Dietary Sodium Restriction for Heart Failure: A Systematic Review of Intervention Outcomes and Behavioral Determinants

Abstract

The 1500mg/day dietary sodium-restriction commonly recommended for patients with heart failure has recently been questioned. Poor adherence to sodium restricted diets makes assessing the efficacy of sodium restriction challenging. Therefore, successful behavioral interventions are needed. We reviewed sodium restriction trials and descriptive studies of sodium restriction to: 1) determine if sodium restriction was achieved in interventions among heart failure patients; and 2) characterize predictors of successful dietary sodium restriction. Among 638 identified studies, 10 intervention trials and 25 descriptive studies met inclusion criteria. We used content analysis to extract information about sodium restriction and behavioral determinants of sodium restriction. Dietary sodium was reduced in seven trials; none achieved 1500 mg/day (range=1938–4564mg/day). The interventions implemented in the interventional trials emphasized knowledge, skills, and self-regulation strategies, but few addressed the determinants correlated with successful sodium restriction in the descriptive studies (e.g., social/cultural norms, social support, taste preferences, food access, self-efficacy). Findings suggest incorporating determinants predictive of successful dietary sodium restriction may improve the success of interventional trials. Without effective interventions to deploy in trials, the safety and efficacy of sodium restriction remains unknown.

INTRODUCTION

Heart failure, which affects approximately 26 million people worldwide,(1) represents the most common indication for a sodium restricted diet.(2, 3) This recommendation is based on a hypothesis that low cardiac output leads to activation of renin which increases sodium retention.(4) This renin-angiotensin-aldosterone system-activated sodium retention, coupled with excess dietary sodium, is theorized to result in fluid retention in heart failure patients.(5) This theory is supported by observational studies that associate sodium restricted diet with lower symptom burden,(6) hospitalization,(7) mortality,(7) and event-free survival.(8) Lapses in sodium restricted diet adherence have been associated with readmission and mortality.(9, 10) American Heart Association guidelines recommend a 1500 mg/day sodium restricted diet for patients with early stage heart failure.(5)

However, data indicating the efficacy of sodium restricted diets as treatment for heart failure is inconsistent(11) and has recently been called into question. Sodium restriction has been associated with hospital readmission(12, 13) and mortality(12) as well as a dysregulated neurohormonal profile among compensated congestive heart failure patients.(14) It remains unclear if these inconsistencies are due to poor dietary measurement, reverse causality, or a biological mechanism of sodium homeostasis.(12, 15, 16)

Adherence to sodium restricted diets is poor among heart failure patients.(7, 16) This calls into question whether the trials assessed to form the current guidelines and recommendations accurately reflect the influence of sodium intake and sodium restricted diets on heart failure outcomes. Dietary interventions that successfully achieve sodium restriction are necessary to assess the effectiveness of sodium restriction on heart failure-relevant outcomes.

Improved behavioral interventions for heart failure self-management, and specifically for sodium restriction, are needed(17) but the path forward is unclear. The quality of behavioral interventions can be improved by identifying and targeting determinants of behavior change. However, the determinants incorporated into dietary sodium restriction interventions have not been examined systematically. Determinants of behavior are organized in a number of frameworks, e.g., Integrated Behavior Model.(18) Frameworks (e.g., (19)) that specifically address food choice and dietary behavior change identify determinants that should be considered in behavioral nutrition interventions. These determinants are key to improving trials of sodium restriction’s effect on heart failure outcomes.

Reviews have examined self-management interventions that include sodium restricted diet: One systematic review found self-management interventions correlated with improved health behavior(17), another review indicated that self-care behaviors improved with education.(20) The European Society of Cardiology recommends tailored skills-based education with social support.(21) No systematic reviews have yet addressed sodium restricted diet interventions in heart failure patients.

In this review, we aimed to determine if sodium restriction has been achieved in interventions among heart failure patients and then characterize why it is or is not. We conducted a systematic review of randomized controlled trials and descriptive studies of sodium restricted diets to 1) assess efficacy of sodium restricted diet interventions tested in randomized controlled trials, 2) identify determinants of dietary sodium restriction targeted in the trialed sodium restricted diet interventions, and 3) identify determinants of dietary sodium restriction identified in descriptive studies of sodium restriction. Then, we compared determinants targeted in randomized controlled trials with those identified in descriptive studies.

