J Bone Miner Metab (2014) 32:718–724
DOI 10.1007/s00774-013-0548-4
ORIGINAL ARTICLE
Homocysteine, an additional factor, is linked to osteoporosis
in postmenopausal women with type 2 diabetes
Li Jianbo • Hongman Zhang • Lingfei Yan •
Min Xie • Yan Mei • Chen Jiawei
Received: 11 June 2013 / Accepted: 11 November 2013 / Published online: 24 December 2013
Ó The Japanese Society for Bone and Mineral Research and Springer Japan 2013
Abstract We explored the relationship between plasma
total homocysteine concentration and osteoporosis in
postmenopausal patients with type 2 diabetes. Postmeno-
pausal patients with type 2 diabetes (n = 258) were
enrolled in a cross-sectional hospital-based study. Osteo-
porosis was documented by dual energy X-ray absorpti-
ometry. Plasma total homocysteine concentration was
measured using fluorescence polarization immunoassay.
Risk factors for osteoporosis and determinants of homo-
cysteine were obtained from blood samples and interviewer
questionnaire. We found that plasma total homocysteine
levels were higher in subjects with osteoporosis and dia-
betes than without [(9.5 ± 2.0) vs. (10.4 ± 2.4) lmol/l,
p = 0.001]. The association of homocysteine with osteo-
porosis was independent of possible risk factors for oste-
oporosis in diabetes (e.g., duration of diabetes, HbA1c,
body mass index, serum 25-hydroxyvitamin D, thiazolid-
inediones, and retinopathy) and determinants of homo-
cysteine concentration (age, serum folate and vitamin B12,
renal status, and biguanide use) [OR 1.40 (1.02–1.90),
p = 0.036]. In addition, bone mineral density was closely
correlated with homocysteine as a continuous variable after
adjusting for age [r = -0.64 (-0.69 to -0.58),
p = 0.002]. Furthermore, per increase of 5.0 lmol/l,
plasma homocysteine was related to osteoporosis, after
controlling for per unit increase of other factors [OR 1.42
(1.07–1.96), p = 0.027]. The optimal cut-off point for the
L. Jianbo (&)
H. Zhang
L. Yan
M. Xie

Y. Mei
C. Jiawei
Diabetes and Osteoporosis Study Group, Endocrinology
and Metabolism Department, The First Affiliated Hospital
of Nanjing Medical University, Guangzhou Road 300,
Nanjing 210029, China
e-mail: ljbzjlx18@yahoo.com.cn
123
plasma homocysteine level distinguishing diabetic patients
with osteoporosis from without was 10.18 lmol/l. The
results suggest that plasma total homocysteine concentra-
tion is independently associated with the occurrence of
osteoporosis in postmenopausal patients with type 2 dia-
betes. Future prospective studies are warranted to clarify
the relationship.
Keywords Homocysteine
Osteoporosis

Postmenopausal women
Type 2 diabetes
Introduction
Type 2 diabetes mellitus (T2DM) has become a major
public health problem [1]. T2DM may increase a risk for a
fracture as a result of osteoporosis. Once a fracture has
occurred, healing is delayed. Therefore predication and
early intervention of osteoporosis is increasingly impera-
tive among diabetes related disorders. However, the
underlying pathogenesis of osteoporosis in diabetes has not
been well defined.
Homocysteine is associated with decreases of bone blood
flow and biomechanical properties [2]. Hyperhomocystein-
emia results in increasing production of oxidation products,
homocysteine thiolactone and homocysteine mixed disul-
phides, which damages endothelium by excessive sulphation
of connective tissues [3, 4]. These may decrease bone mineral
density (BMD) and strength. Furthermore, hyperhomocy-
steinemia is more common in patients with type 2 diabetes
[5]. Therefore, the hypothesis that homocysteine may act as a
causal factor and even as a predictor of diabetes associated
osteoporosis deserves consideration.
There are some determinants of plasma homocysteine
levels [6]. Homocysteine concentration is closely related to
J Bone Miner Metab (2014) 32:718–724
folate or renal status in the elderly [7]. Homocysteine
metabolism is dependent on vitamin B12 and folate con-
centration [8]. Certain genetic abnormalities also affect
homocysteine metabolism [9]. In addition, other factors
such as lifestyles (e.g., smoking, alcohol, and coffee intake)
and drugs affect plasma homocysteine levels [6, 10].
