To the Editor:

Hypertension is a major cause of morbidity and mortality, and its prevention has a major public health advantage in terms of primary prevention. Hypertension is regarded as a main risk factor for cardiovascular and cerebrovascular diseases, and lowering of blood pressure (BP) has repeatedly been recommended in combination with that of other risk factors, such as components of the metabolic syndrome and lifestyle-related factors.1

Parikh and colleagues2 presented a risk score for predicting the near-term incidence of hypertension, which was validated by Kivimaki and colleagues.3 The risk score, calculated based on age, sex, systolic BP, diastolic BP, body mass index, smoking status, and parental history of hypertension, can be used to predict the development of hypertension in the short-term (ie, within 4 years). They reported that none of the biochemical variables had any advantage in primary care, although fasting blood glucose was needed to exclude individuals with diabetes mellitus from the target population.

Bloch and associates4 previously presented a question on the need for a simple office-based tool to predict incidence of hypertension and adequacy of clinical judgment. It has also been debated that causes of diseases in individual cases and determinants of disease of the incidence in the population are different.5,6 The population-based and high-risk identification approaches represent complementary approaches for the development of health policies; however, the primary purpose of each of these approaches is different. The development of a risk score for predicting hypertension is useful for high-risk identification but it is not directly used for BP-lowering programs, such as regular exercise, maintenance of an optimal weight, and prohibition of calories with low salt intake. Thus, there is the opinion that although the hypertension risk score is useful for screening to identify individuals with subclinical damage, the population approach to alleviate the development of the disease should be simultaneously employed. If not, prevention and treatment strategies restricted to individuals with hypertension will overlook other BP-related diseases.1 While the incidence of hypertension is related partly to genetic factors, the main regulation is achieved by implementing changes in lifestyle, including environmental factors, and the authors speculate that health policy planners feel that risk scores are useful for projecting the future burden of hypertension and for allocating resources in communities.

Although there is limitation related to ethnicity in this study (the score was developed for white persons living in developed countries), the authors applied the risk score to male car-manufacturing workers in Japan on the assumption that the risk score for predicting hypertension may also be useful for prevalence estimation.

A total of 3486 male workers (age range, 35–59 years) at a company in Gunma Prefecture, Japan, were the primary participants of this study. All the patients had responded to a questionnaire containing questions on current medical and treatment history and the status of some lifestyle factors. Smoking habit was categorized as current smoking/no smoking. Parental history of hypertension was categorized as yes/no, which was different from the case in the original scoring system. Systolic and diastolic BPs were determined by an automated BP measurement device (TM-2540C; A & D Company, Ltd, Tokyo) with the patient in the sitting position after 3 minutes of rest. Venous blood samples were drawn from each patient after they had fasted for 9 hours, and samples were stored at −80°C. The plasma glucose level was determined using the hexokinase method. Body mass index was calculated as weight (kg)/height (m2). As baseline data, patients with a history of treatment for diabetes and/or hypertension, those with a BP of ≥140/90 mmHg, and those with a fasting plasma glucose of ≥126 mg/dL (7.0 mmol/L) were all excluded. Finally, 2444 male participants were included in the analysis. Informed consent was obtained from all the study participants, and the study protocol was approved by the ethics committee of the company at which the workers were employed. “Hypertension after 1 year” was judged if systolic BP was ≥140 mm Hg and/or diastolic BP was ≥90 mm Hg. Patients taking antihypertensive medication were also regarded as having hypertension.

As shown in the Figure, the prevalence of hypertension stratified by the risk score for predicting hypertension after 1 year fluctuated because of the limitation of sample size, but the value was higher than the incidence data after 2 years, presented by Parikh and colleagues.2 As compared with cardiovascular mortality, cerebrovascular mortality is higher in Japan than that in other developed countries, according to the World Health Organization’s mortality database. The authors conclude that a validation study is needed in different ethnic groups to use the data on the prevalence of hypertension stratified by the risk scores, which reflect the near-future burden of hypertension in US patients.


Figure Figure.  Near-term incidence and prevalence of hypertension stratified by the risk score for predicting hypertension.

Download figure to PowerPoint


  1. Top of page
  2. References