METHODS

First, we searched PubMed, CINAHL, and Cochrane databases for English-language studies published January 2000-December 2018 with the following terms: heart failure or congestive heart failure; dietary sodium restriction, salt restriction, salt limit, sodium limit, dietary salt, low-sodium, or low salt; and intervention, education, behavior, self-manage, self-care, knowledge, disease management, or secondary prevention. We included peer-reviewed randomized controlled trials of interventions for heart failure patients, with sodium restriction as a primary or secondary outcome, and quantitative and qualitative descriptive studies of adherence to sodium restricted diets. Eligible randomized controlled trials reported sodium restricted diet efficacy and behavioral determinants, whereas descriptive studies did not assess sodium restricted diet efficacy, instead focusing on determinants of sodium restricted diet adherence. Our search strategy is summarized in Figure 1. A review protocol was not pre-registered.

Figure 1.

Systematic search and review process

Next, we conducted a content analysis. Two reviewers coded each study for determinants of dietary behavior change, using the Integrated Behavior Model(18) and Contento framework(19) as a guide. Reviewers met regularly throughout the coding process, developing a comprehensive list of determinants represented in the included studies and resolving all differences by consensus.

Finally, we compared determinants addressed in the randomized controlled trial sodium restriction interventions and those identified in the descriptive studies.

RESULTS

Among 638 identified studies, 10 intervention trials and 25 descriptive studies met inclusion criteria.

Intervention studies

Ten randomized controlled trials that included sodium restricted diet were published in 2000–2018(22–31) and are summarized in Table 1. Studies were conducted in the US,(24, 26, 29–31) Canada,(23) Mexico,(25) Brazil,(22) Sweden,(28) and Japan(27) and ranged from 46 to 480 participants. Primary aims of these studies included: sodium reduction,(22, 26, 30) sodium restricted diet adherence,(23, 24) clinical status,(23, 27, 31) symptoms,(27) and self-care(29, 31); studies that did not include sodium reduction or sodium restricted diet adherence as primary aims included sodium reduction among secondary aims. Seven of these studies claimed to be effective in reducing sodium intake(23, 25, 26, 28, 30) or increasing sodium restricted diet adherence,(27, 31) based on 3-day food records (3DFR),(25, 26, 30) 24-hour urinary sodium excretion,(23, 25, 26, 28) or self-reported adherence.(27, 31) One study was inconclusive, finding self-reported reductions in sodium intake with 24-hour dietary recalls, but no reduction in urinary sodium excretion.(22) Two trials found no difference between intervention and control.(24, 29)

Table 1.

Description of sodium restriction randomized trials for heart failure patients

Reference	Location	Sample Size	Population	Intervention	Comparison	Outcome	Average sodium Intake in experimental group at post-test
Alvez, 2012(22)	Brazil	n=46	Outpatients at tertiary university hospital	2 nutrition education sessions	usual care	Reduced self-reported, but not urinary Na intake. Urinary Na was 3-fold higher than self-report	4564 mg/d (urinary sodium)
Arcand, 2005(23)	Toronto, Canada	n=47	Ambulatory heart failure clinic patients	2 g/day Na RX + 2 RD counseling sessions	2 g/day Na RX	Reduced Na based on 3-day food record	2140 mg/day
Athar, 2018(24)	USA	n=97	general medicine inpatients in urban academic medical center	ultrasound image of inferior vena cava and explanation of relation to fluid status	ultrasound image with general heart failure information	No difference in self-reported sodium restriction adherence	Not measured
Colin-Ramirez, 2004(25)	Mexico City, Mexico	n=65	heart failure clinic patients	sodium and fluid restriction diet RX	usual care	Decreased urinary Na excretion @ 6 mos. compared to control	1942mg/day
Dunbar, 2013(26)	Atlanta, USA	n=61	Patient/Family member dyads at outpatient heart failure clinic	family heart failure education or family partnership intervention	usual care	Lower urinary Na in family partnership group @ 8 months than usual care	1938 mg/day measured through 3DFR
Otsu, 2012(27)	Japan	n=94	Outpatients at cardiovascular clinic	self-monitoring and monthly educational sessions	usual care	Self-reported sodium reduction @ 12 months, not @ 24 months	Not measured
Philipson, 2013(28)	Sweden	n=97	Patients with heart failure history from three hospitals	RD counseling session	general information about heart failure provided by RD	Significant reduction in urinary sodium compared to control group @ 12 weeks	Not calculated
Veroff, 2012(29)	USA	n=480	Medicare Advantage patients with heart failure from large not-for-profit health plans	Mailed “Living with heart failure” DVD and booklet	mailed basic heart failure information	No significant change in sodium restriction	Not measured
Welsh, 2013(30)	Kentucky, USA	n=52	heart failure patients recruited from cardiology clinic, community hospital, and university hospital	weekly home or phone counseling sessions	usual care	Lower Na intake in intervention group @ 6 months	2262 mg/day
Young, 2016(31)	USA	n=100	heart failure patients discharged from rural critical access hospital	inpatient self-management training + telephone sessions and toolkit	usual care	No significant difference in estimated daily sodium intake based on urine test	3749 mg/day

