Epidemiology
Elevated blood pressure, cholesterol, and bodyweight are all established risk factors for CVD and, in the case of bodyweight, for other diseases, such as diabetes, certain cancers, and osteoarthritis. The associations between blood pressure (Asia Pacific Cohort Studies Collaboration 1999, 2003a; Prospective Studies Collaboration 2002); cholesterol (Asia Pacific Cohort Studies Collaboration 2003b; Law, Wald, and Thompson 1994; Prospective Studies Collaboration 1995); and body mass index (BMI) (Asia Pacific Cohort Studies Collaboration 2004; Willett and others 1995) and CVD are direct and continuous from relatively low levels, indicating that optimal levels are about 115/75 millimeters of mercury (mmHg), 3.8 millimoles per liter (mmol/l), and 21 kilograms per square meter (kg/m2), respectively (figure 45.1).
[Figure
45.1]
Although some studies suggest J- or U-shaped associations (Calle and others 1999; Cruickshank 1994; D'Agostino and others 1991; Farnett and others 1991; Field and others 2001; Iso and others 1989; Kannel, D'Agostino, and Silbershatz 1997; Stewart 1979; Troiano and others 1996), low levels of these risk factors are unlikely to cause CVD. Rather, such associations more likely reflect incipient disease, which itself produces both a fall in risk-factor levels and an increase in CVD risk (Alderman 1996; Flack and others 1995; MacMahon and others 1997; Manson, Willett, and Stampfer 1995; Neaton and Wentworth 1992; Sleight 1997a; Sleight 1997b; Stevens and others 1998). No trial evidence points to a J-curve association for blood pressure, despite including patients with below average blood pressure (Hansson and others 1999; McMurray and McInnes 1992; Pfeffer 1993; Staessen and others 1997).
The continuous associations between blood pressure, cholesterol, and bodyweight and CVD demonstrate the lack of a biological justification for current threshold levels, such as those that define hypertension. Indeed, most of the disease burden resulting from these three risk factors occurs in the large majority of the population with nonoptimal levels but without hypertension, hypercholesterolemia, or obesity. Hence, this chapter avoids those terms and instead uses high blood pressure, high cholesterol, and high bodyweight, defined as nonoptimal levels of these risk factors (that is, over 115/75 mmHg, 3.8 mmol/l, or 21 kg/m2, respectively).
The strength of the proportional associations of these risk factors with CVD is similar for most population subgroups. Although they attenuate with age, they remain strong and positive in the oldest age groups. Overall, in middle-aged populations, a 10 mmHg lower systolic blood pressure (SBP) is associated with a roughly 30 to 40 percent lower stroke risk and 20 to 25 percent lower ischemic heart disease (IHD) risk, a 1 mmol/l lower cholesterol level is associated with about a 15 to 20 percent lower stroke risk and 20 to 25 percent lower IHD risk, and a 2 kg/m2 lower BMI is associated with an 8 to 12 percent lower stroke and IHD risk and an approximately 20 to 30 percent lower diabetes risk.
