Psychometric evaluation and refinement of the 7DHW questionnaire for the German population

Holistic conceptualizations of health emphasize the integration of physical, psychological, social, occupational, and environmental dimensions. However, psychometrically robust instruments that operationalize such multidimensional frameworks remain limited. This study aimed to examine the internal structure, reliability, and construct validity of the Holistic Scale of Body and Mental Health and Wellbeing, based on the 7 Dimensions of Holistic Wellbeing (7DHW) model, in a German adult sample.

A cross-sectional study was conducted with 677 adults in Germany (age range 22-86 years). Participants completed the 7DHW questionnaire online. Psychometric evaluation included exploratory and confirmatory factor analyses (EFA and CFA) to assess dimensionality, internal consistency analyses using Cronbach's α and McDonald's ω, and known-groups comparisons across gender and age. Item-level diagnostics were conducted to evaluate scoring and conceptual coherence, particularly for poorly performing sub-scales.

The findings provided partial support for the conceptual framework while revealing substantial multidimensionality within several domains. Dimensions A (Self-Esteem), C (Social Relationships), D (Environment), and G (Sense of the Future) demonstrated strong internal consistency, and Dimension E (Meaningful Work) showed acceptable reliability. The original Body Image (B) and Health Knowledge and Behavior (F) dimensions showed inadequate psychometric performance. Following item-level diagnostics, Dimension B was refined through reverse coding and item reduction, resulting in acceptable reliability, although its factorial structure remains provisional. Dimension F demonstrated a heterogeneous structure and was better represented as a set of exploratory sub-domains (e.g., preventive behavior, dietary behavior), with variable reliability across components. Known-groups analyses revealed expected differences by gender and age across selected dimensions and sub-domains, supporting construct validity.

The results indicate that several domains are inherently multidimensional and may be better conceptualized as clusters of related subcomponents rather than strictly unidimensional scales. Overall, the instrument demonstrates promising psychometric properties at the domain level, while highlighting the need for further refinement and independent validation of specific dimensions.