Skip to main content
NCBI home page
The Journal of International Medical Research logoLink to The Journal of International Medical Research
. 2026 Aug 19;54(8):03000605261474625. doi: 10.1177/03000605261474625

Retrospective cross-sectional associations of whole blood cell count and hemoglobin level with lumbar bone mineral density in older adults

Fengxian Cao 1,*, Zhengfu Wang 1,*, Fengxi Chen 1,✉, Qiuyan Cao 2, Qingyan Tang 1, Hongzheng He 1, Lei Zhang 3
PMCID: PMC13490813  PMID: 42615610

Abstract

Introduction

The increasing prevalence of osteoporosis has exerted a substantial global economic burden. Low hemoglobin levels and low bone density frequently co-occur in older adults. This study aimed to elucidate the relationship between whole blood cell count and bone density among individuals aged ≥50 years.

Methods

A retrospective cross-sectional study was conducted from January 2024 to March 2024, which enrolled 83 participants aged ≥50 years. Demographic and clinical characteristics were recorded. Comprehensive complete blood cell count tests were performed, and lumbar bone mineral density (L1–L4) was measured using dual-energy X-ray absorptiometry. Linear regression and restricted cubic spline analyses were performed.

Results

Red blood cell count and hemoglobin levels were positively correlated with lumbar bone mineral density. This association remained significant in males but was not observed in females. Nonlinear relationships were observed between hemoglobin levels, platelet levels, and lumbar bone mineral density, with platelet levels exhibiting an inverted U-shaped curve.

Conclusion

In this retrospective cross-sectional study involving older Chinese individuals, hemoglobin levels demonstrated a positive, nonlinear association with lumbar bone mineral density, particularly among males. Although monitoring of hemoglobin levels may serve as a useful adjunct for early risk assessment of lumbar bone metabolic disorders, these findings are preliminary and causal inferences cannot be drawn due to the cross-sectional design. Further large-scale longitudinal studies are warranted to validate these results.

Keywords: Osteoporosis, bone mineral density, hemoglobin, complete blood count, lumbar spine

Introduction

Osteoporosis (OP) is a skeletal disorder characterized by deteriorated bone microstructure and loss of bone mass, which significantly increases the risk of fracture. 1 Bone mineral density (BMD), regarded as the gold standard for diagnosing OP, is measured using dual-energy X-ray absorptiometry (DXA). It serves as an approximation of the amount of bone mass present in the body. 2 Owing to population aging, OP is emerging as a global epidemic. Currently, it is estimated that >200 million individuals have OP. According to the statistics released by the International Osteoporosis Foundation (IOF), globally, one in three women and one in five men aged ≥50 years will experience osteoporotic fractures during their lifetimes. 3 The consistent rise in OP prevalence of OP imposes a substantial economic burden on countries around the world.4,5 Although the relationship between complete blood count (CBC) parameters and bone health has been explored in various populations, most of the existing research has been conducted in European and Korean cohorts.6–10 There is a notable scarcity of studies focusing on older populations in the Yangtze River Delta region of China, an area with distinct economic and natural environments that may influence the association between hematological parameters and bone metabolism. 11 These geographical and population-specific gaps in the literature underscore the need for regionally focused investigations to better understand these relationships in the context of China's rapidly aging society. 12

A CBC is a routine blood test, which measures the white blood cell (WBC) count, red blood cell (RBC) count, hemoglobin (Hb) level, and platelet count (PLT). 13 This test helps determine an individual's blood health status during disease assessment and diagnosis. 14 Given that CBC is a routinely performed, inexpensive test, investigating its association with BMD could offer a practical and accessible approach for early identification of bone health deterioration in aging populations. Anemia, which is characterized by a low Hb level, is a prevalent public health issue in low-, middle-, and high-income countries across the globe. 15 Among middle-aged and older individuals, low Hb levels have been recognized as a risk factor for OP, falls, fractures, and physical deterioration.16,17 Several studies have indicated that low Hb levels exhibit a negative, independent correlation with BMD.6,7 A study exploring the correlation between adult Hb levels and BMD has demonstrated a nonlinear relationship between adult Hb levels and lumbar BMD. 18 A cross-sectional study involving individuals aged 55–85 years has demonstrated that Hb levels exhibit a positive correlation with BMD in the first to fourth lumbar vertebrae, femoral neck, and total hip. 19 Another cross-sectional study, which involved 495 patients with type 2 diabetes mellitus (T2DM), has indicated that Hb levels are related to OP in male patients, especially those aged >50 years. 20 Additionally, a retrospective study has revealed that Hb levels are inversely related to BMD in postmenopausal women. Moreover, patients with OP have higher complete blood cell counts and Hb levels than those without OP (NOP). 21 However, another study has demonstrated a lack of correlation between Hb levels and BMD. 9 Owing to the constraints associated with the selection of the study population and sample size, the findings regarding the relationship between Hb levels and BMD have been inconsistent. Moreover, only a few studies have delved deeply into the association between Hb levels and lumbar BMD.

Therefore, we conducted a retrospective cross-sectional study to further validate the relationship between CBC parameters (including WBC count, RBC count, PLT, and Hb levels) and lumbar BMD in an older population in China.

Materials and methods

Study population

This retrospective cross-sectional study analyzed the data of middle-aged and older individuals who completed a 1-min OP risk self-assessment (Supplementary Figure 1) at Jinxi People's Hospital, Kunshan City from January 2024 to March 2024. After excluding 11 individuals who did not provide consent for the use of their anonymized data, 83 participants were ultimately included in the analysis. The specific inclusion criteria were as follows: (a) patient age >50 years; (b) availability of complete BMD and blood indicator data; and (c) no current or prior use of any medications known to affect complete blood cell counts or BMD measurements. Figure 1 illustrates the selection process of this study. The study protocol was approved by the Ethics Committee of Kunshan Jinxi People's Hospital (Approval Number: LCYJIEC2025-001-01). The requirement for written informed consent was waived due to the retrospective nature of the study, which used fully anonymized data in accordance with national regulations and ethical guidelines.

Figure 1.

Figure 1.

Open in a new tab

Flow chart of participants selection.