No interventions consistently achieved the 1500 mg/day recommended sodium intake level for adults with heart failure when sodium intake was estimated with 24-hour urinary sodium measurements (according to varying conversion assumptions). Average sodium intake at post-test ranged from 1938mg/day(26) to 4564mg/day(22).

Descriptive studies

Twenty-five studies describing determinants of sodium restricted diet adherence(32–56) were published in 2000–2018, summarized in Table 2. The included studies were conducted in the US(32–40, 42–47, 51, 52, 54, 56), the Netherlands,(48, 55) Taiwan,(41) China,(49) Canada,(50) Korea,(53) and Australia(46) and used qualitative,(33, 40–42, 49, 52, 55) quantitative(32, 34–37, 39, 44–48, 50, 51, 53, 54, 56) or mixed methods(38, 43) to explore determinants of sodium restriction.

Table 2.

Description of studies of determinants of sodium restriction in heart failure patients

Reference	Location	Study Type	Sample Size	Population
Bennett, 2005(32)	USA	Descriptive, Cross-sectional	Sample 1: n=101, Sample 2 n=205	Sample 1: heart failure patients from VA heart clinics and private practice, Sample 2: inpatients from county and VA hospitals
Bentley, 2005(33)	Southern USA	Qualitative descriptive study	n=20	Convenience sample of community-dwelling heart failure patients receiving care at heart failure clinic
Bidwell, 2018(34)	Southeastern USA	Secondary analysis of cross-sectional baseline data from prospective trial	n=114 patient-caregiver dyads	Patient-caregiver dyads from 3 outpatient heart failure clinics
Chung, 2006(35)	Kentucky, Ohio, Australia	Comparative descriptive study	n=68	heart failure outpatient clinic patients
Chung, 2015(36)	Kentucky, Georgia, Indiana	Secondary analysis	n=379	heart failure outpatients who objectively followed sodium restricted diet
Chung, 2017(37)	Kentucky, USA	Cross-sectional descriptive study	n=74	Community-dwelling outpatients with chronic heart failure from 2 community hospitals and 1 academic medical center
Dickson, 2013(38)	USA	Mixed-methods concurrent nested study	n=30	Black adults with confirmed heart failure recruited from heart failure clinic and inpatient units at urban medical center
Dunbar, 2016(39)	Southeastern USA	Secondary analysis of baseline data from randomized controlled trial	n=117 patient-family dyads	Patient-family dyads from 3 outpatient heart failure clinics
Gary, 2006(40)	USA	Descriptive study	n=32	Women over 50 with diagnosed heart failure from outpatient clinic
Jiang, 2013(41)	Taiwan	Qualitative descriptive study	n=12	Elderly Chinese heart failure patients
Heo, 2009(42)	Southern USA	Qualitative descriptive study	n=20	Convenience sample of patients with heart failure from outpatient clinics
Holden, 2015(43)	Southern USA	Descriptive observational study	n=30 patients, n=14 informal caregivers	Patients with heart failure and their informal caregivers
Hooker, 2018(44)	USA	Cross-sectional study of baseline data from randomized controlled trial	n=99 patient-caregiver dyads	Patient-caregiver dyads identified as part of randomized controlled trial with n=314 participants
Kollipara, 2008(45)	Dallas, TX, USA	Observational	Sample 1: n=97, Sample 2: n=48	heart failure patients treated at urban public hospital serving indigent patients
Lennie, 2008(46)	USA and Australia	Observational	n=246	Patients with heart failure referred from academic medical centers
