This study aims to validate the instrument known as the dynamic Neurocognitive Adaptation Scale (dNA), derived from the identical scale previously validated in English on a sample of 815 subjects residing in the United States. All the clinical centers involved will recruit approximately 265 subjects, administer the aforementioned scale, and collect demographic and medical history information. This information will be necessary and sufficient for the first part (Stage #1) of the instrument's validation. Actually, according to the literature, the number of participants required to validate a scale with 20 items- which has already been validated in another language-is approximately 250 participants. Other studies suggest a minimum of 200 subjects to test cross-cultural consistency and reliability. The dNA scale, already validated in English, reports a KM O index greater than .80 (i.e., good), a specific index (Kaiser-Meyer-Olkin) for confirming the appropriateness of the sample. Specifically, for the CFA, this sample size calculation accounts for the need to detect a medium effect size, expressed in terms of Root Mean Square Error of Approximation (RMSEA), with an expected value of 0.05, a statistical power (1-β) of 0.80, and a significance level α of 0.05. Taking potential dropouts into account, approximately 265 subjects will be recruited to ensure the minimum sample size required for statistical power, and this number will be verified experimentally during the validation phase using the KMO index.
Each of the 8 participating centers, depending on their recruitment capacity, will enroll approximately at least 30 participants. The majority of the final sample (i.e., 265 participants) will consist of healthy older adults (HC), while a smaller percentage will include subjects with subjective memory complaints (SMC), mild cognitive impairment (MCI), or Alzheimer's disease (AD).
The first phase (Stage #1) will be followed by a second part (Stage #2) focused on exploring the correlation between the dNA score and neuropsychological variables (particularly those related to memory and executive functions), with the aim of investigating a measure of adaptation that is primarily cognitive in nature. This adaptation will provide a measure of cognitive efficiency, commonly referred to in the literature as cognitive reserve or resilience. In this phase, adherence to a Mediterranean-style diet (i.e., the Mediterranean diet, assessed via a specific questionnaire) will also be explored as a protective factor against general inflammatory processes and cognitive decline associated with Alzheimer's disease (AD).
During the third phase (Stage #3), a measure of neural adaptation will be explored, collecting data on neural efficiency. In the literature, this is a form of resilience or adaptation often described as neural reserve. This component will be explored using structural magnetic resonance imaging (MRI) and functional magnetic resonance imaging (fMRI).
Demographic information, medical history, dietary habits, and neuropsychological data collected by the participating centers will be compiled and managed at the Universities of Chieti and Foggia, under the coordination and responsibility of Professor Michela Balsamo (CH) and Professor Leonardo Carlucci (FG), respectively, both of whom have established expertise in statistics and psychometrics, particularly in the validation of instruments within the field of psychology. The two universities will carry out the data analysis phase aimed at validating the scale (Stage #1), also accounting for key demographic variables (age, sex, education level).
For the primary analyses, principal component analysis (PCA) and exploratory factor analysis (EFA) will be conducted to identify the underlying structure of the dNA instrument and determine the number of components, forming the basis for a subsequent confirmatory factor analysis (CFA) in the validation sample. Descriptive statistics will be used to characterize the study sample. Prior to EFA, sample adequacy (KMO) and variance homogeneity (Bartlett's test) will be assessed, while internal consistency will be evaluated using Cronbach's alpha and inter-item correlations. CFA will be performed using a structural equation modeling (SEM) approach with maximum likelihood estimation to test the hypothesized factor structure. Model fit indices will be calculated, including chi-square (p-value \> 0.05), root mean square error of approximation (RMSEA; cutoff ≤ 0.06), normed fit index (NFI), comparative fit index (CFI; cutoff ≥ 0.90), Tucker-Lewis index (TLI; cut-off ≥ 0.95), and standardized root mean square residual (SRMR; cut-off ≤ 0.08). Finally, the coefficient of determination (R²) will be calculated to assess the proportion of variance explained between factors.
Subsequently, during the second phase (Stage #2), dNA scale scores will be analyzed in relation to the aforementioned dietary habits and neuropsychological variables, particularly memory and executive functions. Correlation analyses will be conducted across the scale domains (cognitive, physical, creative, and social) using distance correlations. The domains will also be examined across the seven time periods considered in the dNA scale (i.e., childhood, adolescence, young adulthood, adulthood, midlife, senior age, and old age) using repeated-measures ANOVA (RM-ANOVA) with Greenhouse-Geisser sphericity corrections. Within-subject effects of time, as well as interactions between time, gender, and education level, will be assessed. Planned contrasts will be conducted within domains for specific pairwise time comparisons and significant interactions, followed by post hoc comparisons using the Bonferroni-Holm correction. Additionally, correlations between the scale and adherence to a Mediterranean diet, as well as performance on cognitive tests assessing memory and executive functions, will be evaluated. Moderation analyses will test whether higher dNA scores may moderate negative effects on neuropsychological outcomes (i.e., lower scores in memory and executive functions) and neurophysiological measures (i.e., poor anticorrelation between the default mode network and task-positive networks such as DAN/FPCN).
All participating clinical centers will collect demographic, medical history, dietary habit, neuropsychological, and, where available, imaging data (MRI, fMRI). These data will be used in the third phase (Stage #3), which will be focused on identifying the neural correlates of adaptation (i.e., neural efficiency) during aging. Imaging data collected by the participating clinical centers will be transferred to the University of Padua, which will oversee data receipt and analysis under the coordination and responsibility of Dr. Marco Marino, whose research focuses on the neural substrates underlying neurocognitive adaptation during aging.
In this phase, both structural (e.g., volumetric and cortical thickness) and functional (e.g., resting-state fMRI) analyses will be performed. Regions of interest (ROIs) will be defined based on prior literature and mapped from standard (MNI) space to individual brain space. Spherical ROIs will be created, and their activity will be summarized using principal component analysis. Functional connectivity maps will be generated by correlating ROI time series with whole-brain activity and then normalized to MNI space for group-level analyses, with statistical significance corrected for multiple comparisons (FDR). Graph theory analysis will be applied to the functional network generated with AAL Atlas (Tzourio-Mazoyer et al., 2002) using the GRETNA software (Wang et al., 2015). Five global indices derived from the graph analysis will be analyzed and correlated with the results of the dNA scale. The effects of diagnosis and functional connectivity on cognitive outcomes (memory scores and other neuropsychological tests) will be examined using ANCOVA and regression models, including interaction effects and group-specific analyses. Additional regression analyses will assess differences in the relationship between cognitive performance and connectivity metrics. Age, handedness, and education will be included as covariates. This phase will receive scientific support from the University of Leuven (Dr. Dante Mantini) and the Cleveland Clinic (Dr. Filippo Cieri); however, these institutions and their respective researchers will not have access to the data.
