cohort study designed to assess the incidence of arterial hypertension in the French working population. Fifty-three work-site physicians were recruited on a voluntary basis. Their geographic locations were representative of metropolitan France. Physicians from all workplaces (eg, industry, services, transports) were included, with the exception of the agricultural sector. Because an annual medical examination is mandatory at all French work sites, the design of the cohort study was based on this annual visit to ensure a high follow-up rate. The volunteer physicians agreed to enroll a population of ≥800 wage earners. According to the initial agreement with the work-site physicians, each enrolled either the entire population or a random sample of the working population that he or she was responsible for. Each subject gave informed consent to participate in the protocol.
Information was collected by the work-site physician during the annual examination. BP was measured with a validated automatic device (OMRON 705CP).15 A training session was organized for all work-site physicians to standardize BP measurements. Systolic and diastolic BPs were measured at 5, 6, and 7 minutes with a cuff properly adapted to the arm size, with subjects in the sitting position. The mean of these 3 measurements was used to estimate BP during all visits. Hypertension was defined by current antihypertensive treatment and/or BP ≥ or ≥ mm Hg. For patients not currently treated for hypertension, a diagnosis of hypertension was established in 2 ways: from the results of the first visit only or from the results of a second follow-up visit. This second follow-up visit concerned all untreated persons with a BP ≥ mm Hg on the first visit. This additional BP measurement session, planned 1 month later, could be performed in 78.5% of the invited population. By use of the second criterion, subjects were defined as hypertensive if their BP was ≥ or ≥ mm Hg on each of 2 visits or if they were under current antihypertensive treatment at the first visit.
Height and weight were measured, and body mass index (BMI) was calculated as weight divided by height squared (kg/m2). Obesity was defined by a BMI ≥30 kg/m2.16 Heart rate was noted at 5, 6, and 7 minutes, and the mean of the 3 measurements was used. Tobacco and alcohol consumption, participation in sports activities, educational level, marital status, and health services use were obtained through subject interviews. Alcohol consumption (wine, beer, and liquor) was quantified in glasses per day, and ≥4 U/d was the limit used to define heavy drinkers. The quality, homogeneity, and completeness of the data were monitored by a coordinating center in Grenoble (L.H., R.d.G.).
Qualitative variables were compared by use of the χ2 test, and quantitative variables were compared through the use of variance analysis. Adjustment of age for qualitative variables was performed with the Mantel-Haenszel χ2 test. All statistical analysis were performed with the SAS statistical package.17
Between January 1997 and April 1998, 17 359 men and 12 267 women were enrolled (Table 1). The mean ages of men and women were nearly identical (38.8±9.5 versus 38.9±9.6 years). Most subjects (96% of men and 95% of women) were employed on a long-term contract. Among men, 14% were upper executives, 27% middle executives, 11% employees, 37% skilled workers, and 11% unskilled workers; among women, these percentages were 8%, 22%, 43%, 13%, and 14%, respectively. The prevalence of obesity (BMI ≥30 kg/m2) was 8.9% among women and 8.5% among men. Of the 3712 men and 1061 women untreated for hypertension with BP ≥ mm Hg on the initial visit, 2914 men (78.5%) and 866 women (81.6%) were examined on a second occasion 1 month later. The mean time interval between visits 1 and 2 was 54±40 days. Fewer than 10% of the invited subjects were reevaluated at a time interval <21 days or >105 days. Subjects who did not attend the second visit were younger than those who did (men, 40.7±9.2 versus 42.7±8.8 years, P=0.01; women, 43.7±9.2 versus 45.4±8.0 years, P=0.01). Subjects who participated in the second visit also had lower systolic and diastolic BPs than subjects who did not (men, 146.8±±8.9 versus 148.8±±8.9 mm Hg, P<0.001 and P<0.001, respectively; women, 144.0±±7.1 versus 146.3±±7.9 mm Hg, P<0.05 and P<0.05). Mean BMI did not differ between participants and nonparticipants at the second visit.
Comparison of Hypertension Prevalence Rates Based on 1 Visit With Those Based on 2 Visits
When the hypertension diagnosis was based on 2 separate visits, its prevalence was lowered by 41% in men and 36% in women. Among persons diagnosed as hypertensive, the percentages of men and women aware of having hypertension increased by 49% and 40%, respectively. Similarly, the proportions of hypertensive men and women under current treatment were increased by 14% and 11%. BP control among all hypertensive subjects increased by 69% and 55% among men and women (Table 2).