Luyster, 2009(47)	USA	Cross-sectional	n=88	heart failure patients with a mean age of 70 years, treated with ICD
Nieuwenhuis, 2011(48)	Netherlands	Cross-sectional correlational study	n=84	heart failure patients
Rong, 2017(49)	China	Qualitative descriptive study	n=15	Subjective sample of Chinese heart failure patients
Schnell-Hoehn, 2009(50)	Canada	Cross-sectional	n=65	Convenience sample of ambulatory heart failure patients recruited from outpatient clinics
Sethares, 2014(51)	Northeastern USA	Exploratory, longitudinal descriptive design	n=78	heart failure patients recruited from community hospital
Sheahan, 2008(52)	USA	Qualitative design	n=33	Older women aged 65–98 residing in 3 congregate living facilities in the high-risk “coronary valley” of USA
Song, 2009(53)	Seoul, South Korea	Prospective study	n=254	heart failure patients referred from outpatient heart failure clinics at two tertiary medical centers
Subramanian, 2008(54)	Midwestern USA	Cross-sectional	n=259	heart failure patients receiving care at 2 VA hospitals
Van der Wal, 2010(55)	Netherlands	Qualitative descriptive study	n=15	heart failure patients participating in multicenter trial
Wu, 2017(56)	Midwestern USA	Secondary analysis of cross-sectional study	n=244	heart failure patients from clinics associated with 3 large community hospitals/academic medical centers

Content Analysis

Nine determinants of sodium restricted dietary behavior consistent with health behavior theories were addressed across the 35 studies in our review: knowledge and skills, outcome expectations, social and cultural norms, social support, taste preferences, food access and availability, perceived barriers and self-efficacy, self-regulation, and habit. We added mental health (i.e., anxiety and/or depression) as a determinant of sodium restriction because it was also common in included studies. The final set of determinants included in the content analysis is summarized in Figure 2. Below we present a summary of how determinants were addressed in randomized controlled trials and descriptive studies. The findings of this content analysis are summarized in Table 3.

Figure 2.

Conceptual framework dietary behavior change mediators in heart failure

Table 3.

Content Analysis of intervention targets

Reference	Knowledge and Skills	Outcome Expectations	Social and cultural Norms	Social support	Taste preferences	Food access and availability	Perceived barriers, self-efficacy	Anxiety, Depression	Self-regulation	Habit
Randomized controlled trials
Alvez, 2012	X								X
Arcand, 2005	X								X
Athar, 2018	X	X
Colin-Ramirez, 2004	X
Dunbar, 2013	X	X		X					X
Otsu, 2012	X			X					X
Philipson, 2013	X		X
Veroff, 2012	X
Welsh, 2013	X	X			X				X
Young, 2016	X								X
Descriptive Studies
Bennett, 2005	X	X	X		X	X		X
Bentley, 2005	X	X	X	X	X	X
Bidwell, 2018	X			X
Chung, 2006	X	X	X	X	X	X
Chung, 2015	X		X	X
Chung, 2017	X	X	X		X				X
Dickson, 2013	X	X	X	X	X		X
Dunbar, 2016				X				X
Gary, 2006	X		X		X			X
Jiang, 2013	X	X	X
Heo, 2009	X	X	X	X	X	X
Holden, 2015	X	X		X	X			X		X
Hooker, 2018				X			X
Kollipara, 2008	X
Lennie, 2008	X		X	X			X
Luyster, 2009								X
Nieuwenhuis, 2011							X
Rong, 2017	X		X		X	X				X
Schnell-Hoehn, 2009							X	X
Sethares, 2014		X	X		X	X
Sheahan, 2008	X	X	X	X	X
Song, 2009								X
Subramanian, 2008	X						X
Van der Wal, 2010	X	X			X	X
Wu, 2017			X