Collection of body mass index (BMI) data

Trained physicians employed a 1-min self-assessment tool for OP risk evaluation within the Geriatric Assessment Management System to identify participants at OP risk and exclude factors that could influence the study results. A single technician measured the height and weight of all the participants. During height measurement, participants were required to remove their shoes and stand on the height-measuring device with their head, hips, and ankles in contact with the three designated points on the device. The height was recorded at the intersection of the highest point of the device and the vertical line of the cylinder and is expressed in meters with two decimal places. For the weight measurement, participants were asked to wear light clothing and remove their shoes before standing at the base of the weight-measuring device. They were required to maintain a proper posture, with the body in an upright position. The reading of the pointer on the weight-measuring device was then recorded in kilograms. BMI was calculated by dividing the weight by the square of the height (kg/m2).

BMD measurement using DXA

DXA is the most widely recognized method for measuring BMD owing to its rapid measurement process, user-friendly operation, and low level of radiation exposure. BMD measurements of the lumbar spine (L1–L4) were performed for all participants using DXA. The measurement was performed using a fourth-generation pen-tip wire bundle DXA (General Electric Company (GE), USA). To ensure the validity of comparisons, all patients were examined by the same experienced technician using the same densitometer.

Blood parameters

CBC values and Hb level were determined using the Mindray BC-7500CRP fully automatic hematology analyzer. CBC measurements included PLT; WBC, neutrophil (NEU), lymphocyte (LYMP), monocyte (MONO), eosinophil (EO), basophil (BO), and RBC counts; Hb, hematocrit (HCT), plateletcrit (PCT), and mean corpuscular hemoglobin (MCH) levels; mean corpuscular volume (MCV); mean corpuscular hemoglobin concentration (MCHC), red blood cell distribution width (RDW); platelet distribution width (PDW); and mean platelet volume (MPV).

To reduce interference from other factors, fasting venous blood was collected from the participants after they had fasted for at least 12 h. We drew 5 mL of blood samples from the forearm into a K2-ethylenediaminetetraacetic acid (K2-EDTA) anticoagulant tubes. Immediately after the blood sample was drawn, the tube was inverted and mixed gently 8 times, and the sample was analyzed within 24 h.

Statistical analyses

Continuous variables were presented as median (interquartile range (IQR)) or mean (SD) values. With respect to sex, the demographic characteristics and blood test indices of all study participants were analyzed using independent samples t-tests and Kruskal–Wallis tests. A univariate linear regression model was employed to analyze the association of lumbar BMD (L1, L2, L3, L4, and L1–L4) with demographic and clinical characteristics. Sex, age, and BMI were the main risk factors for OP.22,23 Linear regression analysis adjusted for age and BMI was conducted through sex stratification to identify independent factors related to CBC parameters (WBC count, RBC count, PLT, and Hb level) and BMD of the lumbar spine (L1, L2, L3, L4, L1–L4). Values with p <0.05 were regarded as statistically significant. Finally, restricted cubic spline (RCS) analysis was applied to further explore the relationship between CBC parameters (WBC count, RBC count, PLT, and Hb level) and lumbar spine BMD (L1, L2, L3, L4, and L1–L4). RCS models were fitted using the ‘rms’ package in R, with 4 knots placed at the 5th, 35th, 65th, and 95th percentiles of the predictor distributions. This knot placement was chosen to provide sufficient flexibility to capture potential nonlinear patterns and minimize the influence of extreme values, in line with standard recommendations for RCSs. The nonlinearity was tested using analysis of variance (ANOVA). Statistical analyses were carried out using the Statistical Package for the Social Sciences (SPSS) software (version 27.0) and RStudio version 2.41.

Results

Clinical characteristics

Ultimately, 83 individuals aged ≥50 years were included in the final analysis. Table 1 shows the characteristics of older individuals stratified by sex (18 males and 65 females). Height (1.68 (1.60, 1.71) vs. 1.54 (1.50, 1.57) m, p < 0.001) and Hb levels (133.00 (21.00) vs. 124.00 (14.00) g/L, p = 0.031) were significantly higher among males than among females; additionally, L1, L2, L3, L4, and L1–L4 BMD levels were significantly higher among males than among females (p < 0.05) (Table 1).

Table 1.

Population characteristics and laboratory measurements.