A few studies show that diabetes is related to a BMD
decrease [11] and osteoporosis in postmenopausal women
[12]. However, a study on homocysteine in association
with osteoporosis in postmenopausal women with type 2
diabetes is limited. We meticulously examined the rela-
tionship to indicate a possible role of hyperhomocystein-
emia in this scenario.
Materials and methods
Diabetes status and subject selection
Diabetes status was biochemically confirmed according to
the WHO diagnostic criteria for the classification of dia-
betes [13]. Chinese patients with type 2 diabetes registered
consecutively as outpatients from the clinic or inpatients
from the ward of our hospital (a tertiary care hospital)
between August 2010 and October 2011 were routinely
examined for bone density and screened for fractures.
Osteoporosis was confirmed in postmenopausal women
according to the WHO diagnosis and assessment [14]. The
subjects were mainly local, from 7 districts in Nanjing and
randomly enrolled in into two stratified groups, with some
major confounders matched between the diabetic patients
with osteoporosis and without. A stratified multi-staged
random sample design was adopted in the study. The
subjects were first stratified by age with 5-year age bands
(45–49, 50–54, 55–59, 60–64, 65–70), and within each
band, subjects were matched between the two groups. From
the stratum of age, years since menopause were further
stratified on 5-year bands (1–4, 5–9, 10–15), and within
each band, subjects were matched. Likewise, within the
latter stratum, renal function was stratified into 10-lmol/l
bands and subjects were matched. The final 258 subjects
were selected to participate the study from these individual
stratums using simple random sampling. All postmeno-
pausal female volunteers were aged 45–70 years and pro-
vided signed informed consent. The study was approved by
the hospital and university scientific and ethic committee.
Patients on vitamins, with severe renal dysfunction (serum
creatinine [147 lmol/1), severe liver disease (e.g., aspar-
tate aminotransferase or alanine aminotransferase [3 times
the normal level), heart failure NYHA class III or IV, or
any other conditions (e.g., hypercortisolism, hyperpara-
thyroidism, hypogonadism, and hyperthyroidism) or drugs
(e.g., glucocorticoids, sex steroids, warfarin, and
719
bisphosphonates) affecting BMD were excluded. The
selected patients were documented on their previous regi-
mens of hypoglycemic agents or insulin during this study:
73 % on metformin (biguanide), thiazolidinediones 46 %,
77 % on sulfonylureas, 30 % on insulin, and 89 % on more
than one of these agents.
Osteoporosis assessment
BMD of all subjects was measured using dual-energy
X-ray absorptiometry (QDR Discovery, Hologic, Inc.,
USA) by one qualified examiner who was blinded to other
data in this study. Participant positioning and scan analysis
procedures were standardized for all scans with CV\ 0.01;
T- and Z-scores mainly at the lumbar spine (L1–L4) and
femoral neck were collected and analyzed. BMD was
measured in grams per square centimeter (g/cm2). Osteo-
porosis in postmenopausal women was diagnosed when the
BMD was less than or equal to 2.5 standard deviations
below that of a young adult reference population [14]. In
addition, fracture status was assessed by the same qualified
radiologist at the site of the lumbar spine and hip using
X-ray films or MRI, if necessary. A fracture was diagnosed
if 20 % or greater reduction in the site of the bone tested
was observed by two investigators who were blinded to
each other’s readings.
Clinical feature measurement
Body weight was measured before breakfast on the
same scales in light clothing with no shoes, and upright
height was measured on the same wall-mounted stadi-
ometer. Individual body mass index (BMI) was then
calculated as weight (kg)/height (m)2. Smoking status,
total calcium intake, alcohol intake, history of myocar-
dial infarction, stroke and parental history of (vertebral
and non-vertebral) fracture were documented through a
questionnaire. The intake of each nutrient including
calcium was calculated using the following formula:
(reported intake frequency daily) 9 (portion size in
grams) 9 (nutrient content per 100 g)/100 [15]. Physi-
cal activity was classified on the basis of frequency and
duration of mild, moderate and strenuous activities.
Kilocalories of energy expended was calculated (MET
score, kcal h/week/kg). Retinopathy was screened using
stereoscopic retinal photographs.
Blood homocysteine and other biochemical parameter
measurement
Fasting plasma homocysteine concentration was measured
using a commercially available fluorescence polarization
immunoassay (AXSYM, Abbott, USA). This was repeated
123
720
twice in all patients. Fasting plasma glucose and serum
creatinine were measured using an automatic analyzer
(AU5400, Olympus). FSH, LH, E2, fasting plasma insulin,
and serum osteocalcin were measured using chemilumi-
nescence immunoassay (ROCHE E170, Switzerland).