The 10 randomized controlled trials addressed fewer determinants than the descriptive studies. The six determinants identified in the content analysis that were addressed in the randomized controlled trials included: knowledge and skills,(22–31) outcome expectations,(24, 26, 30) social and cultural norms,(28) social support,(26, 27) taste preferences,(30) and self-regulation.(22, 23, 26, 27, 31)

All 10 determinants were discussed across the 25 descriptive studies: knowledge and skills,(32–38, 40, 42, 43, 45, 46, 49, 52, 54, 55) outcome expectations,(32, 33, 36–38, 41–43, 51, 52, 55) social and cultural norms,(32, 33, 35–37, 40–42, 46, 49, 51, 52, 56) social support,(33–35, 38, 42–44, 46, 52) taste preferences,(32, 33, 35, 37, 38, 40, 42, 43, 49, 51, 52, 55) food access and availability,(32, 33, 35, 42, 49, 51, 55) perceived barriers and self-efficacy,(32, 37, 38, 44, 48, 50, 54, 56) anxiety and/or depression,(32, 39, 40, 43, 47, 50, 53) self-regulation,(37) and habit.(43, 49)

Knowledge and skills.

All randomized controlled trials intervened on knowledge and skills. Interventions included group education,(22, 23, 26, 27) one-on-one counseling,(24, 25, 27, 28, 31) written materials,(22–26, 28, 31) and a DVD.(26, 29) Education was provided in-person,(22–28, 30, 31) and over the phone.(26, 28, 30, 31) The study that provided a DVD and written handout through the mail in lieu of in-person education was ineffective.(29) Knowledge and skills included information about sodium,(22, 23, 25–28, 30) label reading,(23, 30) recipe adaptation,(26, 28) eating in restaurants,(23) alternative seasoning techniques,(23, 28, 30) and meal planning.(26, 30) Education was provided by registered dietitians,(23, 25, 26) nurses,(26, 30) and ultrasonographers.(24)

Knowledge and skills were important determinants of sodium restricted diet adherence in 17 (65%) of the descriptive studies.(32–38, 40–43, 45, 46, 52, 54, 55) One(45) found only knowledge and skills to be an important determinant of sodium restricted diet adherence, while the other 16 found that other determinants were important in addition to knowledge and skills.

Outcome expectations.

Three of ten randomized controlled trials incorporated outcome expectations by addressing beliefs about the consequences of a behavior and includes perceived risks and perceived benefits.(19) All three interventions discussed the rationale for following sodium restricted diet; two(24, 30) explicitly linked this education to improved outcomes and decreased symptom burden. One(30), which highlighted the relationship between sodium restricted diet and fewer heart failure symptoms, resulted in lower sodium intake at 6 months. Another(24), which aimed to motivate patients by using ultrasound to show patients damage to their inferior vena cava, was unsuccessful.

Eleven (44%) of the descriptive studies illustrated how outcome expectations were related to sodium restricted diet.(32, 33, 35, 37, 38, 41–43, 51, 52, 55) Individuals were motivated to participate in self-care, including a sodium restricted diet, in order to experience decreased symptom burden and hospitalization,(55) and improved health and longevity.(51) Descriptive studies also elucidated that those who did not understand or experience the relationship between diet and symptom burden were less likely to adhere to sodium restricted diet.(32, 33, 41, 43) One study reported patients felt that heart failure was inevitable, and therefore unrelated to health behaviors like sodium restricted diet.(38)

Social and cultural norms.

