Introduction

Hypertension is a major independent risk factor for cardiovascular and renal disease,1 is highly prevalent in adults worldwide2 and is a condition that can be both prevented and treated. Therefore, the distribution of blood pressure (BP) in the population, as well as the awareness, treatment and control of high BP is a major public health concern. International comparisons in the past decade have suggested a particularly high level of BP and low awareness, treatment and control in Germany 1998 compared with other developed countries.3,4 Since 1998, strongly declining stroke mortality rates in Germany5 have suggested BP improvement but nationwide BP data based on standardized measurements only became available with the German Health Interview and Examination Survey for Adults (DEGS1) 2008–11.6 The aim of this study is a reappraisal of BP in Germany from 1998 to 2011 based on two national health examination surveys conducted in 1998 and 2008–11 as well as on a methodological study providing a calibration formula because of a change of the BP measurement device.7

Materials and Methods

Study population

Two national health examination surveys have been conducted in reunified Germany: the German National Health Interview and Examination Survey 1998 (GNHIES98) and the German Health Interview and Examination Survey for Adults 2008–11 (DEGS1). Both consist of a nationwide two-stage clustered sample from local population registers6 (7124 GNHIES98 participants aged 18–79 years from 120 communities and 7988 DEGS1 participants from 180 communities). DEGS1 has largely identical survey methods and measurement protocols as GNHIES98 in order to facilitate the estimation of time trends. All GNHIES98 participants were invited to reparticipate in DEGS1 (response rate 64%, n=3795).8 An additional cross-sectional sample of 4193 adults aged 18–79 years were recruited for DEGS1 based on the same sampling design (response 42%).

Measurements

The GNHIES89 BP measurement protocol and observer training was based on the highly standardized WHO MONICA Project protocol.8 The protocol was continued in DEGS1, but the BP measurement device was changed from a standard mercury sphygmomanometer (Erkameter 3000, Erka, Bad Tölz, Germany) to an automated oscillometric device (Datascope Accutorr Plus, Mahwah, NJ, USA), including new sets of manufacturer-provided cuffs and cuff selection rules. The selection of a new instrument was based on fulfillment of the requirements of agreement with the gold standard mercury sphygmomanometer set out in the Association for the Advancement of Medical Instrumentation (AAMI) and British Hypertension Society (BHS) protocols.9 Three BP measurements were taken at the right arm at 3-min intervals after a non-strenuous part of the examination and an additional 5-min rest (GNHIES98 3 min). As only the second and third measurements were used in the analysis, resting time was longer than 5 min in both surveys. During the procedure, individuals were sitting on a height-adjustable chair, their back supported, with a slightly bent elbow lying on a table at the level of the right atrium. Both feet were straight on the floor and legs were not crossed. The correct cuff size was identified by measuring the upper arm circumference (AC) between the acromion and the olecranon. Three different cuff sizes were used in DEGS1 (bladder 10.5 × 23.9 cm for AC 21.0–27.9 cm, bladder 13.5 × 30.7 cm for AC 28.0–35.9 cm and bladder 17.0 × 38.6 cm for AC 36.0–46.0 cm). In GNHIES89, the bladder sizes were 8 × 20 cm for AC<20 cm, 12 × 28 cm for AC 20–40 cm and 14 × 40 cm for AC>40 cm. The correct position of the cuff above the brachial artery was additionally checked with a mark on the cuff.

The International Standard Classification of Education (ISCED) 1997 was used to categorize the school and professional education level into low: ISCED 1 and 2; medium: ISCED 3B, 3A, 4A and high: ISCED 5A, 5B and 6.10 Body height was measured with a portable stadiometer (Holtain Ltd., Crymych, UK), and body weight was measured with a calibrated electronic scale (SECA column scale 930, Seca Inc., Birmingham, UK). Body Mass Index was weight in kilograms divided by squared height in meters. A medical interview covered i.a. previous medical diagnoses, for example, diabetes, coronary heart disease and stroke. Dyslipidemia was defined as total cholesterol level ⩾190 mgdl−1 or previous medical diagnosis of dyslipidemia.