The prevalence estimates varied by patient age. The younger the patients were, the wider the difference was between estimates based on 1 and 2 visits (Table 2). As a result, only 40% of men <30 years of age who had been labeled hypertensive after 1 visit were still labeled that way after the second visit. This percentage was higher in women (56%). These percentages were 72% in men and 74% in women >50 years of age. The differences between the 2 estimates were smaller when the mm Hg cut point was used (data not shown).
Prevalence, Awareness, Treatment, and BP Control Based on Estimates From 2 Visits
The prevalence of hypertension (≥ mm Hg) was 16.2% in men and 9.4% in women on the basis of estimates from 2 visits. The prevalence was 10.3% and 7.4%, respectively, for a threshold of ≥ mm Hg (Figure). In both genders, prevalence increased with age. It was 36.8% in men >50 years of age (Table 3). As a whole, only 12.5% of treated hypertensive men and 33.2% of treated hypertensive women had their BP levels lowered by the treatment to < mm Hg. A low awareness, particularly in men, and a low percentage of treated hypertensive subjects with BP controlled by treatment in both genders were the 2 main factors accounting for this poor result (Table 3). The prevalence of hypertension, the proportion of hypertensive subjects aware of their hypertension, and among subjects aware of their hypertension, the proportion of subjects under current treatment increased with age in both genders. In contrast, the proportion of treated subjects whose hypertension was controlled decreased with age. As a result, BP control among all hypertensive subjects decreased with age (Table 3). Among subjects who reported having hypertension, 33.6% of men and 46.5% of women had recorded BP levels < mm Hg (Table 4).
With mm Hg used as a cut point, prevalence estimates decreased, and current treatment increased. Only 61% of men and 77% of women had BP levels controlled by antihypertensive treatment.
One hundred sixteen men and 74 women reported having hyperglycemia. Among these subjects, BP control did not differ significantly (P=0.21 and P=0.54, respectively) from the other subjects when the mm Hg threshold was used. If, as suggested by the World Health Organization-International Society of Hypertension (WHO-ISH) guidelines,18 the mm Hg threshold was used to define BP control, only 11.2% of these men and 25.7% of these women had BP controlled by antihypertensive treatment.
Severity of Hypertension
BP was ≥ mm Hg (grades 2 and 3) in 25% of unaware men and 22% of unaware women. This proportion was higher among aware untreated hypertensive men (40%) and women (33.0%) and among treated men and women uncontrolled by the treatment (42% and 32%, respectively; Table 5).
When estimates based on the first visit were used, among 4387 hypertensive men, 99 had severe undetected hypertension (BP ≥ mm Hg). In 84 subjects, hypertension was detected but untreated; in 88, hypertension was treated albeit uncontrolled. As a result, severe hypertension (BP ≥ mm Hg, 1-visit estimate) was observed among 271 (6.2%) of the 4387 hypertensive men. Among 1413 hypertensive women, undetected, detected but not treated, and treated but uncontrolled severe hypertension was observed in 26, 19, and 30 women, respectively. As a result, severe hypertension (BP ≥ mm Hg, 1-visit estimate) was observed among 75 (5.3%) of the 1413 hypertensive women. Finally, uncontrolled hypertension accounted for 32% of men (88 of 271) and 40% of women (30 of 75) having a BP level ≥ mm Hg (1-visit estimate).
In this population of French working women and men 15 to 69 years of age, the prevalence of hypertension was 9% and 16%, respectively. It is worth noting that in the older age groups, the prevalence of hypertension was higher: one third of men age >50 years had a BP level ≥ mm Hg. As far as hypertension control at the population level is concerned, 1 man out of 2 is not aware of his condition. Three quarters of these patients are under treatment, although only one third have hypertension that is controlled by the treatment. These figures are better among women: three quarters were aware, almost 9 out of 10 were under treatment, and 1 out of 2 were controlled by the treatment. In both genders, awareness and control under treatment are thus the main problems to be addressed. In terms of severe hypertension, the lack of adequate BP control among treated hypertensive patients is also the main cause of BP levels ≥ mm Hg among all hypertensive subjects.
The results obtained in this population cannot be extended to the French population without discussion. These subjects belong to a working population. Their health status is likely to be better than in the general population because of the “healthy worker effect.” It is also likely that they use the healthcare system better because of the mandatory annual medical examination, which gives them a regular opportunity to meet with health personnel.