Serum 25-hydroxyvitamin D was measured using chemi-
luminescence immunoassay (LIAISON TOTAL Assay,
DiaSorin Inc., MN, USA). HbA1c was measured using a
high-performance liquid chromatogram (Bio-Rad D10,
USA). Serum C-reactive protein (CRP) was assessed using
nephelometry (Siemens BNII, Germany). Serum folate and
vitamin B12 were determined using automated test assays
(Access, Bechman, USA). Urinary albumin concentration
was measured using immunonephelometry (DCA 2000,
BAYER). Urinary creatinine concentration was measured
using an alkaline picrate method. Individual urinary albu-
min to creatinine ratio was then calculated as albumin
(mg)/creatinine (g).
Statistical analysis
We used SPSS version 13 for Windows (SPSS Inc.,
Chicago, IL, USA) software for statistical analysis.
Descriptive data were expressed as the median with 25th
and 75th quartiles for skewed data or mean (SD) for
normally distributed data. Continuous variables were
tested by Student’s test or the Mann–Whitney U test,
where appropriate. Percentages were compared using the
chi-squared test. Non-normally distributed variables
were log-transformed before analysis. The relations
between osteoporosis and continuous variables were
assessed using ANOVA for variables with homogeneous
variances, or a non-parametric test if non-homogeneous.
The potential variables were first assessed in univariate
analyses. The variables were then analyzed using logistic
regression analysis. Plasma homocysteine was later
added to subsequent models controlling for possible risk
factors for osteoporosis and the known determinants of
homocysteine. Finally, an analysis was made of plasma
homocysteine per unit increase as an independent
determinant for osteoporosis using multivariate logistic
regression models. All odds ratios were given with their
95 % confidence intervals [OR (95 % CI)]. Spearman
correlation coefficients (95 % CI) between BMD and
homocysteine were also assessed. The fit of each model
was tested and the Nagelkerke R2 approximation was
compared. Receiver-operating characteristic (ROC) ana-
lysis was conducted with MedCalc Software version 12.0
for Windows (Mariakerke, Belgium) to assess the
accuracy of plasma homocysteine levels in distinguish-
ing between diabetic patients with osteoporosis and
without. The optimal cut-off point was identified by
calculating the area under the curve (AUC). The data
123
J Bone Miner Metab (2014) 32:718–724
were in the form of cross-sectional observations, and
P \ 0.05 was considered statistically significant.
All procedures were conducted according to the 19th
revision of the Declaration of Helsinki.
Results
The study was completed by 258 patients. Osteoporosis
was confirmed in 133 postmenopausal women with T2DM,
consisting of 115 cases of vertebral osteoporosis with 14
cases of vertebral fractures, 28 cases of femoral neck
osteoporosis without fractures, and 10 cases of the two
coexisting with osteoporosis. In a univariate analysis for
variables (Tables 1, 2), T2DM with and without osteopo-
rosis were compared. There were no differences in fol-
lowing variables: years since menopause, active smokers,
age, fasting insulin, the BMI, total calcium intake, alcohol
intake, energy expended from physical activity (MET), the
history of myocardial infarction and stroke, serum folate,
vitamin B12 and 25-(OH) vitamin D, urinary ACR, and
serum creatinine, etc. Osteoporosis was identified in the
subjects with higher levels of HbA1c, a longer duration of
diabetes, an increased incidence of diabetic retinopathy,
and higher serum CRP, as compared to non-osteoporosis
with diabetes. Both groups did not have obviously different
percentages of thiazolidinediones use. The osteocalcin
level was elevated in the osteoporosis group. Diabetic
patients with osteoporosis had significantly lower BMD
and a higher rate of fractures than without (Table 2).
Plasma homocysteine levels were higher in diabetic indi-
viduals with osteoporosis than without (Table 1). In addi-
tion, a 1-lmol/l increase in plasma homocysteine
concentration was associated with a 1.3–2.8 % incidence
of osteoporosis. The two groups had similar variables that
may potentially influence the homocysteine concentration,
e.g., urinary ACR, serum creatinine, serum folate, vitamin
B12 concentrations, and the percentage of metformin use.
