original contributions
nature publishing group
Hyperinsulinemia and Homeostasis Model Assessment of Insulin Resistance as Predictors of Hypertension: A 5-Year Follow-Up Study of Korean Sample Ki-Chul Sung1, Soo Lim2 and Robert S. Rosenson3 Background The relationships between insulin level and indexes of insulin resistance (IR) to predict incident hypertension have been explored only in cross-sectional or prospective studies with small numbers of patients. We investigated whether plasma insulin concentration and the homeostasis model assessment of insulin resistance (HOMAIR) are associated with the onset of hypertension in a population of apparently healthy and relatively lean Korean adults.
common among older subjects than it was in younger subjects and was associated with a high baseline body mass index (BMI). In multivariable logistic models, elevated serum insulin, and HOMAIR were associated with an increased risk of incident hypertension in both sexes. In a multivariable analysis using quartiles of insulin and HOMAIR, the odds ratio (OR) for incident hypertension was the highest in the highest quartile of insulin and HOMAIR. The highest quartile of insulin and HOMAIR was associated with a 1.5–1.7 times increased risk of incident hypertension.
Methods We selected 10,894 of 15,638 subjects who were normotensive at baseline during general health status evaluations in 2003 and 2008. The baseline and follow-up examinations included analyses of fasting glucose, insulin level, and lipid profile. Alcohol consumption, smoking status, exercise habits, and education level were also evaluated using a standard questionnaire.
Conclusions This 5-year follow-up study provides evidence that both a high circulating insulin level and HOMAIR are significant risk factors for the development of hypertension in a relatively lean and healthy population.
Results Of the 10,894 subjects, hypertension developed in 881 (8.1%) during the ensuing 5 years. Incident hypertension was more
American Journal of Hypertension, advance online publication 26 May 2011; doi:10.1038/ajh.2011.89
Keywords: blood pressure; HOMAIR ; hypertension; insulin
Experimental studies have identified multiple molecular and cellular mechanisms through which elevated insulin level or insulin resistance (IR) elevates blood pressure.1–4 In addition to a review that demonstrated a relationship between IR and hypertension,5 earlier published studies6–8 have addressed this issue and found similar results. According to these hypotheses, IR and compensatory hyperinsulinemia are primary events, and enhanced sympathetic activity and diminished adrenal medullary activity are important links among defects in insulin action, the development of hypertension and associated metabolic abnormalities.5 However, the contribution of elevated
insulin level or IR in the prediction of newly diagnosed hypertension has been limited by the cross-sectional natures of previous studies,9,10 prospective studies of cohorts that included relatively small numbers of study subjects,11 and the inability to demonstrate independence using established risk factors for hypertension.12 Thus, we investigated the prognostic significance of insulin level and the homeostasis model assessment of insulin resistance (HOMAIR), a marker of IR, to predict the incidence of hypertension in a prospective, population-based study of apparently healthy Korean adults.
The first two authors have contributed equally to this work. 1Department of Internal Medicine, Division of Cardiology, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul, Korea; 2Department of Internal Medicine, Seoul National University College of Medicine, Seoul National University Bundang Hospital, Seongnam, South Korea; 3Department of Cardiology, Mount Sinai School of Medicine, New York, New York, USA. Correspondence: Ki Chul Sung (
[email protected])
Subjects. We reviewed the electronic medical records of 15,638 subjects who underwent general health status evaluations in 2003 with repeated health check-ups in 2008 at Kangbuk Samsung Hospital, Seoul, Korea. A majority of the subjects were employees of industrial companies or their family members. The employers paid most of the costs of the medical examinations, and a considerable proportion of the subjects
Received 12 December 2010; first decision 17 January 2011; accepted 6 April 2011. © 2011 American Journal of Hypertension, Ltd.