Some biases must be acknowledged. In our study, antihypertensive treatment means drug treatment. The percentage of hypertensive subjects under treatment is therefore underestimated because a non-negligible proportion of hypertensive subjects, particularly in the mild hypertension group, may be under nonpharmacological treatment, including reduction in weight or alcohol or salt intake. Nonpharmacological treatment was not considered in this first phase of the study. Data from the longitudinal part will be available on this topic. This lack of information may be responsible for an underestimation of the prevalence of treated hypertension and of the proportion of therapeutically controlled BP levels, because those subjects are registered as normotensive in our study.
The definition of hypertension was a second problem. As in other studies, a person being treated with an antihypertensive drug is considered hypertensive. This leads to an overestimation of the prevalence because overtreatment of hypertension resulting from measurement error has been reported to be common.19,20 For example, in this study, particularly among young persons, overdiagnosis of hypertension is suggested by the fact that one third of men reporting high BP levels had in fact normal BP levels. This percentage is higher among young persons, reaching two thirds in men <30 years of age.
Despite recommendations from expert committees indicating the need for measuring BP levels on several occasions before the diagnosis of arterial hypertension is accepted, the consequences of establishing the diagnosis on a single visit have, to the best of our knowledge, not been investigated in epidemiological studies. Dramatic changes were observed when hypertension was assessed on the basis of BP measurements obtained on 2 different occasions. Taking BP during 1 visit has been shown to detect poor BP control as accurately as 24-hour ambulatory BP measurement.21 However, estimates of hypertension prevalence can be expected to be lower than those made on the basis of 1 visit. Consequently, as a result of the decrease in the denominator, which is the number of labeled hypertensive subjects, the proportion of all hypertensive subjects who are aware and treated and have controlled BP levels can be expected to be higher than previously estimated. Compared with results based on 1 visit, the estimation of prevalence was much lower (about −40%), the proportion of aware hypertensive subjects was much higher (≈45%), and the proportion of treated subjects with BP controlled by the treatment among all hypertensive subjects was much higher (≈60%).
These differences were even greater among young persons. Despite the differences in estimates, improving the awareness and the control of high BP with treatment are the priorities in improving the control of hypertension in the population.
Our results are concordant with those obtained in the mid-1990s in France in the Monitoring Trends and Determinants in Cardiovascular Disease (MONICA) population studies.13 It is encouraging to observe that results on awareness were about the same in the late 1990s, with a diagnostic threshold of mm Hg, as in the 1980s, with a mm Hg diagnostic threshold.12–14 In contrast with previous French studies, the percentage of hypertensive persons under current treatment was high.
Compared with the US NHANES III study, which is based on 2 visits, the proportion of awareness observed in our study was lower in the same age range groups.4 Similarly, the results observed in France differ in the lower proportion of hypertensive subjects under current treatment who have BP levels controlled by treatment. In contrast, the percentage of treated subjects among aware hypertensives is much higher in France. Differences in health insurance systems between the 2 countries might explain these differences. As a result of the larger proportion of the hypertensive subjects under treatment in France, a higher proportion of subjects with low compliance may be treated, leading to poorer results. Among men and women aware of being hypertensive, 15% and 9%, respectively, had BP levels > mm Hg and were untreated. It is likely that they did not follow up with a doctor as recommended by the work-site physician, who, in France, is not authorized to prescribe medications. It would be quite insufficient, however, to think of poor results in awareness and BP control as a result of patient behavior. Some subjects might have forgotten BP measurements or screening. However, one might hypothesize that the diagnosis of hypertension might not have been given to hypertensive patients with moderately elevated BP levels. This hypothesis is in keeping with the lack of decision in patients with inadequate BP control reported in a community health plan. Surprisingly, hypertensive patients with elevated BP levels were more carefully and more frequently monitored. However their treatment was not changed.22 In the United Kingdom, a high rate of high BP without any subsequent action being taken has been reported.2 A reluctance to accept mm Hg as the threshold for establishing the diagnosis of hypertension might be involved. Accordingly, the threshold for drug intervention increased from to mm Hg as the age range increased from <45 to >64 years.2
As far as BP control through treatment is concerned, our results show considerable room for improvement. As a result of the strategy, 13% of the hypertensive men in our population had BP levels controlled by the treatment compared with 14% in the group of 18 to 49 year olds in NHANES III. Among women, 33.3% in our population could be compared with 38% in the US population.4 This percentage of subjects controlled by treatment is also very close to that observed in 2 populations representative of the French general practitioners in 1995 and 1999, 24% and 32%, respectively.5,23
In conclusion, it should be kept in mind that estimates of hypertension prevalence and therapeutic control based on 1 visit are overestimated or underestimated, respectively, compared with estimates based on 2 visits. Despite these differences, insufficient awareness of high BP and insufficient BP control through treatment are the 2 main problems to be addressed to improve BP control at the population level. On the other hand, the high proportion of hypertensive subjects >50 years of age in this healthy population calls for primary prevention of arterial hypertension through weight, alcohol, and salt reduction at a population level as a public health goal.
