The N utrition I ntervention C ombined E xercise- Support programme: a NICE way to combat frailty in heart failure?

This invited commentary refers to ‘Effects of a multimodal support intervention (the NICE-Support programme) on frailty and quality of life in patients with heart failure: a randomized controlled trial’, by T.-C. Wang et al., https://doi.org/10.1093/eurjcn/zvaf187. In Europe, approximately 2% of the population suffers from heart failure, and the number of affected individuals increases with age.1 Aging2 and heart failure3 are both individually associated with frailty; consequently, the prevalence of frailty in patients with chronic heart failure is substantially greater than in the general elderly population (pooled prevalence of ∼40% in people with heart failure, vs. 10% in the general population of older adults).3 A large meta-analysis has shown that patients with heart failure and frailty have a higher overall mortality risk, greater hospital readmissions, and worse prognosis.4 Frailty and heart failure share several pathophysiological mechanisms like reduced physiological reserve, and both conditions are associated with decreased muscle mass, inflammation, and metabolic dysfunction.5 Multimorbidity, commonly defined as the presence of two or more chronic conditions, increases the likelihood of frailty in heart failure, and advanced heart failure stages (e.g. NYHA III/IV) are associated with a higher risk of frailty.3,5 The current European heart failure guidelines6 recommend individualized interventions that should include cardiac rehabilitation, nutritional interventions, and optimization of guideline-directed medical therapy, irrespective of frailty status. However, frailty is acknowledged and these components would equally attenuate or reverse frailty.6 In a more recent state-of-the-art review of frailty in heart failure, Uchmanowicz et al.7 highlighted the very high prevalence in HFpEF (up to 90% in HFpEF), prognostic impact on mortality and hospitalizations (regardless of phenotype), and the potential for reversibility through interventions like rehabilitation and optimization of guideline-directed medical therapy. They subsequently recommended employing multidimensional tools to support routine screening at diagnosis and follow-up, alongside tailored interventions. Given the increasing call for screening and interventions attuned to frailty in the heart failure setting, it is perhaps unsurprising that nurse-led interventions are being evaluated for their capacity to affect a change in this key prognostic outcome. The NICE-Support intervention,8 visually presented in Figure 1, is one such intervention that has recently featured in the European Journal of Cardiovascular Nursing. In brief, it included nurse-led counselling (optimization of symptoms, medication, comorbidities, psychological status), provision of self-care advice, resources, a fitness tracker, and personalized diet and exercise plans that were tailored over time, partnered with continuous education and support via print and digital resources. Over a 24-week period, the intervention was shown to significantly reduce frailty status in the intervention arm (baseline and 24 weeks: 24.29 ± 11.35, 4.08 ± 6.40) vs. control [baseline and 24 weeks: 24.05 ± 11.58, 15.94 ± 12.85 (all mean and SD)]. Similarly impressive effects were observed in quality-of-life scores in the intervention arm (baseline and 24 weeks: 45.87 ± 22.90, 6.18 ± 11.41) vs. control [baseline and 24 weeks 42.89 ± 22.31, 20.81 ± 19.81 (all mean and SD)]. Visualization of the NICE-Support intervention. Experienced heart failure practitioners might note the similarity between the NICE-Support intervention and well-established heart failure disease management programmes (DMPs) which typically include medical therapy optimization, education, nutrition, and exercise counselling.6 Thus, these programmes offer a valid source for comparison. Multi-component nurse-led DMPs have a long history—in the early 2000s, Strömberg et al.9 demonstrated the positive effects of the Swedish model’ on morbidity, mortality, and self-care in heart failure. Similar nurse-led programmes in the UK, the Netherlands, and Germany have confirmed the effectiveness of these programmes.10 A plethora of trails since these early signals ensued, some of which indicated small or no benefit.11 However, the most recent meta-analysis suggests these class of interventions do have a positive effect on key outcomes, including quality of life.12 What is less frequently measured or described as part of these programmes, however, is frailty. We have collectively mused on the findings and discussed how the intervention managed to achieve such large effects. A mean reduction of almost 40 points on the Minnesota Living with Heart Failure Questionnaire is considerable. To the unfamiliar reader, this disease-specific measure of quality of life is scored out of 100, with lower scores indicating better quality of life; a minimal clinically important difference of 3–5 points has been established.13 It is possible that the rapid sequencing of the intervention, driven by a robust protocol and comprehensive staff training plan, ensured that the intervention was delivered