TL;DR
TL;DR: Among 10,770 United States adults followed through 31 December 2019, lower hemoglobin-albumin-lymphocyte-platelet (HALP) scores tracked higher all-cause mortality after accounting for continuous metabolic syndrome score (cMetS).
A study published in Frontiers in Nutrition on 20 May 2026 reported that adults in the lowest HALP tertile carried a 15.4% ten-year mortality risk, compared with 11.8% and 12.2% in the middle and highest tertiles, according to Frontiers | HALP, a routine nutrition-inflammation index, and mortality across the cMetS spectrum: NHANES with supportive external cohort evidence.
A routine index tested across metabolic burden
Researchers Keting Deng, Yang Yao, Biping Cheng and Xuan Wang examined whether HALP added prognostic information beyond cMetS, according to Frontiers | HALP, a routine nutrition-inflammation index, and mortality across the cMetS spectrum: NHANES with supportive external cohort evidence.
Deng was affiliated with the Department of Clinical Laboratory at The First Affiliated Hospital of Xi'an Medical University in Xi'an, Shaanxi, China.
Yao was affiliated with the Department of Critical Care Medicine at The First Affiliated Hospital of Xi'an Medical University.
Cheng was affiliated with the Department of Functional Examination at The First Affiliated Hospital of Xi'an Medical University.
Wang, the corresponding author, was affiliated with the Department of Cardiovascular Medicine (VIP Ward) at Yan'an University Xianyang Hospital in Xianyang, Shaanxi, China.
The background premise was that cMetS quantifies cardiometabolic burden but may not fully capture systemic vulnerability related to nutrition, immunity and thrombo-inflammation.
The investigators therefore evaluated HALP, described as a hemoglobin-albumin-lymphocyte-platelet index, across the cMetS spectrum.
HALP was framed as an inexpensive routine index derived from routine laboratory tests.
The authors concluded that HALP was independently and nonlinearly associated with mortality across the cMetS spectrum and provided modest prognostic information beyond cMetS.
They further concluded that HALP may complement metabolic burden assessment by identifying systemic vulnerability.
Who was studied and for how long
The primary analysis included 10,770 participants aged 20 years or older from NHANES 1999–2010.
Participants were followed through 31 December 2019 for mortality outcomes.
The design used survey-weighted Cox regression with restricted cubic splines, adjusting for cMetS and covariates.
Additional analyses included absolute risk estimation, joint HALP-cMetS classification, Harrell's C statistic, decision curve analysis, competing-risk models and supportive external cohort analysis.
Baseline characteristics were presented according to survey-weighted cMetS tertiles in NHANES 1999–2010.
Associations of HALP tertiles with all-cause mortality and cardiovascular mortality in NHANES 1999–2010 were presented in a dedicated results table.
The external cohort component was described as supportive rather than primary, with particular limits for cardiovascular mortality.
What HALP and cMetS represent
HALP integrates indices of oxygen carrying capacity, nutritional and inflammatory status, immune competence and thrombo-inflammatory activity, according to the study background.
Its components are hemoglobin, albumin, lymphocyte and platelet measures obtained from routine laboratory tests.
The report noted that HALP is not yet a widely established clinical risk score, while describing it as inexpensive, routinely available and biologically integrative.
That profile was presented as the rationale for population-based risk assessment research involving HALP.
cMetS was developed to integrate multiple metabolic components into a single quantitative metric.
It captures continuous variation in waist circumference, blood pressure, lipid levels and glucose metabolism.
The study noted that traditional clinical definitions of metabolic syndrome are dichotomous and threshold based.
By contrast, cMetS was described as providing a more refined assessment of metabolic dysregulation and facilitating evaluation of graded associations in population studies.
The authors noted that prognosis varies considerably even among individuals with similar cMetS levels.
That residual heterogeneity was interpreted as indicating that systemic factors beyond conventional metabolic burden contribute to individual risk.
Lower HALP linked to higher all-cause mortality
Among the 10,770 participants, lower HALP was associated with higher all-cause mortality.
Compared with the lowest HALP tertile, the adjusted hazard ratio was 0.78 with a 95% confidence interval of 0.69–0.88 for the middle tertile.
Compared with the lowest HALP tertile, the adjusted hazard ratio was 0.86 with a 95% confidence interval of 0.77–0.95 for the highest tertile.
Those estimates were adjusted for cMetS and covariates in survey-weighted Cox regression.
The association was described as nonlinear and L-shaped in restricted cubic spline analysis.
The pattern indicated elevated risk concentrated at the lower end of the HALP distribution rather than a uniform gradient across all values.
The all-cause mortality association persisted after adjustment for cMetS.
The investigators therefore described HALP as independently associated with mortality across the cMetS spectrum.
