Making Sense of Individual Responses to Sodium Reduction

More than 130 million adults in the US have hypertension,¹ and national rates of blood pressure (BP) control have declined in recent years.² This high burden of hypertension is driven in part by the US diet. The average US adult consumes 3.4 g of sodium daily, approximately 70% of which comes from the packaged and pre-prepared foods that define much of our culinary culture.³ Several randomized clinical trials have demonstrated that lowering sodium intake leads to lower BP, particularly among individuals with higher baseline BP.⁴ Population-level randomized interventions have shown that sodium restriction with potassium-based salt substitutes reduces the incidence of hypertension and stroke.^5,6 Accordingly, dietary sodium restriction is considered by many health care practitioners and public health officials to be a cornerstone of hypertension prevention and management. However, others argue that sodium restriction should not be recommended broadly due to insufficient evidence of cardiovascular risk reduction,⁷ and that sodium restriction may be best targeted only to individuals with high sodium intake and “salt-sensitive” hypertension.

In the 1970s, Guyton and Hall published several seminal articles that provided insights into the mechanistic underpinnings of hypertension, and their framework for salt-sensitive hypertension is still taught in physiology courses today.⁸ They observed that high sodium intake suppresses renin, which in turn results in a reduction in angiotensin II and aldosterone production, yielding a net vasodilatory and natriuretic effect. At any given BP, exceptionally high sodium intake increases BP slightly due to increased plasma volume initially, but in normal physiologic states, downregulation of the renin-angiotensin-aldosterone system prevents excessive increases in BP. However, Guyton and Hall postulated that in salt-sensitive individuals, the renin-angiotensin-aldosterone system develops a higher “set point” at which it is suppressed. As a result, high sodium intake in these individuals results in a net increase in plasma volume and peripheral vascular resistance, thus raising BP.⁹ Over the several decades since Guyton and Hall’s work, a large body of research has evolved and identified many interrelated mechanisms that contribute to this phenomenon, several of which may be targeted by novel therapies.¹⁰ Most of this work hinged on the idea that salt sensitivity is specific to individuals with distinctively dysfunctional sodium handling. Trials evaluating the effect of sodium intake on BP often excluded individuals taking antihypertensive medications to minimize noise from medication changes (dosing times or regimen). Although a number of small studies suggest ways in which sodium restriction might complement or enhance antihypertensive treatment, this remains an important area for further research.

In this issue of JAMA, Gupta and colleagues¹¹ present the results of a crossover trial in which 213 middle-aged and older adults at 2 US centers (University of Alabama at Birmingham and Northwestern University in Chicago) were assigned to very low-sodium and very high-sodium diets for 1 week each in random order. Differences in BP were assessed using 24-hour ambulatory BP monitoring at the end of each week. The low-sodium diet was restricted to 0.5 g of sodium daily as part of a standardized diet that also included 4.5 g of potassiumand1g of calciumdaily. The high-sodium diet encompassed 2.2 g of sodium daily added to usual intake using bouillon cubes. Participants were a median of 61 years of age (range, 50–75 years), 65% were female, 25% were normotensive and not taking antihypertensive medication, 20% had well-controlled hypertension while taking antihypertensive medications, 31% had uncontrolled hypertension despite taking antihypertensive medications, and 25% had uncontrolled hypertension and were not taking antihypertensive medications. The authors observed a median 4–mm Hg decline in mean arterial pressure (the primary end point, based on precedence from prior studies of salt sensitivity¹²) and a 7–mm Hg decline in systolic BP with the low-sodium diet vs the high-sodium diet. Notably, 73% of participants experienced a decline in mean arterial pressure, signifying a spectrum of responsiveness to sodium restriction. In exploratory analyses, there was no difference in response to sodium restriction across subgroups of age, sex, race, hypertension diagnosis, antihypertensive therapy, baseline BP, diabetes status, body mass index, study site, or order of receiving the low- and high-sodium diets.

Although the authors observed meaningful differences in sodium intake between the randomization groups, the participants did not reach their prescribed levels of sodium intake. Sodium intake was estimated using 24-hour urine sodium, which demonstrated that participants took in a median of 4.5 g of sodium daily during their usual diet, 5.0 g daily during the high-sodium diet (prescribed as 2.2 g of sodium added to their usual diet), and 1.3 g daily during the low-sodium diet (prescribed as a standardized diet with 0.5 g of sodium daily). The trial population represented individuals with higher baseline sodium intake (>1 g more per day) than the general US population. Very little difference was seen in BP between the usual and high-sodium diets, consistent with the minimal observed difference in 24-hour urine sodium between the 2 diets.

Failure of participants to reach the prescribed level of sodium intake reflects some underlying limitations of the study. First, the diets were not fully controlled. The high-sodium diet was very pragmatic, with participants being given 2 bouillon cubes to add to their usual diet. The low-sodium diet was standardized but, importantly, it included high intake of potassium and calcium that could further lower BP. Although there were no statistically meaningful differences in urinary potassium levels between the dietary groups, there are known limitations to single 24-hour urine collections for quantification of potassium intake.¹³ There is evidence that potassium is primarily important for BP reduction in the context of high sodium intake.^5,6,14 However, this study is unable to inform ongoing debates of the relative importance of either micronutrient. Ultimately, when this trial’s intervention is evaluated in the context of the existing literature, it may be better depicted as a low-sodium, high-potassium, and high-calcium intervention than purely as a sodium reduction trial.

Additionally, the sodium content of the low-sodium diet was exceptionally low, which may be why participants did not reach the prescribed level. In contrast, the American Heart Association recommends dietary restriction to less than 1.5 g of sodium daily among individuals with hypertension (or, if not feasible, at least a 1-g daily reduction in sodium)¹; this is the most stringent guideline in this regard. Furthermore, there was likely insufficient time to reach a steady state of sodium and potassium. Participants received each diet for only 1 week, with no washout period in between. Prior evidence suggests that dietary sodium restriction takes several weeks to reach full effect, and that 1 week is not nearly enough to see a full response. For example, the DASH-Sodium trial randomized 412 adults with prehypertension or stage 1 hypertension who were not taking antihypertensive medications to 3 levels of sodium intake (1.2 g, 2.3 g, or 3.5 g daily) for 4 weeks each.¹⁵ The primary trial demonstrated a dose-response relationship of sodium restriction with BP lowering, with a decline in systolic BP of 7 mm Hg during the lowest vs highest dietary sodium diet, similar to the current trial. In a post hoc analysis, the low-sodium diet reduced BP incrementally on a weekly basis without plateauing, suggesting that the full effects of sodium restriction occur more than 4 weeks after initiating dietary changes.¹⁶ Finally, BP is highly variable, even when measured using ambulatory BP monitoring. While it may appear that nonresponse represents salt insensitivity, measurement error cannot be excluded as the cause of some of the extreme responses observed. Thus, caution should be applied in attributing biologic mechanisms to different changes in BP in response to the intervention.

More than 50 years after Guyton and Hall developed their conceptual model of salt-sensitive hypertension, we now have evidence that BP response to sodium intake may be more of a spectrum than a binary threshold. The vast majority of middle-aged and older adults with high baseline sodium intake experienced BP lowering during a very low-sodium, high-potassium, and high-calcium diet vs a very high-sodium diet, independent of hypertension status and antihypertensive medication use. Altering diet in an era in which the world is increasingly reliant on processed and pre-prepared foods may be unachievable. Public health efforts focused on developing policies to lower sodium and increase potassium content in these types of food products could have enormous benefits for BP control.