Analysis

For the final data analysis, the second and third measurements were averaged. GNHIES98 BP data were calibrated for comparison with DEGS1 BP data based on a methodological study, which is described in detail elsewhere.7 In brief, comparisons of 315 measurement pairs according to the principles of the International Protocol revision 2010 for the validation of BP measuring devices in adults of the European Society of Hypertension11 revealed higher mean systolic BP (SBP) and diastolic BP (DBP) if measurements were conducted according to the standard-mercury-old-protocol compared with the automated-Datascope-new-protocol measurements. Differences increased with BP level, pulse pressure, cuff width to AC ratio difference, age and sex. Calibration formulas for SBP and DBP were used as described previously.7 Both GNHIES98 and DEGS1 data were weighted according to sex, age, federal state, German/non-German nationality, community size and education in order to reflect the German population as of 31 December 2010. DEGS1 weights of former GNHIES98 participants additionally included adjustment for re-participation probability.6 In order to take into account both the weighting and the correlation of the participants within a community, the confidence intervals were determined using using SPSS 20 procedures for complex samples. The weighted distribution of BP was depicted graphically with kernel density estimates (calculated with SAS 9.4, SAS Institute, Cary, NC, USA).

Definitions were: (1) Hypertension: SBP⩾140 mm Hg or DBP⩾90 mm Hg or treatment as defined below; (2) Awareness: Hypertension in individuals who had ever been told by a physician to have elevated BP or hypertension; (3) Treatment: intake of ATC-coded medication with antihypertensive main effect (diuretics (ATC code C03), beta-blockers (C07), calcium channel blockers (C08), ACE inhibitors (C09) and antihypertensive drugs (C02)) in the past 7 days. However, as the indication for taking these drugs may be other than hypertension, the antihypertensive medication was only used for defining prevalent and treated hypertension if the participants were aware of having hypertension. (4) Control: SBP<140 mm Hg and DBP<90 mm Hg in treated hypertensives.

Results

This study is based on data from 7108 GNHIES98 and 7095 DEGS1 participants aged 18–79 years with complete BP data (Table 1).

Table 1 Characteristics of the study population
Full size table

The overall prevalence of hypertension, including controlled hypertension, remained almost unchanged between 1998 and 2008–11 (Table 2). Nearly one in three adults aged 18 to 79 years had hypertension in Germany in 2008–11 (men 33.4%, women 29.9%, total 31.6%). There was a small increase in hypertension prevalence in men from 1998 to 2008–11 (from 30% to 33%), while the prevalence in women remained almost unchanged at 30%. Uncontrolled hypertension (BP⩾140/90 mm Hg, treated or untreated), however, decreased substantially from 23% to 15% (women from 22% to 13%, men from 24% to 18%; Table 2). Optimal BP (<120/80 mm Hg, regardless of treatment) increased from 33% to 41% (women from 40% to 53%, men from 25% to 29%). The entire distribution of SBP and DBP shifted towards lower values in both men and women (Figure 1). Mean SBP decreased from 129 to 124 mm Hg, that is, by approximately 4 mm Hg in 18–79-year-old adults in Germany from 1998 to 2008–11 and mean DBP by 5 mm Hg (a decrease of 7 mm Hg in both SBP and DBP in women, men 3 mm Hg in SBP and DBP; Table 3). The decrease of mean SBP and DBP was most pronounced in the subgroup of treated hypertensives (decrease of SBP from 144 to 130 mm Hg, DBP decrease from 85 to 75 mm Hg, Table 3). However, mean SBP and DBP was also lower in the subgroup with no hypertension (121 mm Hg vs 119 mm Hg for SBP and 74 mm Hg vs 71 mm Hg for DBP in men and women; however, in men this difference was only significant for DBP).

Table 2 Prevalence (95% CI) of prehypertensive and hypertensive blood pressure and of hypertension in adults aged 18–79 years living in Germany 1998 and 2008–11 (weighted to the population 2010)
Full size table
Figure 1
figure 1

Density curves representing the distributions of SBP and DBP in adults aged 18–79 years in Germany 1998 and 2008–11, weighted to the population as of 31 December 2010.