Variable All patients Sex p-value
Male (N = 18) Female (N = 65)
L1 BMD (g/cm2, Mean (SD)) 0.91 (0.02) 1.08 (0.26) 0.86 (0.19) <0.001
L2 BMD (g/cm2, Mean (SD)) 0.95 (0.03) 1.12 (0.26) 0.91 (0.20) <0.001
L3 BMD (g/cm2, Mean (SD)) 0.99 (0.03) 1.15 (0.30) 0.94 (0.19) 0.01
L4 BMD (g/cm2, Mean (SD)) 1.02 (0.03) 1.19 (0.29) 0.97 (0.20) 0.007
L1–L4 BMD (g/cm2, Mean (SD)) 0.97 (0.02) 1.14 (0.27) 0.92 (0.19) <0.001
Age (years, M (P25, P75)) 75.10 (66.00, 81.50) 76.25 (72.38, 86.20) 75.00 (63.75, 80.80) 0.194
Height (m, M (P25, P75)) 1.55 (1.50, 1.60) 1.68 (1.60, 1.71) 1.54 (1.50, 1.57) <0.001
Weight (kg, Mean (SD)) 56.18 (10.27) 61.60 (14.60) 54.70 (8.30) 0.07
BMI (kg/m2, Mean (SD)) 23.02 (3.42) 22.28 (4.27) 23.22 (3.16) 0.303
WBC count (×10^9/L, M (P25, P75)) 6.60 (5.27, 8.52) 6.84 (5.83, 9.15) 6.60 (5.14, 8.25) 0.395
NEU count (%, Mean (SD))) 71.09 (13.02) 71.30 (14.80) 71.00 (12.60) 0.938
LYMP count (%, Mean (SD)) 21.04 (11.11) 18.80 (11.30) 21.70 (11.00) 0.326
MONO (%, M (P25, P75)) 5.80 (4.70, 7.40) 6.45 (5.28, 8.98) 5.70 (4.55, 7.35) 0.12
EO (%, M (P25, P75)) 0.70 (0.10, 2.20) 0.90 (0.10, 4.05) 0.70 (0.10, 1.80) 0.228
BO (%, M (P25, P75)) 0.30 (0.20, 0.50) 0.30 (0.10, 0.43) 0.30 (0.20, 0.50) 0.253
NEU (×10^9/L, M (P25, P75)) 4.40 (3.30, 6.90) 4.60 (3.75, 7.33) 4.40 (3.25, 6.35) 0.5
LYMP (×10^9/L, M (P25, P75)) 1.20 (0.90, 1.70) 1.15 (0.78, 1.60) 1.30 (0.90, 1.85) 0.486
MONO (×10^9/L, M (P25, P75)) 0.40 (0.20, 0.60) 0.50 (0.30, 0.73) 0.40 (0.20, 0.50) 0.05
EO (×10^9/L, M (P25, P75)) 0.04 (0.01, 0.12) 0.07 (0.01, 0.24) 0.04 (0.01, 0.09) 0.286
BO (×10^9/L, M (P25, P75)) 0.02 (0.02, 0.03) 0.03 (0.02, 0.03) 0.02 (0.02, 0.03) 0.788
RBC (×10^12/L, M (P25, P75)) 4.13 (3.88, 4.53) 4.32 (3.95, 4.68) 4.09 (3.84, 4.51) 0.276
Hb (g/L, Mean (SD)) 125.54 (16.33) 133.00 (21.00) 124.00 (14.00) 0.031
HCT (%, Mean (SD)) 38.20 (4.73) 40.20 (6.00) 37.70 (4.20) 0.047
MCV (fL, Mean (SD)) 91.38 (4.30) 94.40 (5.20) 90.60 (3.60) <0.001
MCH (pg, M (P25, P75)) 30.00 (29.10, 31.10) 31.25 (30.40, 31.80) 29.60 (28.90, 30.40) <0.001
MCHC (g/L, Mean (SD)) 328.31 (6.87) 330.00 (7.00) 328.00 (7.00) 0.138
RDW-CV (%, M (P25, P75)) 12.80 (12.40, 13.30) 12.80 (12.30, 13.50) 12.80 (12.40, 13.30) 0.719
RDW-SD (%, M (P25, P75)) 43.20 (41.50, 44.80) 43.95 (42.68, 46.48) 42.90 (41.35, 44.55) 0.063
PLT (×10^9/L, M (P25, P75)) 202.00 (158.00, 262.00) 174.50 (132.75, 263.00) 208 (162.5, 259.5) 0.299
PCT (%, M (P25, P75)) 0.19 (0.14, 0.25) 0.185 (0.13, 0.2325) 0.20 (0.15, 0.26) 0.426
PDW (fL, Mean (SD)) 16.24 (0.41) 16.40 (0.40) 16.20 (0.40) 0.018
MPV (fL, M (P25, P75)) 9.40 (8.70, 10.30) 10.10 (8.55, 10.60) 9.40 (8.70, 10.05) 0.41
Open in a new tab

L1 BMD: lumbar spine 1 bone mineral density; L2 BMD: lumbar spine 2 bone mineral density; L3 BMD: lumbar spine 3 bone mineral density; L4 BMD: lumbar spine 4 bone mineral density; L1–L4 BMD: lumbar spine 1–4 bone mineral density; BMD: bone mineral density; BMI: body mass index; WBC: white blood cell; NEU: neutrophil; LYMP: lymphocyte; MONO: monocyte; EO: eosinophil; BO: basophil; RBC: red blood cell; Hb: hemoglobin; HCT: hematocrit; MCV: mean corpuscular volume; MCH: mean corpuscular hemoglobin; MCHC: mean corpuscular hemoglobin concentration; RDW: red blood cell distribution width; PLT: platelet; PCT: plateletcrit; PDW: platelet distribution width; MPV: mean platelet volume; M: median; P25: 25th percentile; P75: 75th percentile; RDW-CV: red blood cell distribution width-coefficient of variation.

The p-value for the difference between men and women was calculated using independent samples t-test for normal distribution and Kruskal–Wallis test for non-normal distribution.

Factors related to lumbar BMD in univariate linear regression analysis

BMI (β = 0.217 (0.000084, 0.028), p = 0.049), RBC count (β = 0.253 (0.013, 0.157), p = 0.021), Hb level (β = 0.332 (0.002, 0.007), p = 0.002), and PCT (β = 0.325 (0.005, 0.025), p = 0.003) were associated with L1–L4 BMD (Table 2). Additionally, BMI was positively correlated with L1 BMD; RBC count with L1, L2, and L3 BMD; and Hb level and PCT with L1, L2, L3, and L4 BMD (p < 0.05) (Supplementary Tables 1 to 4). When stratified by sex, among male participants, BMI (β = 0.541 (0.006, 0.062), p = 0.02), Hb level (β = 0.47 (0.00003, 0.012), p = 0.049), and PCT (β = 0.47 (0.000086,0.042), p = 0.049) were positively correlated with L1–L4 BMD. Among female participants, age was negatively correlated with lumbar spine BMD (p < 0.05) (Table 2). Additionally, among male participants, BMI was positively correlated with L1, L2, L3, and L4 BMD, whereas Hb level and PCT were positively correlated with L1 and L2 BMD (Supplementary Tables 1 to 4).

Table 2.

Univariate linear regression analysis of factors associated with L1–L4 bone mineral density (g/cm2) in participants stratified by sex.