Methods
AMERICAN JOURNAL OF HYPERTENSION | VOLUME 24 NUMBER 9 | 1041-1045 | september 2011
1041
original contributions repeated the exam annually or biannually. We took advantage of this opportunity to conduct a follow-up study. Based on the 2003 data, we excluded individuals if they had a history of diabetes (n = 236), fasting plasma glucose ≥126 mg/dl (n = 382), history of hypertension (n = 723), systolic blood pressure (SBP) ≥140 mm Hg (n = 1,133), diastolic blood pressure (DBP) ≥90 mm Hg (n = 1,891) or if they had received antihypertensive or cardiovascular medications that affect blood pressure. Other exclusions included missing data for any of the following parameters: height or weight information (n = 26), exercise (n = 309), alcohol consumption (n = 399), smoking (n = 361), and education (n = 581). Since some individuals met more than one exclusion criterion, a total of 10,894 subjects were included in the final analysis. The study was approved by the institutional review board of Kangbuk Samsung Hospital. The initial health examinations performed in 2003 included medical histories, physical examinations, questionnaires about health-related behavior, and anthropometric and biochemical measurements. Medical histories and prescription drug use were assessed by the examining physicians, and all participants were asked to respond to a questionnaire regarding health-related behavior. Questions about alcohol intake assessed the frequency of alcohol consumption on a weekly basis and the amount consumed on a daily basis. We defined persons who reported that they smoked as current smokers. Regular exercise was assessed using a physician-administered questionnaire in which subjects were asked if they had participated in any regular exercise at least once a week, three times a month or none at all. Measurement of biochemical parameters. Blood samples were drawn from the antecubital vein after subjects had fasted >12 h. The fasting serum glucose, triglycerides, total, low density lipoprotein-cholesterol, and high density lipoproteincholesterol, uric acid, blood urea nitrogen, and creatinine levels were measured enzymatically using an automatic analyzer (Advia 1650 AutoAnalyzer; Bayer Diagnostics, Leverkusen, Germany). Insulin concentrations were measured with an immunoradiometric assay (Biosource, Nivelles, Belgium), with intra- and inter-assay coefficient of variations of 4.7% and 12.2%, respectively. Blood pressure was measured using a mercury manometer, between 08:00 am and 10:00 am, after the subject had been sitting upright for at least 10 min. Trained nurses measured seated blood pressure with a standard mercury sphygmomanometer. The first and fifth Korotkoff sounds were used to estimate SBP and DBP, respectively. When the SBP or DBP exceeded 140 mm Hg or 90 mm Hg, it was remeasured after a 5-min rest, and the results averaged. High blood pressure was defined as a SBP ≥140 mm Hg or a DBP ≥90 mm Hg. Height and weight were measured after an overnight fast with study participants wearing a lightweight hospital gown and no shoes. Body mass index (BMI) was calculated as weight (in kilograms) divided by the square of height (in meters). The level of IR was estimated using HOMA: fasting serum insulin (μU/ml) × fasting serum glucose (mmol/l)/22.5. 1042
Hyperinsulinemia and HOMAIR as Predictors of Hypertension
Definition of hypertension at the 2008 follow-up examination. Hypertension was defined as SBP ≥140 mm Hg, DBP ≥90 mm Hg, or history of hypertension during 2003–2008. Statistical analysis. The descriptive statistics for continuous variables were described as mean, s.d., median, and inter-quartile range (Q1, Q3). The categorical variables were described using frequencies and percentages. Comparisons of baseline cardiovascular risk factors according to the presence or absence of incident hypertension were analyzed using Student’s t-test or the Mann–Whitney test for continuous variables. Also, categorical variables were analyzed using the χ2 test or Fisher’s exact test. We conducted the Cochran–Armitage test to evaluate P values for trends. Insulin and HOMAIR did not follow a normal distribution, and log HOMAIR correlated relatively well with average results for euglycemic clamp tests previously conducted in Koreans.13 A natural log transformation (ln) was performed before we calculated the odds ratios (ORs) for incident hypertension in both sexes. Serum insulin and HOMAIR were divided into quartiles, and we mathematically adjusted the ORs for incident hypertension using multiple logistic regression models with potential confounding variables including baseline glucose, triglyceride, high density lipoprotein-cholesterol, baseline BMI, BMI change, SBP, DBP, age, sex, education level, smoking status, alcohol consumption, and regular exercise. Combined receiver operating characteristics curve analyses were performed to determine the predictors of hypertension with the comparison of area under the curve of each variable with the inclusion of confounding factors in the model. Area under the curve values were used to compare the discrimination abilities of insulin level and HOMAIR for hypertension. The data were analyzed, and statistical analyses for the data were performed using SPSS version 15.0 software (SPSS, Chicago, IL). All reported P values were two-tailed, and those
