Altered RBC deformability in diabetes: clinical characteristics and RBC pathophysiology

Abstract

Background: Reduced red blood cell deformability (RBCD) is associated with diabetic vascular complications, but early pathophysiological RBC changes and predictive demographic and clinical factors in populations with diabetes are unclear. An understanding of early diabetes-specific RBC changes associated with impaired RBCD is essential in investigating mechanisms that predispose to diabetic vascular complications.

Methods: We conducted an outpatient cross-sectional study of participants in a well-controlled diabetes cohort (N81) and nondiabetic controls (N78) at the National Institutes of Health. First, between-group differences in RBCD measures were assessed with shear stress-gradient ektacytometry. Differences in structural RBC parameters were assessed using osmotic gradient ektacytometry and NaCl osmotic fragility. Functional RBC changes were assessed using hemoglobin-oxygen dissociation: p50.

Results: All shear-stress gradient RBCD measures were significantly altered in the diabetes cohort vs. nondiabetic controls, even after adjustment for confounding covariates (p < 0.001). Adjusted for diabetes-status and demographic factors, significant predictors of reduced RBCD included older age, Black race, male gender, hyperglycemia, and vascular complications (all p < 0.05). Reduced RBCD was also associated with aberrant osmotic-gradient parameters, with a left-shift on osmotic gradient profile indicative of dehydrated RBCs in diabetes. A structure-function relationship was observed with reduced RBCD associated with reduced osmotic fragility (P < 0.001) and increased hemoglobin-oxygen dissociation (P < 0.01).

Conclusions: Findings suggest impaired RBCD incurs similar demographic and clinical risk factors as diabetic vascular disease, with early pathophysiological RBC changes indicative of disordered RBC hydration in diabetes. Findings provide strong evidence for disordered oxygen release as a functional consequence of reduced RBCD.

Clinical trial number: NCT00071526.

Fig. 1

A Enrollment and patient flow of study participants in the diabetes cohort and nondiabetic control group, B Flow diagram for RBC physiology studies conducted in samples obtained from participants both groups. ^a Shear stress-gradient ektacytometry. ^b Osmotic-gradient ektacytometry. ^c Sodium chloride (NaCl) technique for osmotic fragility. ^d Hemoglobin-oxygen dissociation rate assessed by partial pressure at 50% oxygen saturation (p50)

Fig. 2

Shear stress gradient RBC deformability (RBCD) measures in the diabetes cohort (N = 81) and nondiabetic controls (N = 78). A RBC deformability curves of mean elongation indices across a gradient of shear stresses (0.3–30 Pa) in the diabetes cohort and nondiabetic controls. The difference of the mean values in both groups is shown in the insert; p-values reflect joint Wald tests for diabetes-diagnosis effects of repeated measures modeling across shear-stress levels using transformed values. B Mean elongation index at each individual shear stress value in both groups. Squares indicate Hodges-Lehmann shifts, with horizontal points indicating 95% Confidence Interval (CI). C Derived RBCD shear stress-gradient measures: maximal elongation index: EImax^a, and shear stress required to reach 50% of EImax: SS_1/2^b. Squares indicate Hodges-Lehmann shifts, with horizontal points indicating 95% CI. *Adjusted for potentially confounding covariates (Supplementary Fig. 1). ^a Maximum elongation index (EImax). ^b Shear stress value required to reach half-maximal elongation index (SS_1/2)

Fig. 3

Demographic and clinical variables and their predictive association with RBCD assessed using diabetes-status-adjusted linear regression of transformed values for SS_1/2. Squares indicate Box-Cox transformed mean shifts in SS_1/2, conditional on diabetes status, with horizontal points indicating 95% CI. ¹Black and White groups were comprised of non-Hispanic participants. “Other” was comprised of Asians, Hispanic/Latino ethnicity, and multiracial participants. ²Presence of micro- or macrovascular complications. ³Type 1 or Type 2 diabetes. BMI, body mass index; eGFR, estimated glomerular filtration rate; PCR, Protein-to-creatinine ratio

Fig. 4

Relationships between RBC deformability (SS_1/2) and osmotic-gradient osmolality measures (Omax, Omin, Ohyper) in the diabetes cohort (N = 49) and nondiabetic controls (N = 40). A Representative scheme of osmotic gradient curves and parameters. (B-C) Mean osmotic elongation indices (B), and corresponding mean osmolality values (C) in both groups. Squares indicate Hodges-Lehmann shifts, with horizontal points indicating 95% CI. (D-F) Associations between RBCD (SS_1/2) and osmotic-gradient osmolality measures: Omin (D), Omax (E) and Ohyper (F) within the diabetes cohort (red circles) and within nondiabetic controls (blue circles) ¹Maximum osmotic elongation index (O.EImax) and corresponding osmolality (Omax) ²Minimum osmotic elongation index (O.EImin) and corresponding osmolality (Omin) ³Hypertonic osmotic elongation index (O.EIhyper) and corresponding osmolality (Ohyper)

Fig. 5

RBC Deformability (SS_1/2) and Osmotic fragility (Omin) in the diabetes cohort (N = 48) and nondiabetic controls (N = 77). A Associations between Omin and osmotic fragility measured using the classic NaCl technique (See Methods) in the diabetes cohort (red circles) and nondiabetic controls (blue circles); p-values reflect joint Wald tests for diabetes-diagnosis effects of repeated measures modeling across SS_1/2 tertiles using transformed values. B Associations between SS_1/2 and NaCl osmotic fragility the diabetes cohort (red circles) and nondiabetic controls (blue circles). (C-D) Mean Omin values stratified by SS_1/2 tertiles within the diabetes cohort (C) and within nondiabetic controls (D). E Relationships between RBCD (SS_1/2) and mean osmotic fragility (Omin) in subgroups stratified by SS_1/2 tertiles within the diabetes cohort (red circles) and nondiabetic controls (blue circles); 95% confidence intervals result from repeated measures modeling across SS_1/2 tertiles using transformed values. SS_1/2 tertiles within the diabetes cohort: low: 0.75–1.37 Pa; medium:1.38–1.75; high:1.76–2.9 Pa, and nondiabetic controls: low:0.85–1.28 Pa; medium:1.29–1.41 Pa; high: 1.43–2.47 Pa

Fig. 6

RBC Deformability (SS_1/2) and hemoglobin-oxygen dissociation rate (p50) in the diabetes cohort (N = 58) and nondiabetic controls (N = 75). A Associations between RBCD (SS_1/2) and hemoglobin-oxygen dissociation rate (p50) within diabetes cohort (red circles) and nondiabetic controls (blue circles); p-values reflect joint Wald tests for diabetes-diagnosis effects of repeated measures modeling across SS_1/2 tertiles using transformed values. B, C Mean p50 values in subgroups stratified by SS_1/2 tertiles within the diabetes cohort (B) and nondiabetic controls (C). D Relationships between RBCD (SS_1/2) and mean p50 values in subgroups stratified by SS_1/2 tertiles within the diabetes cohort (red circles) and nondiabetic controls (blue circles); 95% confidence intervals result from repeated measures modeling across SS_1/2 tertiles using transformed values. SS_1/2 tertiles within the diabetes cohort: low: 0.75–1.37 Pa; medium:1.38–1.75; high:1.76–2.9 Pa, and nondiabetic controls: low:0.85–1.28 Pa; medium:1.29–1.41 Pa; high: 1.43–2.47 Pa

Fig. 7

Scheme of osmotic-gradient profile changes in diabetes