In a multivariate model of multiple logistic regression
analysis for each of above factors and clinical variables,
influence of any given variable was adjusted by the other
factors, and homocysteine levels were assessed as contin-
uous variables. Interestingly, the presence of retinopathy,
serum CRP, and osteocalcin levels that were elevated in
osteoporosis in the univariate analysis (Table 1) were not
closely related to osteoporosis, after adjusting for other co-
variables in this model. However, the relation of diabetic
duration [OR 1.14 (1.02–1.23), p = 0.041] and HbA1c
[OR 1.10 (1.01–1.18), p = 0.044] to osteoporosis was still
significant, respectively (Table 3). Plasma homocysteine
was shown to be associated with osteoporosis after con-
trolling for other risk factors, including thiazolidinediones
use for osteoporosis and known determinants of
J Bone Miner Metab (2014) 32:718–724 721

Table 1 Variables analyzed in diabetic postmenopausal subjects with Table 2 Bone mineral density (BMD) in diabetic subjects with and
and without osteoporosis without osteoporosis
Variables Osteoporosis Osteoporosis p value Variables Osteoporosis Osteoporosis p value
absent present absent (n = 125) present (n = 133)
(n = 125) (n = 133)
Age (years) 59.6 (48.7, 65.3) 63.3 (50.2, 67.7) 0.437
Age (years) 59.6 (48.7, 65.3) 63.3 (50.2, 67.7) 0.437 Lumber spine
Years since 4.9(3.5, 10.6) 5.1 (4.3, 11.2) 1.109 BMD (g/cm2)
menopause (years) L1 1.012 (0.180) 0.843 (0.204) 0.006
Estrogen2 (pmol/l) 34.2 (2.3) 33.9 (1.8) 0.735 L2 1.098 (0.201) 0.826 (0.196) 0.002
FSH (IU/l) 70.2 (5.5) 77.8 (4.6) 0.527 L3 1.008 (0.179) 0.817 (0.220) 0.005
LH (IU/l) 38.1 (2.3) 37.9 (2.5) 0.641 L4 1.013 (0.204) 0.820 (0.189) 0.001
Fasting glucose 10.5 (1.6) 10.9 (3.2) 0.275 Femoral neck 0.625 (0.127) 0.554 (0.117) 0.003
(mmol/l) BMD (g/cm2)
HbA1c (%) 6.7 (1.8) 7.2 (2.7) 0.019 Fracture (%) 3/125 11/133 0.000
Duration of diabetes 6.2 (3.9, 7.6) 6.8 (4.8,10.0) 0.012
(years) All values are the mean (SD) for normally distributed data
Active smoker (%) 6.8 7.2 0.958
Fasting insulin 21.3 (3.1) 20.8 (2.7) 1.165 Table 3 Odds ratios (95 % CI) associated with confounders for
(mIU/l) osteoporosis and homocysteine in multivariable logistic regression
Body mass index 23.7 (2.2) 22.4 (2.7) 0.730 Variables OR (95 % CI) p-value
(kg/m2)
Total calcium intake 783.5 (594.6) 765.2 (403.1) 1.121 Age (years) 0.90 (0.76–1.05) 0.642
(mg/day) Duration of diabetes (years) 1.14 (1.02–1.23) 0.041
Active alcohol 8.2 8.4 0.247 HbA1c (%) 1.10 (1.01–1.18) 0.044
drinker (%) 2
Body mass index (kg/m ) 0.71 (0.30–1.05) 0.937
MET (kcal h/week/ 23.4 (17.5) 22.9 (19.7) 1.120
kg) Parental history of fracture (%) 0.72 (0.42–1.11) 0.605
History of 3.6 3.7 0.972 Presence of retinopathy (%) 1.01 (0.81–1.13) 0.138
myocardial Thiazolidinediones use (%) 0.90 (0.63–1.21) 0.516
infarction (%) Active smoker (%) 0.73 (0.27–1.11) 0.963
History of stroke (%) 1.4 1.5 0.308 Active alcohol drinker (%) 0.78 (0.62–1.05) 0.722
Presence of 5.8 6.0 0.045 Serum folate (nmol/l) 0.92 (0.83–1.15) 0.385
retinopathy (%)
Serum B12 (pmol/l) 0.73 (0.45–1.06) 0.837
Parental history of 19.8 21.2 0.340
Urinary ACR (mg/g) 0.80 (0.69–1.03) 0.534
fracture (%)
Serum creatinine (lmol/l) 0.78 (0.30–1.27) 0.805
Thiazolidinediones 45.6 46.1 0.215
use (%) Serum 25-hydroxyvitamin D (nmol/l) 0.73 (0.31–1.15) 0.991
C-reactive protein 7.13 (1.5) 7.7 (2.7) 0.032 C-reactive protein (CRP) (mg/l) 0.62 (0.22–1.07) 0.835
(CRP) (mg/l) Osteocalcin (lg/l) 0.72 (0.30–1.12) 0.731
Osteocalcin (lg/l) 30.7 (4.8) 33.4 (5.2) 0.038 Metformin use (%) 0.83 (0.42–1.29) 0.808
Serum 79 (10.5) 82 (8.6) 0.727 Homocysteine (lmol/l)a 1.40 (1.02–1.90) 0.036
25-hydroxyvitamin
D (nmol/l) All values are odds ratio (OR), 95 % confidential interval (95 % CI).