Social norms consist of beliefs about how others think one should behave and perceptions about what others are doing.(19) Only one randomized controlled trial provided dietary education in line with patients’ social and cultural norms, and was effective in reducing urinary sodium.(28) On the other hand, social and cultural norms were associated with sodium restricted diet in 14 descriptive studies.(32, 33, 35–38, 40, 41, 41, 42, 46, 49, 51, 52, 56) Individuals struggled to follow sodium restricted diet if their family did not follow it with them(35, 42). Conversely, individuals whose family members followed sodium restricted diet were more successful.(36) Many did not feel their families would accept low-sodium foods.(40) Eating alone, or eating different low-sodium foods from family members, were barriers to sodium restricted diet.(32, 33, 52) Individuals were more likely to follow sodium restricted diet when they believed it was important to their family or doctor.(37, 56) Lastly, cultural norms, such as traditional medicinal beliefs and cultural foods, influenced sodium restricted diet adherence.(38, 49)

Social support.

Two of ten randomized controlled trial interventions considered social support, though the interventions differed greatly. One successful intervention(26) included a caregiver, most often a spouse, in all aspects of the intervention, including supportive communication and self-care role-playing. In an unsuccessful randomized controlled trial(27), investigators provided a letter to each patient’s family.

Eleven descriptive studies (44%) found social support to influence sodium restricted diet adherence.(33–36, 38, 39, 42–44, 46, 52) Three studies found sodium restricted diet interfered with social life.(33, 46, 52) Informal caregivers influenced sodium restricted diet in five studies(26, 34, 36, 43, 44) and healthcare providers influenced sodium intake in one study.(43)

Taste Preferences.

One randomized controlled trial provided(30) educational activities tailored to taste preferences by helping patients to identify their favorite low-sodium foods at restaurants, plan sample menus, and familiarize themselves with low-sodium substitutes. This intervention resulted in lower sodium intake.

Taste was an important determinant of sodium restricted diet adherence in 12 (48%) of the included descriptive studies.(32, 33, 35, 37, 38, 40, 42, 43, 49, 51, 52, 55) Low palatability of sodium restricted diets was a barrier to adherence,(32, 33, 35, 42, 51) as was preference for saltier foods.(37, 40, 42, 52, 55)

Food Access and Availability.

No randomized controlled trials intervened on food access and availability. However, food access and availability was associated with sodium restricted diet adherence in eight descriptive studies.(32, 33, 35, 40, 42, 49, 51, 55) The cost of healthy food,(32, 40, 49, 51, 55) as well as the difficulty identifying low-sodium foods in grocery stores and on restaurant menus(33, 35) and insufficient low-sodium choices (32, 42, 51) were challenges to sodium restricted diet adherence.

Perceived Behavioral Control and Self-efficacy.

Self-efficacy refers to an individual’s confidence in carrying out an intended behavior.(19) No randomized controlled trials intervened on self-efficacy or perceived behavioral control, while six descriptive studies identified self-efficacy as an important determinant of adherence to a sodium restricted diet.(38, 44, 46, 48, 50, 54) Perceived behavioral control, such as control over diet due to family roles, was also associated with adherence.(46)

Anxiety and Depression.

No randomized controlled trials intervened on mental health. Seven descriptive studies found that anxiety and depression were associated with decreased adherence to sodium restricted diets, with increased anxiety and depression decreasing self-care, including sodium restricted diet.(32, 39, 40, 43, 46, 50, 53)

Self-regulation.

Self-regulation refers to the ability to handle challenges, manage emotions, and avoid temptations.(57) Six of ten randomized controlled trials intervened on self-regulation. Three studies, one of which was effective in reducing sodium intake based on 3-day food records, helped subjects with goal setting;(22, 23, 31) three studies, one of which was successful based on urinary sodium, included self-monitoring.(26, 27, 31) Randomized controlled trials also addressed self-regulation via care planning,(23) reminders and cues,(26) strategies for controlling emotions,(27) and keeping a diary.(30) Only one descriptive study related self-regulation to sodium restricted diet adherence, finding that those who struggled to adhere to sodium restricted diet reported “lacking the will power” to change their diet(37).

Habit.