Full size image
Table 3 Mean systolic blood pressure (SBP) and diastolic blood pressure (DBP) in adults aged 18–79 years living in Germany (mean in mm Hg, 95% CI) (weighted to the population 2010)
Full size table

Awareness among hypertensives increased from 69% to 82% (women from 74% to 87%, men from 65% to 78%), treatment among hypertensives increased from 55% to 72% (women from 62% to 79%, men from 48% to 65%), treatment among aware hypertensives increased from 79% to 88% (women from 84% to 92%, men from 73% to 84%), control increased from 23% to 51% (women from 25% to 58%, men from 20% to 45%) and control among treated hypertensives increased from 42% to 72% (women from 41% to 73%, men from 43% to 70%) (Table 4). The proportion of the population with hypertension medication (subjects aware of hypertension and taking medication with antihypertensive main effect) increased substantially from 16.3% (95% CI 15.0–17.7) in 1998 to 22.7% (21.5–23.9) in 2008–11. The prevalence of users of medication with antihypertensive main effect irrespective of indication increased from 19.7 (18.2–21.3) to 25.7 (24.4–27.0).

Table 4 Awareness, treatment and control among participants with hypertension (weighted to the population 2010)
Full size table

All of the described improvements were more pronounced in women than in men and in the elderly. Additionally, young men between 18 and 29 years had an opposite trend with an average 1.5 mm Hg higher SBP (however, 2 mm Hg lower DBP), higher prevalence of hypertension (8.5% vs 4.1%, P=0.02), which was mostly uncontrolled (Table 5). Unlike other groups, young men did not show an improvement of hypertension awareness, treatment and control between 1998 and 2008–11.

Table 5 Factors associated with hypertension prevalence, awareness, treatment and control in adults aged 18–79 years living in Germany
Full size table

In multivariate analysis, hypertension was more likely in men, with increasing age as well as in the presence of overweight and obesity, high cholesterol, diabetes, coronary heart disease or stroke (Table 5). Current smoking was inversely associated with hypertension. Awareness, treatment and control among hypertensives were more likely in women, with increasing age, and in the presence of obesity, diabetes and coronary heart disease or stroke. Educational level was not significantly associated with hypertension prevalence or management with the exception of more likely treatment in the high-education group in GNHIES98. Similarly, the socioeconomic status, including education, occupational status and income,12 had no influence on hypertension prevalence and management; only individuals with a middle socioeconomic status were more likely to be aware of hypertension in DEGS1 compared with the low socioeconomic status group (results not shown in the table).

Discussion

These data demonstrate that between 1998 and 2008–11 BP level in Germany has decreased and awareness, treatment and control have increased substantially. However, similar to other countries with decreasing BP, hypertension prevalence, which includes also controlled hypertension, remained almost constant.13, 14, 15

In the 1990s, Germany had the highest SBP and DBP among six European countries, Canada and the United States as well as the lowest or among the lowest hypertension awareness, treatment and control proportions both at the 140/90 mm Hg and the 160/95 mm Hg thresholds.3,4 This was a plausible finding as Germany had also the second-highest mortality of stroke, which is the clinical outcome most closely correlated to hypertension. Stroke mortality in Germany has substantially declined in the past decades, from 180 deaths per 100 000 in 1980 to 128 deaths in 1990, 83 death in 2000 and 51 death in 2010 (all numbers are age-standardized to the population in 1987; www.gbe-bund.de), suggesting that BP may have improved as well. After 1998, the DEGS1 survey is the first study to provide national population-based BP data based on standardized BP measurements in adults in Germany that allow verification of the hypothesized downward BP trend in Germany.

The population-wide decrease of almost 7 mm Hg mean SBP and DBP in women and 3 mm Hg SBP and DBP in men (overall 5 mm Hg) can be considered as large given that a population-wide reduction of SBP of 2 mm Hg has been estimated to involve a reduction of stroke mortality of 10% and of ischemic heart disease mortality of 7%.1 Following this estimation, the observed BP decrease of 5 mm Hg in Germany would be associated with a decrease of stroke mortality of 25%. In fact, stroke mortality decreased by 39% in 2010 compared with 2000, but improved treatment and trends in other risk factors for stroke are likely to contribute to changes in stroke mortality.

