|
Article Excerpt
PERSONS REPORTING AN INTAKE of three or more standard-sized alcoholic drinks per day have an increased prevalence of systemic hypertension (HTN) in cross-sectional analyses and an increased incidence in prospective investigations. These findings are detailed in various reviews (Keil et al., 1993; Klatsky, 2000; Klatsky and Gunderson, 2006; MacMahon, 1987). Although a biological mechanism is not established, the reviews also describe clinical experiments that confirm an alcohol-HTN relationship and offer support for a causal hypothesis. Conflicting reports about associations of lighter alcohol intake with HTN leave unresolved whether the true alcohol-HTN relationship is linear, has a threshold, or, as suggested by several studies (Klatsky, 2000; Klatsky and Gunderson, 2006; Thadhani et al., 2002), is J shaped. Even if only heavy drinking were related to HTN, underreporting by some heavy drinkers could spuriously lower the apparent threshold or transform a threshold alcohol-blood pressure (BP) relationship to an apparently continuous one (Beilin, 1987; Romelsjo et al., 1995; Searles et al., 2002 Verkirk, 1992).
Method
Subjects and alcohol data
The study protocols were approved by the Institutional Review Board of the Kaiser Permanente Medical Care Program. Subjects were 105,378 black and white men and women who voluntarily took a health examination in 19781985 and completed a special alcohol research questionnaire offered during those years. The examination was offered by a comprehensive prepaid health care program in Oakland and San Francisco, CA. The basic examination procedure, without the special questionnaire for most years, was performed from 1964 to 1991. Examinees supplied questionnaire data about sociodemographics, medical history, and habits, and they underwent health measurements (Collen and Davis, 1969). The special alcohol questionnaire (1978-1985) inquired of drinkers about "usual number of drinks in the past year (wine, liquor, beer, or cocktails)." Checked responses yielded this categorization: lifelong abstainers = 9,874; exdrinkers = 3,562; those taking less than one drink per month = 20,397; those taking more than one drink per month but less than one drink per day = 41,066; those taking one to two drinks per day = 21,032; and those taking three or more drinks per day = 9,447. In this article, we refer to persons reporting fewer than three drinks per day as "light-moderate" drinkers.
Alcohol intake and associated BP were defined in persons with multiple examinations as the data on the first examination during 1978-1985. Among persons reporting fewer than three drinks per day on that index examination, determination of a presumptive greater likelihood of heavier drinking was made from other available computerized databases. Persons judged more likely to be underreporters were classified as "positive"; all others were classified as "negative." An individual was positive if he or she (I) reported a usual intake of three or more drinks per day on any examination other than the index one, either on another special research questionnaire (1978-1985) or on the examination questionnaire (1964-1991), or (2) had an inpatient or outpatient diagnosis of an alcohol-related condition (International Classification of Diseases, Ninth Revision [Commission on Professional and Hospital Activities {1978}, codes: 291 [alcoholic psychosis], 303 [alcohol dependence], 305 [alcohol abuse], 790.3 [excess blood alcohol level], 860-860.1 [alcohol poisoning], 265.2 [pellagra], 357.5 [alcoholic polyneuropathy], 425.5 [alcoholic cardiomyopathy], 535.3 [alcoholic gastritis], and 571 [chronic liver disease and cirrhosis]). An individual was negative if he or she never reported three or more drinks per day on any examination and had no computer-stored diagnosis of an alcohol-related condition.
Blood pressure ascertainment
Health measurements at the health examinations included supine systolic and diastolic BP. Examinees were asked to take no food or drink for 4 or more hours before the time of the appointment. At a specific examination stage, patients were asked to lie down on their backs and relax for several minutes. Then BP was measured once by a Roche Arteriosonde Blood Pressure Monitor, Model 1216 (Roche Products, Welwyn, Garden City, United Kingdom).
Other health measurements
These measurements included height and weight for all examinees. At baseline examinations from January 1978 through October 1979, a total of 37,620 persons had aspartate aminotransferase (AST) determinations. From January 1978 through August 1981, a total of 69,904 persons had alanine aminotransferase (ALT) determinations. Except for the examination date, there were no other selection factors for these liver enzyme tests.
Analytic methods
Multiple linear regression methods (analysis of variance) were used to assess differences in mean systolic BP and mean diastolic BP among 12 alcohol categories (negative and positive subgroups across all levels of alcohol intake). They were adjusted for the following covariates: gender, race/ethnicity, age; gender, ethnicity, age (decades); education (no college [referent], some college, college graduate); smoking (never smoked [referent], exsmoker, smoking less than one pack per day, smoking one or more packs per day); body mass index (<25 kg/[m.sup.2] [referent], 25-29 kg/[m.sup.2], [greater than or equal to] 30 kg/[m.sup.2]); and 12 alcohol categories (negative lifelong abstainers as referent, all other positive and negative subgroups).
HTN (systolic BP...
|
