Dengue fever is a mosquito-borne viral infection that produces characteristic abnormalities in routine blood tests, yet these hematologic changes are typically analysed separately for each parameter rather than as a combined multivariate profile. This study investigated whether the joint hematologic profile of adult dengue patients in Bangladesh is systematically displaced from healthy adult reference values. We analysed a cohort of laboratory-confirmed adult dengue cases from a Bangladeshi hospital and focused on four core hematologic indices: haemoglobin, white blood cell count, platelet count, and platelet distribution width (PDW). External adult reference means were used to define a healthy location vector, and robust multivariate inference was carried out using the rank-based location test of Utts and Hettmansperger (1980). Sex-specific (male, female) and pooled (all adults) analyses were performed after careful data cleaning, outlier diagnostics, and checks of non-normality. Across all sex-specific and pooled analyses, the same multivariate profile emerged: haemoglobin, white-cell, and platelet levels were consistently lower than their healthy reference means, whereas PDW was higher, indicating greater platelet-size variability. The Utts–Hettmansperger test strongly rejected the null hypothesis of equality with the healthy reference vector in every analysis, documenting a large and coherent displacement of the dengue group in the four-dimensional hematologic space. Taken together, these results provide robust, distribution-free statistical evidence that adult dengue fever in Bangladesh is associated with a stable, biologically interpretable shift in core blood indices, integrating leukopenia, thrombocytopenia, and altered platelet morphology into a single multivariate summary. This study demonstrates that robust rank-based multivariate location tests can enhance traditional laboratory interpretation by quantifying the joint displacement of key blood indices in infectious-disease cohorts such as adult dengue.
