Prognostic Value of Functional-Nutritional Status in Older Adults With Acute Coronary Syndrome

This is a prospective, two-center observational cohort study of the prognostic value of functional-nutritional status in patients aged 65 years and older hospitalized with acute coronary syndrome (ACS: ST-elevation myocardial infarction, non-ST-elevation myocardial infarction or unstable angina). It is not a randomized or interventional trial; no treatment is assigned by the study. The target enrollment is 270 participants; recruitment continues until 31 December 2027 or until 330 participants are enrolled, whichever comes first.

Baseline (t0) is the date of written informed consent at the first research contact after clinical stabilization, no later than 7 days after admission and before discharge. Baseline assessments are performed at this contact. Patients who die or are discharged before this contact are recorded in the screening log and are not enrolled. Participants in whom non-fulfilment of the eligibility criteria already present at t0 (including revision of the ACS diagnosis by the final diagnosis of the index hospitalization) is confirmed by a second specialist without access to functional-nutritional data are excluded from all analyses and are not followed further; their number and reasons are reported. Withdrawal of consent is recorded at three levels: refusal of contacts only (follow-up continues from medical records and regional health information systems); withdrawal from participation (data collected up to that point are used and follow-up is censored at the withdrawal date); and a request to destroy data (the participant is not included in any analysis). Loss of telephone contact is not withdrawal and is not a reason for exclusion.

Exposure. The principal exposure is a prespecified functional-nutritional burden: the number of impaired domains (0, 1 or 2), calculated as functional impairment plus GLIM malnutrition.

• Functional domain (functional impairment): probable pre-frailty or frailty in the state about two weeks before ACS, defined as an "Age Is Not a Barrier" screening questionnaire score of 3 or more, or a Clinical Frailty Scale (version 2.0) score of 4 or more. These are operational categories; a clinical diagnosis of frailty, which requires comprehensive geriatric assessment, is not established.
• Nutritional domain: malnutrition (moderate or severe) according to the Global Leadership Initiative on Malnutrition (GLIM) criteria (2019 consensus and 2025 update). All five criteria are assessed in every participant, independently of screening (research classification). Phenotypic criteria: unintentional weight loss; low body mass index; reduced muscle mass by calf circumference adjusted for body mass index. Etiologic criteria: reduced food intake or assimilation; inflammation or disease burden, assessed with a standardized documentation checklist, in which a negative answer requires a documented negative assessment or a structured review. Research C-reactive protein is not used for this criterion.
• Malnutrition is present when at least one phenotypic and at least one etiologic criterion are met; absent only when all five criteria have been assessed and this condition is not met; and unknown otherwise. A criterion that could not be assessed is never treated as not met, and unknown values are not treated as negative.

Supporting constructs. Risk of malnutrition is screened with the Mini Nutritional Assessment - Short Form (MNA-SF score of 11 or less). It is a separate construct, is not part of the burden, and is used in a supporting sensitivity analysis. The standard two-step GLIM procedure (MNA-SF screening followed by the GLIM criteria) is a supporting classification that differs from the research classification; discrepancies are reported. Four phenotypes are described: neither impairment, functional only, nutritional only, and both. The Short Physical Performance Battery is performed when possible before discharge and describes in-hospital physical performance; it is not part of the burden.

Laboratory component. Research blood sampling is part of the protocol and is offered to every participant. With separate consent, two additional serum tubes without additive (5.5 mL each, 11 mL in total) are collected after consent at the nearest routine blood draw ordered by the attending physician, through the same venipuncture. Serum is processed and stored according to the rules of the institutional laboratory, and albumin and high-sensitivity C-reactive protein are measured later in batches. The Geriatric Nutritional Risk Index, Controlling Nutritional Status score and Prognostic Nutritional Index are calculated using lymphocyte count and total cholesterol from the first routine tests performed no later than the research sample. Laboratory analyses start at the sampling date (delayed entry), are exploratory, and do not redefine the baseline exposure. Declining the blood sample does not prevent participation.

Outcomes and analysis.

• Primary outcome: time to the first of all-cause death, recurrent nonfatal myocardial infarction, ischemic stroke, or unplanned cardiovascular rehospitalization within 365 days after t0.
• Primary model: Cox regression with the burden as a linear term (1 degree of freedom), adjusted for age, sex, Charlson comorbidity index and ACS type, and stratified by center; the primary hypothesis is tested by a single likelihood-ratio test. Supporting analyses include adjustment for GRACE 2.0 risk and the contribution of the individual domains. Fourteen prespecified sensitivity analyses are planned, including adjustment for Killip class; the nutritional domain defined by MNA-SF risk, by the two-step GLIM procedure, and by GLIM without automatic fulfilment of the inflammation criterion in index myocardial infarction; a time scale from admission with delayed entry at t0; center as a fixed effect; complete-case analysis and delta-adjustment of unassessed phenotypic criteria.
• Secondary outcomes include all-cause death, death or myocardial infarction, and vital and functional status at the follow-up assessment on days 366-395; these three secondary hypotheses are tested with Holm correction (the model for vital and functional status at follow-up is adjusted for age). Functional deterioration is a decline in at least one of four Barthel index items compared with the state before ACS. For outcomes that do not include death, death is treated as a competing event (Aalen-Johansen cumulative incidence); no separate Cox models or p-values are used for non-fatal components.
• Missing data are handled by multiple imputation of the component measurements; GLIM classification and burden are derived in each imputed dataset. Estimates are combined by Rubin's rules and likelihood-ratio tests by the D2 method.
• Before the final analysis, only descriptive analyses of baseline and in-hospital data are performed, and a descriptive analysis of the 181-day interim endpoint after the whole cohort has completed 181 days of follow-up. Interim analyses have no stopping rules and do not test the primary hypothesis; no alpha adjustment is applied.
• Follow-up contacts take place on days 182-211 and 366-395 after t0. All treatment decisions follow routine clinical practice.

The statistical analysis plan was registered at OSF Registries before enrollment (DOI 10.17605/OSF.IO/D7W25), with prespecified amendments.