Can a Bedside Measure of Visceral Adiposity Predict Progression of Prehypertension to Hypertension?

Elevated blood pressure (BP) is a major risk factor for cardiovascular disease (CVD) and death, and observational data suggest the optimal systolic BP may be as low as 90 mm Hg.1 Guideline committees devise diagnostic criteria for hypertension (HTN) based on cohort studies showing the BP levels associated with greatest risk, and on randomized controlled trials (RCTs) demonstrating the thresholds at which BP-lowering interventions reduce morbidity and mortality. But it is likely that many patients with BP below the cutoff for the diagnosis of HTN (130/80 mm Hg in the most recent US guideline2) could benefit from BP lowering. While the absolute risk reduction at the individual patient level would be small, the population-level benefit could be substantial.3 Prehypertension, previously defined as a BP of 120–139/80–89 mm Hg, is one classification of patients with “normal” but suboptimal BP4 and affects approximately 26% of Americans.5 First introduced in the Seventh Report of the Joint National Committee (JNC7), prehypertension has been subsumed by the newer categories of elevated BP (120–129/<80 mm Hg) and stage 1 HTN (130–139/80–89 mm Hg) defined by the 2017 American College of Cardiology/American Heart Association (ACC/AHA) guideline.2 Although it comes from an era when the threshold for diagnosis of HTN was higher, prehypertension remains a useful construct because it includes all patients with suboptimal BP who will not necessarily be prescribed pharmacological therapy initially. Prehypertension frequently progresses to HTN within several years, especially in patients age 65 and older.6 More importantly, prehypertension independently predicts the development of CVD in meta-analyses of observational studies.7,8 Some, but not all, of these studies have shown significant associations with CVD for both the elevated BP and stage 1 HTN categories. In addition, up to 34% of patients with prehypertension have masked HTN,9 which confers elevated risk for cardiovascular events10 and usually merits treatment with antihypertensive medication.2 In light of these data from non-randomized trials, the hypothetical case for BP lowering in prehypertensive patients is compelling. Nevertheless, the US healthcare system is struggling to control BP in patients who are already being treated for HTN, more than half of whom have not achieved their goal BP.5 Before adding to the burden of the HTN workforce, is it possible to identify those prehypertensive patients who are most likely to progress to HTN, and target lifestyle and even pharmacologic interventions to them? Focusing on patients with obesity is a promising approach. The pathophysiology of HTN in obesity is multifactorial, including sympathetic nervous system and renin–angiotensin system activation, increased aldosterone synthesis by adipocytes, insulin resistance, and comorbid chronic kidney disease and obstructive sleep apnea.11 Overweight, defined as a body mass index (BMI) of 25.0–29.9 kg/m2, and obesity (BMI ≥30 kg/m2) account for 26%–40% of new cases of HTN.12,13 Additionally, a considerable portion of the excess risk of coronary heart disease and stroke due to elevated BMI (31% and 65%, respectively) is mediated by HTN.14 Both obesity and weight gain increase the risk of incident HTN in prospective cohort studies.6,15 Despite its ease of use, BMI alone does not distinguish fat mass from lean body mass, so it may not be the best marker of obesity to risk stratify prehypertensive patients. Instead, it appears that visceral adiposity is what drives the development of HTN. In an analysis of normotensive patients in the Dallas Heart Study, BMI no longer independently predicted incident HTN when a measure of visceral adipose tissue was added to the regression model.16 Measures of visceral adiposity range in complexity from waist circumference (WC) to magnetic resonance imaging. The Chinese visceral adiposity index (CVAI) is calculated from demographics (age and sex), anthropometrics (BMI and WC), and commonly collected laboratory values (high-density lipoprotein [HDL] and triglycerides), making it easily accessible to the primary care clinician. In several studies, it was a better predictor of prehypertension and HTN than other readily available obesity indices.17,18 In this issue, Dai et al. investigated the ability of the CVAI to predict BP trajectory in the medium term among patients with prehypertension.19 For their analysis, they used data from the China Health and Retirement Longitudinal Study (CHARLS), a prospective cohort that enrolled adults age 45 and older and collected health data periodically by questionnaire. Participants had office BP and anthropometrics measured and blood samples collected at baseline (2011) and 4 years later (2015). Of the nearly 18,000 participants in CHARLS, the authors identified 2,057 participants with prehypertension; those with a self-reported history of HTN or who were taking antihypertensive medications were excluded. On average, included participants were in their late 50s, equally likely to be male or female, and overwhelmingly rural-dwelling and married, with a low level of education and burden of chronic diseases. Over the observation period, 34% of participants progressed to HTN, 27% regressed to normal BP, and 39% remained prehypertensive. The authors computed CVAI for all participants and dichotomized them into groups with low and high baseline CVAI. In a logistic regression model with numerous covariates, high CVAI was associated with higher odds of progression to HTN (odds ratio [OR] 1.66, 95% confidence interval [CI] 1.35–2.05) and lower odds of regression to normal BP (OR 0.58, 95% CI 0.47–0.73). A restricted cubic spline model showed a linear relationship between increasing CVAI and increasing OR for progression to HTN, as well as between increasing CVAI and decreasing OR for regression to normal BP. Similar results were seen in all subgroups, and in a sensitivity analysis excluding participants with diabetes and dyslipidemia. Notably, when the analysis was restricted to the 1,084 participants who met criteria at baseline for elevated BP but not stage 1 HTN, the results were also comparable (OR of high CVAI for progression to stage 1 or higher HTN 2.15, 95% CI 1.62–2.85; OR for regression to normal BP 0.52, 95% CI 0.38–0.70). Dai et al. should be commended for their contribution to the literature on factors associated with the progression of prehypertension to HTN in general, and on the predictive power of the CVAI in particular. Strengths of their study include the large sample size, prospective nature, adjustment for multiple potential confounders, and appropriate subgroup and sensitivity analyses. There were, however, some limitations. First, more than 80% of the participants with prehypertension in the CHARLS cohort were excluded due to missing data, raising the possibility of selection bias. Second, the population consisted mainly of late-middle-aged, relatively healthy individuals living in rural China, so the results may not apply to patients who are older or younger, are urban-dwelling, live in other parts of the world, or have multiple chronic conditions. Third, BP was measured in, not outside, the office at baseline and follow-up; consequently, the impact of white coat and masked HTN on the findings cannot be determined. Finally, the CVAI was the only measure of obesity in the regression model, so it is conceivable that another index (e.g., A Body Shape Index [ABSI], which is BMI-independent) might predict BP trajectory even better. How should clinicians respond to the results of Dai et al.’s study? Further work is needed to confirm the relationship between visceral adiposity and the progression of prehypertension to HTN. Moreover, we must be humble in the face of a paucity of evidence that BP-lowering interventions are effective for the primary prevention of CVD in patients with prehypertension. An RCT enrolling prehypertensive patients at above-average risk of progression due to high CVAI could help address this unknown. In the meantime, it is tempting to speculate: Should we offer more than lifestyle modifications to patients with elevated BP and high CVAI (perhaps computed by a widget in the electronic medical record)? And should we use the CVAI as an adjunct to current atherosclerotic CVD risk calculators in determining when to recommend pharmacotherapy to patients with stage 1 HTN? More intriguing still, as we enter a new era of highly effective weight loss medications that also lower BP,20 is the question of whether to expand the HTN pharmacopeia to agents that specifically target obesity. By preventing incident cases of HTN, these approaches could remove fuel from the fire of a public health problem already burning out of control. Dr Moreno serves as a medical advisor and has done paid consultant work for Healthline Media.