The PAMELA study - results and perspectives

The Pressioni Arteriose Monitorate E Loro Associazioni, known worldwide as PAMELA Study, was planned and initiated almost twenty years ago. This research project was designed to collect information from a sample of the general population geographically located in Monza (a town located in the north-east of Milan, Italy) and surroundings.

The PAMELA study allowed important new insights to be achieved in the following areas:

• normality values of home and 24-hour blood pressure
• relationship between blood pressure values and target organ damage
• seasonal variations of blood pressure
• prognostic importance of blood pressure variability
• epidemiology and clinical importance of “white coat” and “masked” hypertension
• epidemiology and prognostic impact of metabolic syndrome

I - Initial steps

The study started in 1990 with, as first goal to define the values of normality of blood pressure as self-measured at home - home blood pressure - and ambulatory blood pressure - 24-hour blood pressure - as measured outside the clinic environment. Until then, these normality values had mostly only been based on measurements collected in subjects referred for hypertension to specialists’ clinics, thus introducing an important selection bias.

PAMELA solved this issue for the first time (1), by obtaining several types of blood pressure measurements, i.e. clinic blood pressure with a mercury sphygmomanometer, home blood pressure with a semiautomatic oscillometric validated device, and 24-hour blood pressure with an automatic and validated device. All were taken from a large sample randomly selected from the population of Monza, stratified only for gender and age ranging from 25 to 74 years.
In addition, study protocol included collection of medical history, echocardiograhic determination of left ventricular mass as well as metabolic data. Out of 3200 subjects selected, 64% agreed to take part in the study. This portion of subjects underwent the same measurements 10 years later -participation rate was approximately 70%-, and were followed from randomisation for 148 months altogether a time during which all causes of deaths and cardiovascular events were recorded.

II - Normality values of home and ambulatory blood pressure

Results of the survey provided evidence that:

1. clinic, home and 24-hour blood pressure in the general population show a normal like distribution and are significantly related to one another
2. clinic blood pressure is markedly higher than home and 24-hour blood pressure – mostly regardless of gender and this difference in values widens with advancing age and higher clinic blood pressure values
3. the upper limit of normality of home and 24-hour blood pressure, calculated as corresponding to a clinic blood pressure of 140 mmHg (systolic) and 90 mmHg (diastolic), are 132/83 mmHg (home) and 125/79 mmHg (24-hour) respectively (2).

It was found that the determination of normality values for home and 24-hour blood pressure allowed for a more comprehensive assessment of blood pressure control in treated hypertensives. Indeed, treated hypertensives’ blood pressure values are much higher than normotensives’, and are similar to that of hypertensives not taking antihypertensive therapy: this holds true for clinic and for home and 24-hour blood pressure measurements. The percentage of treated hypertensives with clinic blood pressure lower than 140/90 mmHg was approximately 20%, however blood pressure control improved further when going from clinic to home and 24-hour values (3).

III - Prognostic significance of home and 24-hour blood pressure

PAMELA explored the clinical importance of home and 24-hour blood pressure in the general population, making use of a long follow-up -around 12 years-, and supplying results that were quite different to those from other population-based studies (1-4). In PAMELA, Clinic, home and 24-hour blood pressure were all found predictive of cardiovascular events, with no superiority of “out of office” blood pressures (home and 24-hour) over the clinic blood pressures.

IV - White-coat and masked hypertension

The new availability of these various types of blood pressure assessments allowed recognising that in an individual subject, the normality or abnormality of various measurements may be heterogeneously associated: if we consider clinic and 24-hour blood pressure, one can identify various groups of subjects:

1. One in which both values are normal (true normotensives)
2. subjects with both values elevated (sustained hypertensives), but also
3. subjects with high clinic blood pressure and normal 24-hour blood pressure (isolated ambulatory hypertension, also defined “white-coat hypertension”), as well as
4. subjects with normal clinic blood pressure and elevated 24-hour (isolated ambulatory hypertension, also called “masked hypertension”).

In the PAMELA study, the last two conditions showed a non-negligible prevalence, being - approximatively 12% and 9% respectively, while studies addressing the problem of prognosis in white-coat and masked hypertension have reported different results. PAMELA indicated that the prevalence of left ventricular hypertrophy is clearly lower in subjects characterised by a true normotension than in subjects with white coat or masked hypertension, and cardiac organ damage, on the other hand  was less frequent  in white coat or masked hypertension than in true hypertensives (4).
Results also show that white-coat and masked hypertension increase the 10-year risk of developing an impairment of glucose tolerance (impaired fasting glucose and diabetes mellitus) and sustained hypertension (5,6).