Variable All patients Sex
Male (N = 18) Female(N = 65)
STDβ (95% CI) p-value STDβ (95% CI) p-value STDβ (95% CI) p-value
Age (years) −0.073 (−0.006, 0.003) 0.512 0.185 (−0.008, 0.017) 0.462 −0.293 (−0.009,−0.001) 0.018
BMI (kg/m2) 0.217 (0.000084, 0.028) 0.049 0.541 (0.006, 0.062) 0.020 0.158 (−0.005,0.024) 0.208
WBC count (×10^9/L) −0.082 (−0.021, 0.01) 0.463 −0.161 (−0.054, 0.029) 0.523 −0.111 (−0.022,0.008) 0.379
NEU count (%)) −0.002 (−0.004, 0.004) 0.984 0.022 (−0.009, 0.01) 0.930 −0.019 (−0.004,0.003) 0.879
LYMP count (%) −0.001 (−0.004, 0.004) 0.992 0.067 (−0.011, 0.014) 0.793 0.04 (−0.004,0.005) 0.755
MONO count (%) −0.018 (−0.019, 0.016) 0.875 −0.159 (−0.045, 0.024) 0.530 −0.124 (−0.031,0.01) 0.326
EO count (%) 0.055 (−0.022, 0.037) 0.618 −0.206 (−0.081, 0.035) 0.412 0.066 (−0.025,0.043) 0.602
BO count (%) −0.022 (−0.256, 0.21) 0.844 0.269 (−0.364, 1.179) 0.280 −0.031 (−0.24,0.187) 0.804
NEU count (×10^9/L) −0.078 (−0.023, 0.011) 0.483 −0.12 (−0.052, 0.033) 0.634 −0.117 (−0.024,0.009) 0.355
LYMP count (×10^9/L) −0.023 (−0.081, 0.066) 0.835 −0.013 (−0.215, 0.205) 0.959 0.001 (−0.069,0.07) 0.994
MONO count (×10^9/L) −0.058 (−0.221, 0.128) 0.600 −0.168 (−0.612, 0.314) 0.505 −0.153 (−0.275,0.065) 0.204
EO count (×10^9/L) 0.03 (−0.349, 0.459) 0.787 −0.308 (−1.274, 0.308) 0.214 0.085 (−0.303,0.614) 0.488
BO count (×10^9/L) −0.076 (−5.17, 2.525) 0.496 −0.058 (−17.121, 13.741) 0.819 −0.091 (−4.639,2.169) 0.471
RBC count (×10^12/L) 0.253 (0.013, 0.157) 0.021 0.41 (−0.029, 0.361) 0.091 0.168 (−0.022,0.116) 0.182
Hb level (g/L) 0.332 (0.002, 0.007) 0.002 0.47 (0.00003, 0.012) 0.049 0.151 (−0.001,0.005) 0.231
HCT level (%) 0.325 (0.005, 0.025) 0.003 0.47 (0.000086, 0.042) 0.049 0.154 (−0.004,0.018) 0.219
MCV (fL) 0.142 (−0.004, 0.019) 0.200 0.004 (−0.027, 0.027) 0.987 −0.012 (−0.013,0.012) 0.926
MCH level (pg) 0.18 (−0.005, 0.058) 0.104 0.087 (−0.065, 0.091) 0.730 −0.018 (−0.039,0.034) 0.886
MCHC (g/L) 0.139 (−0.003, 0.012) 0.211 0.224 (−0.012, 0.03) 0.373 0.029 (−0.006,0.008) 0.820
RDW-CV (%) 0.018 (−0.052, 0.061) 0.872 −0.42 (−0.334, 0.023) 0.083 0.154 (−0.019,0.082) 0.220
RDW-SD (%) 0.1 (−0.009, 0.024) 0.370 −0.28 (−0.056, 0.016) 0.260 0.149 (−0.007,0.029) 0.236
PLT (×10^9/L) −0.195 (−0.001, 0.000055) 0.077 −0.264 (−0.003, 0.001) 0.290 −0.144 (−0.001,0.000217) 0.253
PCT (%) −0.192 (−1.21, 0.074) 0.082 −0.249 (−2.988, 1.034) 0.318 −0.155 (−0.968,0.225) 0.230
PDW (fL) 0.211 (−0.003, 0.235) 0.211 0.116 (−0.265, 0.414) 0.648 0.125 (−0.06,0.179) 0.320
MPV (fL) 0.051 (−0.037, 0.06) 0.647 0.241 (−0.071, 0.195) 0.335 −0.07 (−0.059,0.033) 0.578
Open in a new tab

STDβ: standardization coefficient beta; CI: confidence interval; BMI: body mass index; WBC: white blood cell; NEU: neutrophil; LYMP: lymphocyte; MONO: monocyte; EO: eosinophil; BO: basophil; RBC: red blood cell; Hb: hemoglobin; HCT: hematocrit; MCV: mean corpuscular volume; MCH: mean corpuscular hemoglobin; MCHC: mean corpuscular hemoglobin concentration; RDW: red blood cell distribution width; PLT: platelet count; PCT: plateletcrit; PDW: platelet distribution width; MPV: mean platelet volume; RDW-CV: red blood cell distribution width-coefficient of variation.

Correlation between CBC parameters and L1, L2, L3, L4, and L1–L4 BMD on multivariate linear regression analysis

BMI and age were adjusted to analyze the linear correlations between CBC parameters (WBC count, RBC count, PLT, and Hb level) and L1, L2, L3, L4, and L1–L4 BMD. In the unstratified group, after adjusting for BMI and age, Hb levels demonstrated a significant positive correlation with L1, L2, L3, and L1–L4 BMD (β = 0.294 (0.001, 0.007), p = 0.009). RBC count was positively correlated with L1 and L2 BMD (p < 0.05). Among male participants, Hb levels demonstrated a positive correlation with L1 BMD (β = 0.479 (0.000454, 0.011), p = 0.036). However, this association was not observed in female participants (Table 3).

Table 3.

Multivariate linear regression analysis of the association between CBC parameters and lumbar BMD (L1–L4) stratified by sex.