A major contributor of the observed BP decrease appears to be the increased intake of antihypertension medication. However, even in participants with BP <140/90 mm Hg and without antihypertensive medication SBP decreased in women by almost 4 mm Hg and DBP decreased in both men and women. Similarly, in people without hypertension SBP dropped in Canada by 4 mm Hg on average between 1992 and 2009,15 whereas in the same period DBP increased by 1.5 mm Hg in this group.13 A BP decrease in people without hypertension and without treatment may not only be due to improved lifestyle factors but may also reflect a selection effect due to more intensive diagnosis and treatment close to the threshold of 140/90 mm Hg, leaving a healthier group untreated in 2008–11 than in 1998.

Lifestyle trends that could have contributed to BP decrease in Germany are increased physical activity,16 improved nutrition with higher levels of fruit and vegetable intake17 and decreased alcohol consumption per capita in the past decade (http://www.dhs.de/datenfakten/alkohol.html). No data are available so far on recent salt consumption trends in Germany, but from modelling of worldwide data, salt consumption in Germany has been estimated to have only marginally changed in the past decade.18 On the other hand, obesity prevalence has increased in the same period in men in most age groups and in young women,19 possibly attenuating the BP decrease.

The downward SBP and DBP trends in the past decade in Germany are more pronounced than the Global Burden of Disease Study 1980–2008 estimates per decade for Western Europe, which were −2.1 mm Hg for men and −3.5 mm Hg for women.2 Large 10-year SBP downward have been found in a number of the WHO MONICA populations in the 1980s and 1990s (several with reductions of >10 mm Hg),20 in North Karelia in Finland −5 mm Hg from 1992–2002 (of note, long after the initial North Karelia CVD prevention project 1972–77)21 and −10 mm Hg in men and −6 mm Hg in women in Canada from 1986–92 to 2006 (the Canadian Hypertension Education Program being started in 1999).15 Similar to Germany, the prevalence of hypertension remained stable during that period in Canada.15

All hypertension management indicators have improved considerably in Germany. Awareness was 82% and compares favorably to other countries, for example, in a three-country comparison ( USA 81% in 2007–10, Canada 83% in 2007–9 and England 65% in 2006).22 Using this same three-country comparison, German treatment and control proportions reported in this paper are quite similar to the United States, not quite as good as in Canada but better than in England (treatment in Germany 72%, USA 74%, Canada 80%, England 51%; control in Germany 51%, USA 53%, Canada 66%, England 27%). However, these figures show that there is still a large prevention potential in Germany. Similar to the United States, Canada and England, hypertension awareness, treatment and control in Germany is not as good in men as in women.22,23 In addition, young men aged 18–29 years had higher SBP and a higher prevalence of hypertension in 2008–11 than in 1998, that is, the BP trend in this group was opposite. Young men had a slightly more pronounced obesity increase than other age groups; however, this can account only for a rather small part of this trend. Among subjects with hypertension, those who also had coronary heart disease or stroke were more likely to be aware of hypertension, as well as to be treated and controlled. This finding, which can reflect increased targeting as well as increased compliance of high-risk patients, has been also reported from Canada.15

This analysis has several limitations. The data of the two national surveys are very well comparable in many respects, but due to the device change the 1998 standard mercury values had to be calibrated to Datascope values7 and residual differences due to device and cuff differences cannot be excluded. The BP decrease without the calibration would appear much larger. Nevertheless, the changes from 1998 to 2008–11 are so pronounced that the conclusion that BP and BP management in Germany have improved is likely to be robust. Another possible limitation could be selection bias through non-response in the two surveys. Epidemiological surveys worldwide suffer from declining response rates,24 and the GNHIES98 with a response rate of 61% and DEGS1 with a response rate of 62% in reinvited participants and 42% in newly invited participants build no exception. Selective loss to follow-up between GNHIES98 and DEGS1 cannot be excluded and may not be fully compensated by the weighting procedure. However, hypertension status in GNHIES98 was not associated with DEGS1 participation. Last but not least, our results are based on BP measurements taken on a single day, which is not sufficient for a clinical diagnosis of hypertension. However, this type of data are customary in BP epidemiological studies worldwide and allows the evaluation of BP trends.

Despite these limitations, the study supports in a rather robust way the encouraging news that BP has decreased and management has improved in a relevant manner in Germany in the past decade. Declining stroke mortality rates in Germany also support our results. However, there is still a great preventive potential achievable through lifestyle changes and further improved treatment.