PAMELA emphasised the evidence that white-coat and masked hypertension are not clinically innocent conditions, they are pathological states associated with increased cardiovascular risk, at a level between that of normotensives and hypertensives.

V - Blood pressure and cardiac organ damage

PAMELA’s methodology included evaluation of left ventricular mass and geometry by echocardiography. Thus, new, major findings relating to cardiac organ damage resulted:

1. First, in the treated fraction of the hypertensive population with satisfactory blood pressure control, left ventricular mass index values remain greater than those of normotensives, suggesting that even when antihypertensive drug treatment is effective in achieving adequate target blood pressure, it fails to fully restore normal cardiac structure (7).
2. Second, left ventricular mass index is a predictor of death and cardiovascular events independent not only, of clinic blood pressure (as already known) but also, of 24-hour blood pressure - a value obtained outside the clinic environment. During follow-up, rate  of fatal and non-fatal hospitalisations, cardiovascular events as well as all-cause death was four to five-fold in patients with left ventricular hypertrophy as compared to those without and the increased risk remained significant even when data were adjusted for a large number of confounders, including home and 24-hour blood pressure

VI - Blood pressure variability and its prognostic value

A first aspect of this substudy was the description of the long-term aspect of blood pressure variability, i.e. seasonal variation of systo-diastolic values (8). PAMELA showed that blood pressure is highest in winter and lowest in summer, with intermediate values during other two seasons, not only for clinic but also for home and 24-hour measurements, a finding further supported by the fact that these modifications were evident during both day and night-time, with conditions during each period engendering widely different factors controlling the cardiovascular system. We found that a component of systolic and diastolic blood pressure variability, the so called non-erratic pressure, residual pressure, was directly and independently associated with left ventricular mass index (9). In addition, cardiovascular and all-cause mortality were independently predicted by residual diastolic 24-hour variability and by attenuation of the major cyclic component of circadian blood pressure variability (10).

VII - Blood pressure, metabolic profile and metabolic syndrome

Several studies had shown that hypertensives display alterations in glucose and lipid metabolism more frequently than age-matched normotensives do. Data collected, however, were mainly based on clinic blood pressure and no information was available until a few years ago as regarding the relationship between metabolic variables and out-of-office blood pressure. PAMELA addressed this issue by assessing serum total cholesterol, high density lipoprotein cholesterol (HDL-C), fasting blood glucose, and calculating the incidence proportion of hypercholesterolemia, impaired fasting glucose and diabetes mellitus in its study population (11). The cohort was subdivided in four categories according to the presence of optimal (<120/80 mmHg), normal (120-129/80-84 mmHg), high normal (130-139/85-89 mmHg) and elevated blood pressure (=140/90 mmHg)(classification of European Society of Hypertension/European Society of Cardiology) (12). Similarly, subjects were subdivided in categories of home blood pressure and quartiles of 24-hour blood pressure.

Results showed that, with increasing clinic blood pressure,

1. blood glucose and serum cholesterol levels progressively increase while HDL-C progressively decreases in both categories of home and 24-hour blood pressure quartiles.
2. percentage of subjects with impaired fasting glucose or diabetes mellitus also progressively increases from the lowest to the highest category of clinic blood pressure and quartiles of home or 24-hour blood pressure (11).

PAMELA also contributed to expanding the information available on the metabolic syndrome (11). The metabolic syndrome:

1. was found in a considerable fraction of the population (16.2%), with elevation in office blood pressure being the most frequent component and blood glucose abnormality the least frequent component
2. induces frequent elevation in home and/or 24-hour blood pressure, as well as a greater left ventricular mass index and greater prevalence of left ventricular hypertrophy, and
3. markedly increases the risk of cardiovascular and all-cause death (71% and 37% respectively).

Conclusion:

The results of the PAMELA study as reviewed in this paper confirm the importance of epidemiological studies which are capable of providing dynamic information on a given disease, its complications and interactions with other pathological states or risk factors. In the next years new data will hopefully be collected in the context of a new PAMELA study, whose detailed planning is currently ongoing.