WBC count RBC count PLT Hb level
STDβ (95% CI) p-value STDβ (95% CI) p-value STDβ (95% CI) p-value STDβ (95% CI) p-value
All patients −0.092 (−0.022, 0.009) 0.403 0.213 (−0.002, 0.145) 0.057 −0.187 (−0.001, 0.000071) 0.087 0.294 (0.001, 0.007) 0.009
Sex
Male (N = 18) −0.186 (−0.048, 0.018) 0.351 0.348 (−0.058, 0.34) 0.151 −0.091 (−0.002, 0.001) 0.69 0.442 (−0.001, 0.012) 0.072
Female (N = 65) −0.134 (−0.022, 0.006) 0.272 0.146(−0.027,0.108) 0.231 −0.162 (−0.001, 0.000159) 0.18 0.107 (−0.002, 0.005) 0.388
Lumbar L1 BMD
All Patients −0.052 (−0.019, 0.012) 0.634 0.225 (0.003, 0.151) 0.043 −0.167 (−0.001, 0.000119) 0.123 0.281 (0.001, 0.007) 0.012
Sex
Male (N = 18) −0.112 (−0.04, 0.022) 0.558 0.407 (−0.016, 0.335) 0.072 −0.006 (−0.002, 0.002) 0.979 0.479 (0.000454, 0.011) 0.036
Female (N = 65) −0.104 (−0.021, 0.008) 0.388 0.158 (−0.023, 0.115) 0.19 −0.161 (−0.001, 0.000161) 0.177 0.09 (−0.002, 0.005) 0.459
Lumbar L2 BMD
All patients −0.096 (−0.023, 0.009) 0.386 0.268 (0.017, 0.166) 0.017 −0.205 (−0.001, 0.000027) 0.062 0.356 (0.002, 0.008) 0.001
Sex
Male (N = 18) −0.216 (−0.048, 0.015) 0.275 0.453 (−0.003, 0.36) 0.053 0.553 (0.001, 0.012) 0.487 −0.156 (−0.002, 0.001) 0.487
Female (N = 65) −0.126 (−0.023, 0.007) 0.301 0.206 (−0.01, 0.131) 0.091 −0.173 (−0.001, 0.000145) 0.154 0.184 (−0.001, 0.006) 0.134
Lumbar L3 BMD
All patients −0.098 (−0.023, 0.009) 0.378 0.199 (−0.008, 0.147) 0.079 −0.168 (−0.001, 0.000132) 0.129 0.3 (0.001, 0.007) 0.008
Sex
Male (N = 18) −0.248 (−0.06, 0.015) 0.226 0.339 (−0.081, 0.388) 0.181 −0.17 (−0.003, 0.001) 0.468 0.483 (−0.000197, 0.014040) 0.056
Female (N = 65) −0.108 (−0.021, 0.008) 0.382 0.134 (−0.032, 0.107) 0.283 −0.117 (−0.001, 0.000267) 0.345 0.108 (−0.002, 0.005) 0.389
Lumbar L4 BMD
All patients −0.101 (−0.024, 0.009) 0.359 0.146 (−0.027, 0.13) 0.196 −0.162 (−0.001, 0.000142) 0.138 0.22 (−0.000030, 0.006) 0.052
Sex
Male (N = 18) −0.135 (−0.05, 0.027) 0.529 0.209 (−0.15, 0.331) 0.433 0.004 (−0.002, 0.002) 0.988 0.249 (−0.004, 0.011) 0.362
Female (N = 65) −0.161 (−0.025, 0.005) 0.188 0.072 (−0.051, 0.094) 0.559 −0.145 (−0.001, 0.000212) 0.237 0.042 (−0.003, 0.004) 0.733
Open in a new tab

Adjusted for age and BMI.

STDβ: standardization coefficient beta; CI: confidence interval; WBC: white blood cell; RBC: red blood cell; Hb: hemoglobin; PLT: platelet count; BMI: body mass index; CBC: complete blood count; L1 BMD: lumbar spine 1 bone mineral density; L2 BMD: lumbar spine 2 bone mineral density; L3 BMD: lumbar spine 3 bone mineral density; L4 BMD: lumbar spine 4 bone mineral density; L1–L4 BMD: lumbar spine 1–4 bone mineral density; BMD: bone mineral density.

Nonlinear relationship of CBC parameters with L1, L2, L3, L4 and L1–L4 BMD

There was a significant nonlinear relationship between Hb levels and L1–L4 BMD and between PLT and L1–L4 BMD, as indicated by the fitted smoothing curves (p < 0.05) (Figure 2). As depicted in Supplementary Figures 2 and 3, RBC count also demonstrated a nonlinear relationship with L1 and L2 BMD (p < 0.05), whereas PLT demonstrated a nonlinear relationship with L4 BMD (p = 0.029), which exhibited an inverted U-shaped association (Supplementary Figures 3 and 4).

Figure 2.

Figure 2.

Open in a new tab

RCS regression analysis of the association between complete blood count parameters and lumbar bone mineral density. RCS regression analysis of the associations between (a) hemoglobin level and L1–L4 bone mineral density, (b) white blood cell count and L1–L4 bone mineral density, (c) red blood cell count and L1–L4 bone mineral density, and (d) platelet count and L1–L4 bone mineral density.

RCS: restricted cubic spline.

Discussion

In this study involving 83 older adults, there was a positive, nonlinear correlation between Hb levels and lumbar BMD. This association was evident in males but not in females. Additionally, there were associations between RBC count, PLT, and lumbar BMD.

Anemia and OP frequently co-occur among older adults. According to the first epidemiological survey on osteoporosis in China released by the National Health Commission in 2018, the prevalence of OP among people aged 50 and above was 19.2%. 24 Anemia incidence increases rapidly after the age of 50 years and affects 14.1% of older men and 10.2% of older women in the US.25,26 Certain clinical studies have demonstrated an association between anemia and BMD. Previous studies have indicated that Hb levels are positively correlated with BMD.6,7,8 More recently, a study conducted in Shanghai, China, has shown that Hb levels are positively correlated with BMD among older individuals in China. 19 Another cross-sectional study has reported a nonlinear association between adult Hb levels and lumbar BMD. 18 Moreover, inconsistent findings have been reported in the literature. For instance, a cross-sectional study conducted in South Korea has revealed that among non-anemic adults, Hb levels were negatively correlated with low BMD. 10 These contradictory reports might be attributed to factors related to participant selection, including, whether the study enrolled individuals taking medications that could affect BMD and HCT levels. In our study, we excluded risk factors that could influence the study results and have explored nonlinear relationships using general linear models.