Absolute ten-year risk differed by HALP group
Ten-year mortality risk was highest in the lowest HALP tertile at 15.4%.
Ten-year mortality risk was 11.8% in the middle HALP tertile.
Ten-year mortality risk was 12.2% in the highest HALP tertile.
The absolute difference between the lowest and middle groups was 3.6 percentage points.
The absolute difference between the lowest and highest groups was 3.2 percentage points.
Those absolute-risk estimates complemented the relative hazard ratios reported for the middle and highest tertiles versus the lowest tertile.
The absolute-risk analysis was one of the additional evaluations beyond the main Cox models.
It placed the adjusted associations in terms of ten-year mortality risk by HALP tertile.
Combined high metabolic burden and low HALP
Participants with high cMetS and low HALP had the highest risks in joint HALP-cMetS classification.
For all-cause mortality, that combined group had a hazard ratio of 1.32 with a 95% confidence interval of 1.15–1.52.
For cardiovascular mortality, that combined group had a hazard ratio of 1.56 with a 95% confidence interval of 1.16–2.09.
The joint classification directly addressed whether combining HALP with metabolic burden identified higher-risk subgroups.
The highest all-cause risk occurred where elevated continuous metabolic burden coincided with depressed nutrition-inflammation reserve.
The highest cardiovascular risk in the main joint model also occurred in that same high-cMetS and low-HALP stratum.
Those joint estimates were central to the conclusion that HALP may complement cMetS by identifying systemic vulnerability.
How much prediction improved
Adding HALP modestly improved discrimination beyond cMetS and covariates.
Harrell's C statistic moved from 0.8666 to 0.8685 after inclusion of HALP.
The change in discrimination, ΔC, was 0.0018.
Decision curve analysis showed modest net benefit with HALP added.
The authors characterized both the discrimination gain and the net benefit as modest.
That wording indicates incremental rather than transformative prognostic information beyond cMetS and covariates.
The discrimination analysis tested whether HALP provides prognostic value beyond cMetS.
The reported ΔC of 0.0018 quantifies the magnitude of that incremental improvement.
Cardiovascular mortality needed cautious reading
Competing-risk analyses attenuated cardiovascular mortality associations.
That attenuation means the cardiovascular signal was less robust when competing risks were modeled.
A supportive external cohort analysis supported the all-cause mortality association.
The same supportive external cohort analysis was described as limited for cardiovascular mortality.
The overall interpretation therefore rests more firmly on all-cause mortality than on cardiovascular mortality.
The study presentation retained the joint high-cMetS and low-HALP cardiovascular hazard ratio of 1.56 while reporting the competing-risk attenuation.
Readers focused on nutrition and heart health should treat the cardiovascular component as supportive rather than definitive.
Context from a related nutrition-inflammation index
A separate NHANES-based report published 20 May 2026 examined the C-reactive protein–albumin–lymphocyte (CALLY) index and prevalent diabetes, according to Association between the CALLY index as an integrated nutritional biomarker and diabetes prevalence in the NHANES population study.
That cross-sectional analysis included 38,722 participants in NHANES 1999–2010.
Higher CALLY levels were associated with lower diabetes prevalence after adjustment for multiple covariates.
Participants in the highest CALLY quartile had 28% lower odds of prevalent diabetes, with an odds ratio of 0.72 and 95% confidence interval of 0.62–0.84, compared with those in the lowest quartile.
Restricted cubic spline analysis identified an inverse nonlinear relationship with an inflection point at 29.613.
The authors stated that the inflection point describes the observed nonlinear pattern in that dataset but should not be interpreted as a clinically actionable cutoff.
They also stated that the predictive and clinical value of CALLY require prospective validation.
That CALLY report does not establish HALP outcomes, but it documents parallel research interest in composite nutrition-immune-inflammation indicators within the same NHANES 1999–2010 era.
What this means for nutrition monitoring
The HALP analysis did not test a diet, supplement or feeding intervention.
It tested a laboratory composite reflecting hemoglobin, albumin, lymphocyte and platelet information.
The relevant population was general adults aged 20 years or older, not only patients with coronary heart disease or acute myocardial infarction.
Prior studies cited in the report had linked lower HALP to higher all-cause and cardiovascular mortality in coronary heart disease and acute myocardial infarction populations.
The new contribution was evaluation in a broader adult population with continuous quantification of metabolic burden.
For consumers, the practical implication is that routine laboratory measures reflecting nutrition, immunity and thrombo-inflammation may carry prognostic context alongside metabolic measures.
For clinicians and researchers, the implication is that low HALP alongside high cMetS identified the stratum with the highest all-cause and cardiovascular estimates in the main joint model.
Any clinical use would still require prospective validation and establishment of HALP as a clinical risk score, which the authors noted is not yet widely established.