No randomized controlled trials intervened on habit. Two descriptive studies found that habits incompatible with sodium restricted diet made it difficult for participants to adhere to a sodium restricted diet.(43, 49) For example, one participant reported, “I’ve been eating pork all my life. And sometimes it’s hard to break habits.”(43)

DISCUSSION

In our review and content analysis, we found that although most trialed sodium restricted diet interventions claimed success in reducing sodium intake among heart failure patients, few patients in the included studies achieved the recommended sodium intake of 1,500mg/day. Importantly, the sodium restricted diet interventions tested in the included randomized controlled trials focused primarily on changing participants’ sodium intake by increasing their knowledge and skills and self-regulation about sodium. However, our content analysis of the descriptive studies of dietary sodium restriction among heart failure patients elucidated that many more determinants play an important role in the success or failure of a sodium restricted diet. In the descriptive studies, motivational determinants, like outcome expectations, taste preferences, social norms, and self-efficacy, as well as food access and availability, were important in addition to knowledge and skills. However, these were infrequently addressed in the randomized controlled trials. Furthermore, though self-regulation was not prominent in the descriptive studies, it was commonly addressed in the randomized controlled trials. A comparison of the determinants addressed in the sodium restricted diet randomized controlled trials and those identified in the descriptive studies is presented in Figure 3.

Figure 3.

Comparison of intervention targets in experimental and descriptive studies

The mismatch between the evidence from descriptive studies and the intervention design of the randomized controlled trials suggests that providing patients with knowledge, skills, and self-regulation strategies as a primary intervention is unlikely to aid them in adopting a sodium restricted diet.(58, 59) This is consistent with the broader evidence base in nutrition education and behavior change.(19) Interventions intended to change behavior must address motivators, facilitators, as well as structural (i.e., environmental) issues, such as food access.(60, 61) None of the randomized controlled trial interventions in our review addressed environmental/policy determinants, such as food access and food security. Theory-based interventions, which by definition indicate that multiple determinants of behavior change are included, have been identified as holding greater potential for success for behavior changes in heart failure patients.(34, 58) Physiological determinants, such as sodium appetitive drive and/or taste sensitivity, are likely also critical,(16) especially in the context of heart failure.(62, 63) Tastelessness of low sodium food was frequently reported as a barrier to adherence in descriptive studies included in this review. Interventions that use health behavior theory and intervene on structural determinants of diet, including food access and taste preferences, and determinants identified as associated with success, may ultimately be more successful in achieving sodium restriction.

Successful behavioral interventions are needed to assess sodium restriction and this study identifies a mechanism by which sodium restricted diets can be improved. Importantly, these findings suggest that the current debate over whether our current sodium restriction guidelines substantially improve outcomes for patients is affected by type III error, i.e., providing the right answer to the wrong question, because sodium restricted diet randomized controlled trials may not, in fact, represent a comparison of outcomes of participants who do and do not follow a sodium restricted diet.

Limitations

There are several limitations of this review. First, many studies we reviewed lacked rigor in determining sodium intake, relying on self-report and dietary recall. Second, definitions of success varied by study, and intervention design was reported inconsistently, making comparison impossible. Third, the descriptions of the behavioral interventions were often poor, making it difficult to ascertain how determinants were addressed and to what extent, also limiting comparison.

CONCLUSIONS

Restricting dietary sodium to 1500 mg/d is recommended for patients with heart failure, but has not been achieved in randomized controlled trials of sodium restricted diets. Moreover, there is a disconnect between behavioral determinants identified in descriptive studies and those applied in interventional trials. Future randomized controlled trials evaluating sodium restriction should address behavioral determinants beyond knowledge, skills, and self-regulation. The behavioral determinants applied in the interventions must be explicitly reported so they may be compared for effectiveness and reproducibility. Given the acuity of illness in this population, well designed and relatively short-term randomized controlled trials should be able to answer whether and how much sodium restriction would be of benefit to this patient population

ClinicalSignificance_R1.

• Randomized trials of sodium-restricted diets for patients with heart failure have failed to reach the 1500 mg/day recommended intake.

• Sodium restriction interventions in randomized trials do not incorporate many of the determinants shown to be predictive of decreased sodium intake in descriptive studies of sodium restriction among heart failure patients. These include: knowledge/skills, outcome expectations, social/cultural norms, social support, taste, food access, self-efficacy, self-regulation, habit, and mental health.