Hb plays a vital role in the process of oxygen transport. Abnormal Hb levels can affect the oxygenation state as well as lipid metabolism within the bone marrow. This can exert a significant impact on the metabolism of bone cells and ultimately on BMD. 27 Research has indicated that hypoxic environments can influence bone metabolism. Additionally, anemia can cause bone loss, reduce BMD, and aid OP development. 28 With age, there is a decline in Hb levels owing to reduced RBC production resulting from hematopoietic dysfunction. This dysfunction encompasses decreased regenerative capacity and a bias toward myeloid differentiation. 29 Concurrently, BMD also tends to reduce.30,31 Another study has further demonstrated that impaired bone marrow hematopoietic function can impact bone metabolic levels. 32 Previous investigations have suggested that higher Hb levels exert some degree of protection for BMD in older adults. 33 This conclusion is consistent with the findings of our study.

In our study, in addition to Hb levels, PLT also demonstrated an associated with lumbar BMD. The relationship between PLT and bone metabolism remains debatable in the literature. Some studies have reported a negative linear correlation between PLT and BMD, suggesting that higher PLT is a risk factor for bone loss.34,35 However, other studies have identified sex-specific differences, with no significant association among women and a significant negative correlation observed among men. 36 These inconsistent findings may be attributed to differences in study populations, sample sizes, or failure to account for potential nonlinearity. In contrast to the traditional linear view, our study identified a nonlinear, inverted U-shaped relationship between PLT and lumbar BMD. To the best of our knowledge, this specific nonlinear pattern has rarely been reported in previous research. This inverted U-shaped curve suggests that the effect of platelets on bone is not simply unidirectional; a moderate PLT may be beneficial for bone health, whereas both extremely low and high levels could be detrimental. This finding may help explain the discrepancies in prior studies as linear models might have obscured the true complexity of the relationship.

Our study concentrated on older adults; however, recent investigations have also reported associations between CBC parameters and bone density in pediatric populations, including among children with obesity and thyroid diseases.37,38 These findings suggest that hematological markers reflect bone metabolism across ages, reinforcing the utility of routine blood tests in bone health assessment.

This study has certain limitations. First, the cross-sectional design precludes any causal inferences. Second, the modest sample size, particularly the small number of male participants (n = 18), may limit the statistical power. However, notably, despite the relatively small number of male participants, significant associations were still detected in our cohort, suggesting the presence of a meaningful biological effect. This finding aligns with previous research; a study that used the Fracture Risk Assessment Tool and enrolled 662 male patients reported that Hb levels were positively correlated with BMD. 39 Similarly, a study in Korean adults has identified a positive association between Hb levels and BMD, specifically in the male population. 17 Third, we were unable to account for potential confounders such as detailed nutritional status (dietary calcium and vitamin D intake), physical activity, smoking habits, and menopausal status/sex hormones. These unmeasured factors may partially confound the observed associations and should be addressed in future studies. Future large-scale prospective studies that include these variables are warranted to validate and expand upon our findings.

The sex-specific association might relate to the influence of sex hormones on bone health. 40 Low testosterone levels in men are associated with both anemia and reduced BMD, creating a direct physiological pathway that explains the positive correlation observed in men.41,42 In contrast, the relationship in women could be more complex. Given that all our study participants were aged >50 years, most women in this cohort were presumed to be postmenopausal; thus, factors such as estrogen decline and menopausal status may have obscured a direct association of Hb levels with BMD. We did not measure the levels of sex hormones or menopausal status, which might have contributed to the observed sex-based differences.

Moreover, nutritional status and exercise levels exert a significant impact on the relationship between Hb levels and BMD. An unhealthy diet may result in low protein intake and malnutrition. This, in turn, can give rise to poor nutritional conditions, anemia, hypoalbuminemia, and reduced BMD. 43 Research has demonstrated that individuals with higher Hb levels are likely to show superior physical performance and endurance since their muscles can obtain more oxygen and nutrients. Additionally, regular weightlifting and other physical activities are proven to increase BMD.44,45

Low Hb levels and low BMD are not isolated health concerns. As demonstrated by our results, among older adults, low Hb levels can cause a reduction of BMD, which can subsequently trigger well-known health conditions such as infections and fractures. Hb levels can be conveniently measured via routine blood tests, which are relatively inexpensive and easy to perform, thus enabling early detection of low BMD and related problems. Timely corrective measures and interventions can then be implemented to prevent potentially serious health issues. Based on our findings, we suggest that older patients suffering from anemia should be considered for BMD surveillance. In contrast, patients diagnosed with OP may benefit from the inclusion of blood tests in health screening.

Taken together, our research has both merits and limitations. By excluding medications that could impact CBC parameters and BMD, our results are more likely to be applicable to most older adults. The utilization of data from 2024, as opposed to data from 2018 to 2020 (as in previous studies) helped us bridge a gap in the existing research. However, the relatively small size of the sample, which only included older individuals aged >50 years, restricts the generalizability of our findings. Additionally, the study failed to consider other factors such as nutritional status and physical activity levels that could affect the relationship between Hb levels and BMD. Moreover, as a cross-sectional study, it is unable to establish causal relationships. To improve our understanding of the association between Hb levels and BMD, future research should be conducted on larger samples. More comprehensive studies are needed, and long-term follow-up studies can provide deeper insights into the causal mechanisms underlying the relationship between Hb levels and BMD, thus enhancing the overall quality and reliability of the research in this area.

Conclusion

In this cross-sectional study involving older Chinese individuals, Hb levels were positively associated with lumbar BMD in a nonlinear pattern, particularly among males. Although monitoring of Hb levels may serve as a useful adjunct for early risk assessment of lumbar bone metabolic disorders, the cross-sectional design precludes causal inferences, and these findings should be interpreted as preliminary. Further large-scale longitudinal studies are essential to validate these associations and establish causality.

Supplemental Material

sj-docx-1-imr-10.1177_03000605261474625 - Supplemental material for Retrospective cross-sectional associations of whole blood cell count and hemoglobin level with lumbar bone mineral density in older adults
sj-docx-1-imr-10.1177_03000605261474625.docx (1.7MB, docx)

Supplemental material, sj-docx-1-imr-10.1177_03000605261474625 for Retrospective cross-sectional associations of whole blood cell count and hemoglobin level with lumbar bone mineral density in older adults by Fengxian Cao, Zhengfu Wang, Fengxi Chen, Qiuyan Cao, Qingyan Tang, Hongzheng He and Lei Zhang in Journal of International Medical Research

Acknowledgments

The authors would like to thank all the participants of this study.