quickly and systematically. Further, delivering the intervention in the optimal ‘teachable window’, i.e. at a time when people with heart failure might be more receptive to behaviour change,14 and regular reinforcement of information and advice, may well have resulted in more substantial effects. We know that referral to cardiac rehabilitation in heart failure, where the majority of patients would receive similar support as was provided by the NICE-Support intervention, is suboptimal in heart failure.15 This study may be considered evidence that early and sustained input, that is unaffected by the referral delays, waitlists, restrictive criteria and scheduling barriers often associated with routine care,16 can produce optimal outcomes. Improvements in frailty were also large, with a mean change of 20.41 in the intervention arm, comfortably bringing participants into the non-frail category of the Frailty Assessment Scale for Heart Failure.17 No minimal clinically important difference has been established for this scale; however, the change in terms of the number is large and presumably the end mean result of 4.08 is ostensibly normal (a score of 0 is classed as ‘robust’ by the scale creators). The control group, on the other hand, remained closer to the frail threshold of 20—whilst not established by the authors, this is arguably pre-frail. Of note, application of different frailty scales yields different prevalence rate and risk profiles,3,18 and there is no universal standard.18 Combining diet with exercise has been shown to be additive in terms of reversing frailty in older adults with heart failure19; however, dietary change within the context of an intervention has been shown to be complex,20 making them challenging to achieve and sustain.21 It is feasible that personalization and ongoing nurse counselling, combined with high adherence to the exercise prescription, could have optimized chances of effecting such a large change in frailty status. Limited reporting across all outcomes, perhaps driven by journal word count and figure and table restrictions, prevents reflection on what might be driving the change in frailty and quality-of-life scores. It could be, as the authors assert,8 the multi-component and theory-grounded nature of the intervention, which was designed to target key variables that contribute to frailty and quality of life. However, the benefits observed might not be so large, if an alternative measure was used, especially one with a longer history of use and validation. Furthermore, the effect may be wholly driven by the exercise or diet components alone, as meta-analyses have demonstrated that targeting either can lead to improvement in these outcomes,22,23 albeit it to a lesser extent than is observed here. Providing additional data via supplementary material would allow researchers to interrogate and contextualize the findings, gaining a deeper understanding of the intervention and its effects. The TIDieR (template for intervention description and replication) has been used24; however, granular descriptions or case examples would be very helpful for future researchers, who might aspire to combine results in a meta-analysis of effects, replicate, or adapt this intervention. An accompanying process evaluation might help unpack what is driving benefit.25 Lastly, an attention control group, matched for the number of visits, would confirm more robustly the effects are driven by the intervention components alone, and not the result of increased care and attention that naturally accompanies more intensive programmes like the NICE-Support intervention. There is a growing body of evidence indicating the significance of frailty within heart failure. In our discussion of this work, we pondered, perhaps contestably, whether this intervention was conceptually different from trials of DMPs already published and routine care in many clinical settings. Given this, it might be useful to re-analyse previous nurse-led DMPs tested via randomized controlled trial methods (RCTs), or undertake service evaluations of extant DMP-based, nurse-led care, to determine their effects on frailty. Some frailty scores can be generated via artificial intelligence-based mining of structured data, like that collected in RCTs and electronic medical records.26 The latter would generate useful comparative data and help determine the unique factors that made NICE-Support so effective in enhancing frailty and quality of life. In the same vein, complementary analyses interrogating internal and external influencing variables would contribute significantly to the evidence base. In 2012, Clark and Thompson called for a DMP paradigm shift in the conceptualization, implementation, testing, and post hoc interrogation of these programmes.11 Augmenting the NICE-Support data with further analyses, as suggested here, would be the first step in realizing this paradigmatic shift. Faye Forsyth (Conceptualization [lead], Supervision [lead], Writing—original draft [equal], Writing—review & editing [equal]) and Anke Kampmann (Conceptualization [supporting], Writing—original draft [equal], Writing—review & editing [equal]) No funding was obtained. This commentary does not contain new data.