Epidemiology group of the Société Française d’Hypertension Artérielle: J.M. Mallion, P. Ducimetière, P.F. Plouin, and J.P. Cambou. IHPAF Working physicians: Drs B. Agopian, C. Amoureux, L. Arsac, C. Barthelemy, N. Begnard, M.R. Belville, M. Bidault, C. Boijoux, F. Chamard, A. Chamot, B. Chanut, S. Charrière, H. Chebli, N. Cholley, C. Combre, M.C. Combre, C. Costa, C. De Dreuzy, J. Decatoire, B. Defawe, M. Dischant, D. Duquesne, I. Elter, E. Filliard, I. Fortier, C. Gamot, J. Grasset, A. Guiret, H. Hage, M.L. Hinard, A. Hoorelbeke, A. Lafragette, L. Langlois, N. Laplaiche, D. Lariviere, F. Marlin, C. Michel-Panichi, Z.H. Nghiem, C. Paquette, B. Pasquet, S. Picart, S. Piotte, S. Portal, A. Poudevigne, H. Ramambason, D. Rodallec, A.M. Roux, J. Royon, E. Sailly, J.J. Servagent, J. Solanet, M. Stern, R. Wolff, and M.H. Zerbib.
The following pharmaceutical companies have supported the activity of the group and this study: AstraZeneca, Aventis, Bristol-Myers-Squibb, Debat Cardio, Glaxo Welcome, Lipha Sante, Merck Sharp & Dohme Chibret, Information Servier, Wyeth-Lederle. This study was undertaken under the auspices of Société Française d’Hypertension Artérielle and its Epidemiology Group. We would like to thank the staff of the work sites, without whom this work would not have been possible, and we thank Michael J. Fine for reviewing and improving the manuscript.
- Received January 30, 2001.
- Revision received March 5, 2001.
Burt VL, Cutler JA, Higgins M, et al. Trends in the prevalence, awareness, treatment, and control of hypertension in the adult US population: data from the health examination surveys, 1960 to 1991. Hypertension. . 1995; 26: 60–69. Published erratum appears in Hypertension. 1996;27:1192.
Burt VL, Whelton P, Roccella EJ, et al. Prevalence of hypertension in the US adult population: results from the Third National Health and Nutrition Examination Survey, 1988–1991. Hypertension. . 1995; 25: 305–313.
Smith WC, Lee AJ, Crombie IK, Tunstall-Pedoe H. Control of blood pressure in Scotland: the rule of halves. BMJ. . 1990; 300: 981–983.
Poggi L, Chamontin B, Lang T, et al. An attempt to improve blood pressure control in the French population: information and outcome awareness at the general practitioner level. Am J Hypertens. . 1998; 11: 905–908.
Fouriaud C, Jacquinet-Salord MC, Degoulet P, et al. Influence of socioprofessional conditions on blood pressure levels and hypertension control: epidemiologic study of 6,665 subjects in the Paris district. Am J Epidemiol. . 1984; 120: 72–86.
World Health Organization. Obesity: Preventing and Managing the Global Epidemic. Geneva: Switzerland: World Health Organization; June 3–5, 1997.
SAS/STAT User’s Guide, Version 6. 4th ed. Cary, NC: SAS institute Inc: 1989.
Kaplan NM. The appropriate goals of antihypertensive therapy: neither too much nor too little. Ann Intern Med. . 1992; 116: 686–690.
Myers MG, Reeves RA, Oh PI, Joyner CD. Overtreatment of hypertension in the community? Am J Hypertens. . 1996; 9: 419–425.
Chamontin B, Lang T, Vaisse B, et al. Evolution du contrôle de l’HTA en France de 1994 à 1999 en médecine générale. Arch Mal Coeur Vaiss. . 1999; 92(Hors Série): 12.