Footnotes

Ethical approval: This study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Kunshan Jinxi People's Hospital (Approval No. LCYJIEC2025-001-01).

Consent to participate: The requirement for written informed consent was waived by the Ethics Committee of Kunshan Jinxi People’s Hospital due to the retrospective nature of the study, which used fully anonymized data in accordance with national regulations and ethical guidelines.

Consent for publication: Not applicable, as this study does not contain any individual person's data in any form.

Author contributions: Fengxian Cao and Zhengfu Wang contributed equally to this work and share first authorship. Fengxian Cao and Zhengfu Wang: conceptualization, data curation, formal analysis, investigation, methodology, and writing–original draft. Fengxi Chen: conceptualization, project administration, supervision, and writing–review & editing. Qiuyan Cao: data curation, investigation, and methodology. Qingyan Tang and Hongzheng He: data curation, investigation, and resources. Lei Zhang: formal analysis, methodology, and validation. All authors reviewed and approved the final manuscript.

Funding: The authors received no financial support for the research, authorship, and/or publication of this article.

The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability: The data that support the findings of this study are available from the corresponding author upon reasonable request. The data are not publicly available due to privacy or ethical restrictions.

Supplemental material: Supplemental material for this article is available online.

References

  • 1.Clynes MA, Harvey NC, Curtis EM, et al. The epidemiology of osteoporosis. Br Med Bull 2020; 133: 105–117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Lupsa BC, Insogna K. Bone health and osteoporosis. Endocrinol Metab Clin North Am 2015; 44: 517–530. [DOI] [PubMed] [Google Scholar]
  • 3.Sözen T, Özışık L, Başaran NÇ. An overview and management of osteoporosis. Eur J Rheumatol 2017; 4: 46–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Hernlund E, Svedbom A, Ivergård M, et al. Osteoporosis in the European Union: medical management, epidemiology and economic burden. A report prepared in collaboration with the International Osteoporosis Foundation (IOF) and the European Federation of Pharmaceutical Industry Associations (EFPIA). Arch Osteoporos 2013; 8: 136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Qu B, Ma Y, Yan M, et al. The economic burden of fracture patients with osteoporosis in western China. Osteoporos Int 2014; 25: 1853–1860. [DOI] [PubMed] [Google Scholar]
  • 6.Cesari M, Pahor M, Lauretani F, et al. Bone density and hemoglobin levels in older persons: results from the InCHIANTI study. Osteoporos Int 2005; 16: 691–699. [DOI] [PubMed] [Google Scholar]
  • 7.Laudisio A, Marzetti E, Pagano F, et al. Haemoglobin levels are associated with bone mineral density in the elderly: a population-based study. Clin Rheumatol 2009; 28: 145–151. [DOI] [PubMed] [Google Scholar]
  • 8.Korkmaz U, Korkmaz N, Yazici S, et al. Anemia as a risk factor for low bone mineral density in postmenopausal Turkish women. Eur J Intern Med 2012; 23: 154–158. [DOI] [PubMed] [Google Scholar]
  • 9.Oh YH, Moon JH, Cho B. Association between hemoglobin level and bone mineral density in Korean adults. J Bone Metab 2017; 24: 161–173. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Kim HS, Park HM, Lee HSet al. et al. Hemoglobin levels and low bone mineral density in non-anemic older adults: secondary analysis of the Korean national health and nutrition examination survey. Exp Gerontol 2019; 126: 110706. [DOI] [PubMed] [Google Scholar]
  • 11.Liu H, Li X, Guan Y, et al. Comprehensive evaluation and analysis of human Settlements’ suitability in the Yangtze river Delta based on multi-source data. Int J Environ Res Public Health 2023; 20: 1354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Zhang L, Tang J, Xu M, et al. Spatial-temporal evolution of population aging in the Yangtze river delta urban agglomeration of China. PLOS One 2024; 19: e0298199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Dixon LR. The complete blood count: physiologic basis and clinical usage. J Perinat Neonatal Nurs 1997; 11: 1–18. [DOI] [PubMed] [Google Scholar]
  • 14.Palladino M. Complete blood count alterations in COVID-19 patients: a narrative review. Biochem Med (Zagreb) 2021; 31: 030501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Balarajan Y, Ramakrishnan U, Ozaltin E, et al. Anaemia in low-income and middle-income countries. Lancet 2011; 378: 2123–2135. [DOI] [PubMed] [Google Scholar]
  • 16.Bani Hassan E, Vogrin S, Hernandez Viña I, et al. Hemoglobin levels are low in sarcopenic and osteosarcopenic older persons. Calcif Tissue Int 2020; 107: 135–142. [DOI] [PubMed] [Google Scholar]
  • 17.Hirani V, Naganathan V, Blyth F, et al. Low hemoglobin concentrations are associated with sarcopenia, physical performance, and disability in older Australian men in cross-sectional and longitudinal analysis: the concord health and ageing in men project. J Gerontol A Biol Sci Med Sci 2016; 71: 1667–1675. [DOI] [PubMed] [Google Scholar]
  • 18.Cheng X, Zhang L, Cao Bet al. et al. Correlation between serum hemoglobin levels and bone mineral density in adults: a cross-sectional study. Medicine (Baltimore) 2024; 103: e39612. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Liu Y, Zeng Y, Lu J, et al. Correlation of hemoglobin with osteoporosis in elderly Chinese population: a cross-sectional study. Front Public Health 2023; 11: 1073968. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Ye T, Lu L, Guo Let al. et al. Association between hemoglobin levels and osteoporosis in Chinese patients with type 2 diabetes mellitus: a cross-sectional study. Diabetes Metab Syndr Obes 2022; 15: 2803–2811. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Li L, Ge JR, Chen J, et al. Association of bone mineral density with peripheral blood cell counts and hemoglobin in Chinese postmenopausal women: a retrospective study. Medicine (Baltimore) 2020; 99: e20906. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Ashpole NM, Herron JC, Mitschelen MC, et al. IGF-1 regulates vertebral bone aging through sex-specific and time-dependent mechanisms. J Bone Miner Res 2016; 31: 443–454. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Cherukuri L, Kinninger A, Birudaraju D, et al. Effect of body mass index on bone mineral density is age-specific. Nutr Metab Cardiovasc Dis 2021; 31: 1767–1773. [DOI] [PubMed] [Google Scholar]