Model Nagelkerke R2 0.26
Plasma homocysteine 9.5 (2.0) 10.4 (2.4) 0.001 a
(lmol/l) Homocysteine adjusted for osteoporosis and homocysteine co-
founders: age, duration of diabetes, HbA1c, body mass index, parental
Serum folate (nmol/l) 28.4 (7.7) 29.7 (9.8) 0.115 history of fracture, presence of retinopathy, thiazolidinediones use,
Serum B12 (pmol/l) 386 (163) 378 (184) 1.176 active smokers, active alcohol drinkers, serum vitamins 12 and folate,
Urinary ACR (mg/g) 24.8 (3.3) 26.2 (2.9) 0.213 serum 25-hydroxyvitamin D, renal status (serum creatinine and uri-
nary ACR), CRP, osteocalcin, and metformin use
Serum creatinine 97.2 (10.5) 103.6 (10.1) 0.435
(lmol/l)
Metformin use (%) 75.2 69.7 0.568 homocysteine concentrations, and the relation remained
significant [OR 1.40 (1.02–1.90), p = 0.036] when a
All values are the mean (SD) for normally distributed data and median
(IQR 25, 75 %) for skewed data; active smoker: a smoke per day;
biguanide prescription was added to the same model
active alcohol drinker: a drink per day; MET energy expended from (Table 3). In addition, BMD was closely correlated with
physical activity, ACR Albumin to creatinine ratio homocysteine as a continuous variable, and the correlation

123
722 J Bone Miner Metab (2014) 32:718–724

Table 4 Logistic regression model analysis of per 5 mmol/l increase of plasma homocysteine in relation to postmenopausal osteoporosis
Independent variables OR (95.0 % CI) (p value)
Mode 1 Mode 2 Mode 3

Homocysteine (per 5 mmol/l increase) 1.83 (1.26–2.31) (0.003) 1.69 (1.08–2.29) (0.018) 1.42 (1.07–1.96) (0.027)
Duration of diabetes (per 5 year increase) 2.52 (1.34–3.87) (0.015) 2.30 (1.12–3.24) (0.027) 2.21 (1.11–3.15) (0.031)
HbA1c (per 1 % increase) – 1.24 (1.07–1.73) (0.032) 1.21 (1.03–1.67) (0.045)
Urinary ACR (per 5 mg/g increase) 0.73 (0.58–1.01) (0.597)
Serum creatinine (per 10 lmol/l increase) 0.65 (0.24–1.03) (0.930)
Serum folate (per 10 nmol/l increase) – – 0.93 (0.87–1.02) (0.762)
Serum B12 (per 100 pmol/l increase) – – 0.92 (0.81–1.08) (0.835)

All values are the mean (95.0 % CI). Mode Nagelkerke R21 0.23, R22 0.22, R23 0.25
Mode 1: homocysteine adjusted for duration of diabetes
Mode 2: homocysteine adjusted for duration of diabetes and HbA1c
Mode 3: homocysteine adjusted for duration of diabetes, HbA1c, urinary ACR, serum creatinine, serum folate and vitamin B12

was still significant when age was considered [r = -0.64
(-0.69 to -0.58), p = 0.002]. Furthermore, increases in
plasma homocysteine, when categorized in increments of
5.0 lmol/l, (Table 4) were significantly associated with the
occurrence of osteoporosis [OR 1.42 (1.07–1.96),
p = 0.027], after adjusting for per 5 year increase of
duration of diabetes, per 1 % increase of HbA1c, per
10 nmol/l increase of serum folate, per 100 pmol/l increase
of serum B12, per 10 lmol/l increase of serum creatinine,
and per 5 mg/g increase of urinary ACR. Using ROC
analysis, the optimal cut-off point for the plasma homo-
cysteine level in order to distinguish diabetic patients with
osteoporosis from those without was 10.18 lmol/l, with a
sensitivity of 60.7 %, a specificity of 75.0 %, and the
highest AUC equal to 0.71 [(0.66–0.75), p = 0.002].