  • 24.Lyu F, Ramoo V, Chui PL, et al. Prevalence rate of primary osteoporosis in China: a meta-analysis. BMC Public Health 2024; 24: 1518. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Guralnik JM, Eisenstaedt RS, Ferrucci L, et al. Prevalence of anemia in persons 65 years and older in the United States: evidence for a high rate of unexplained anemia. Blood 2004; 104: 2263–2268. [DOI] [PubMed] [Google Scholar]
  • 26.Seitz AE, Eberhardt MS, Lukacs SL. Anemia prevalence and trends in adults aged 65 and older: U.S. national health and nutrition examination survey, 2001-2004 to 2013-2016. J Am Geriatr Soc 2018; 66: 2431–2432. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Fan J, Jahed V, Klavins K. Metabolomics in bone research. Metabolites 2021; 11: 434. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Rutten EP, Franssen FM, Spruit MAet al. et al. Anemia is associated with bone mineral density in chronic obstructive pulmonary disease. COPD 2013; 10: 286–292. [DOI] [PubMed] [Google Scholar]
  • 29.Kuribayashi W, Oshima M, Itokawa N, et al. Limited rejuvenation of aged hematopoietic stem cells in young bone marrow niche. J Exp Med 2021; 218: e20192283. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Stauder R, Valent P, Theurl I. Anemia at older age: etiologies, clinical implications, and management. Blood 2018; 131: 505–514. [DOI] [PubMed] [Google Scholar]
  • 31.Gurevitch O, Slavin S, Feldman AG. Conversion of red bone marrow into yellow-cause and mechanisms. Med Hypotheses 2007; 69: 531–536. [DOI] [PubMed] [Google Scholar]
  • 32.Despars G, St-Pierre Y. Bidirectional interactions between bone metabolism and hematopoiesis. Exp Hematol 2011; 39: 809–816. [DOI] [PubMed] [Google Scholar]
  • 33.Xiu S, Mu Z, Sun L, et al. Hemoglobin level and osteoporosis in Chinese elders with type 2 diabetes mellitus. Nutr Diabetes 2022; 12: 19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Kim J, Kim HS, Lee HSet al. et al. The relationship between platelet count and bone mineral density: results from two independent population-based studies. Arch Osteoporos 2020; 15: 43. [DOI] [PubMed] [Google Scholar]
  • 35.Zhang Y, Zhao B, Huang Z, et al. Sex-specific association between platelet content and bone mineral density in adults: a cross-sectional study. BMC Musculoskelet Disord 2024; 25: 875. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Kikuchi S, Wada A, Kamihara Y, et al. A novel mechanism for bone loss: platelet count negatively correlates with bone mineral density via megakaryocyte-derived RANKL. Int J Mol Sci 2023; 24: 12150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Bala MM, Bala KA. Bone mineral density and complete blood count ratios in children and adolescents with obesity. Eur Rev Med Pharmacol Sci 2022 Jan; 26: 249–256. [DOI] [PubMed] [Google Scholar]
  • 38.Bala MM, Bala KA. Bone mineral density (BMD) and neutrophil-lymphocyte ratio (NLR), monocyte-lymphocyte ratio (MLR), and platelet-lymphocyte ratio (PLR) in childhood thyroid diseases. Eur Rev Med Pharmacol Sci 2022; 26: 1945–1951. [DOI] [PubMed] [Google Scholar]
  • 39.Chuang MH, Chuang TL, Koo Met al. et al. Low hemoglobin is associated with low bone mineral density and high risk of bone fracture in male adults: a retrospective medical record review study. Am J Mens Health 2019; 13: 1557988319850378. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Lee EA, Shin DW, Yoo JH, et al. Anemia and risk of fractures in older Korean adults: a nationwide population-based study. J Bone Miner Res 2019; 34: 1049–1057. [DOI] [PubMed] [Google Scholar]
  • 41.Lewerin C, Nilsson-Ehle H, Jacobsson S, et al. Serum estradiol associates with blood hemoglobin in elderly men: the MrOS Sweden study. J Clin Endocrinol Metab 2014; 99: 2549–2556. [DOI] [PubMed] [Google Scholar]
  • 42.Kenny AM, Prestwood KM, Marcello KMet al. et al. Determinants of bone density in healthy older men with low testosterone levels. J Gerontol A Biol Sci Med Sci 2000; 55: M492–M497. [DOI] [PubMed] [Google Scholar]
  • 43.Lopes SO, Ribeiro SAV, Morais DC, et al. Factors associated with anemia among adults and the elderly family farmers. Int J Environ Res Public Health 2022; 19: 7371. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Herbert AJ, Williams AG, Hennis PJ, et al. The interactions of physical activity, exercise and genetics and their associations with bone mineral density: implications for injury risk in elite athletes. Eur J Appl Physiol 2019; 119: 29–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Zhao X, Yang Y, Yue Ret al. et al. Potential causal association between leisure sedentary behaviors, physical activity and musculoskeletal health: a Mendelian randomization study. PLOS One 2023; 18: e0283014. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

sj-docx-1-imr-10.1177_03000605261474625 - Supplemental material for Retrospective cross-sectional associations of whole blood cell count and hemoglobin level with lumbar bone mineral density in older adults
sj-docx-1-imr-10.1177_03000605261474625.docx (1.7MB, docx)

Supplemental material, sj-docx-1-imr-10.1177_03000605261474625 for Retrospective cross-sectional associations of whole blood cell count and hemoglobin level with lumbar bone mineral density in older adults by Fengxian Cao, Zhengfu Wang, Fengxi Chen, Qiuyan Cao, Qingyan Tang, Hongzheng He and Lei Zhang in Journal of International Medical Research

Articles from The Journal of International Medical Research are provided here courtesy of SAGE Publications

RESOURCES