Discussion
In this study, diabetic patients with osteoporosis had a
greater level of homocysteine than without. The ele-
vated plasma homocysteine levels were associated with
osteoporosis, which was independent of other possible
risk factors for osteoporosis and determinants of high
homocysteine concentrations in postmenopausal women
with T2DM. In addition, we first found that even a unit
increment of plasma homocysteine appeared to increase
occurrence of osteoporosis to some extent, and a
5.0 lmol/l increase in plasma homocysteine was sig-
nificantly linked to the occurrence of osteoporosis
(independent of other variables). Interestingly, most of
our subjects had vertebral osteoporoses and fractures.
This disagrees with one previous study on non-diabetic
elderly patients ([70 years of age), in which high
homocysteine levels due to poor renal function were
associated with an increased risk of hip fractures [16].
This discrepancy was probably attributed to the subject
selection and the ethnicity difference. In addition, we
found that BMD was closely correlated with homo-
cysteine as a continuous variable, even after adjusting
for age (a probable confounder of both BMD and
homocysteine). Furthermore, our subjects with osteo-
porosis had a mean homocysteine level below 11 lmol/
l, and a mean plasma homocysteine concentration of
10.18 lmol/l appeared to be a potential threshold
demarcating osteoporosis from non-osteoporosis in
T2DM. A larger prospective study is needed to confirm
our findings.
Among determinants of homocysteine concentration,
metformin use was an additional confounder, possibly due
to its interference with B12 absorption from the B12-
intrinsic factor complex. A decrease to subnormal levels of
previously normal serum vitamin B12 levels occurred [17].
We added metformin use (73 % participants on) to the
multivariable analysis. Nevertheless, metformin did not
show a significant effect on serum B12 of patients with or
without osteoporosis from our observation (data not
shown). The concentration of serum folic acid, another
potential factor influencing homocysteine concentration,
was not found to be different among the patients, although
there could be some differences in the serum concentration
of the nutrients between the southerners and the northern-
ers in China, probably due to their dietary preference, as
investigated in another study [18]. In addition to age and
other factors, decreased levels of estrogen may in part
contribute to elevated homocysteine concentration in
postmenopausal women [19]. Furthermore, we observed
that diabetic conditions influenced plasma homocysteine
concentration, which is consistent with findings reported
elsewhere [5].

123
J Bone Miner Metab (2014) 32:718–724
In osteoporotic patients with increased homocysteine
levels, there was a slightly elevated percentage of reti-
nopathy, a chronic diabetic complication, pathologically
due to ischemia or hypoxia within retina. This suggests
that homocysteine-related compromised micro-circulation
may at least in part be involved in bone mass loss or
biomechanical properties [2, 20]. Homocysteine may also
be related to inflammation participating in the osteopo-
rosis as indicated by our finding that CRP increased in
diabetic patients with hyperhomocysteinemia. Homocys-
teine decreases the level of natural antioxidants [21] and
thus increases inflammation and the CRP level. In addi-
tion, hyperhomocysteinemia results in increasing produc-
tion of oxidation products, homo-cysteine thiolactone and
homocysteine mixed disulphides [3, 4], which may
directly or indirectly damage bone. Furthermore, we
observed that in osteoporotic patients with hyperhomo-
cysteinemia, serum osteocalcin levels were elevated,
suggesting that a high level of homocysteine may directly
or indirectly affect the activity of osteoblasts. The exact
mechanism of hyperhomocysteinemia-related osteoporosis
or fracture occurring in our patients remains to be eluci-
dated, since the level of serum 25-hydroxyvitamin D in
our subjects was normal and other biomarkers of bone
turnover were not measured in this study. Interestingly,
the level of osteoporosis-related hyperhomocysteinemia
observed in our patients was lower than another reported
level that was associated with the risk of hip fracture in
men and women [22]. Some the authors suggested that a
high homocysteine concentration may weaken bone by
interfering with collagen cross-linking, thereby increasing
the risk of osteoporotic fracture [22]. In our study, HbA1c
and duration of diabetes were independently associated
with osteoporosis, respectively. Long-standing diabetes
results in increased concentration of advanced glycation
end products and causes abnormal collagen cross-linking
in the bone [11, 23], which may explain a fraction of our
diabetic patients with fractures without osteoporosis
(normal BMD relative to general population). Therefore,
in such circumstance, the cut-off value of homocysteine
for an increased risk for osteoporosis may be lowered to
some extent, or apart from hyperglycemia, homocysteine
is an additional factor linked to osteoporosis in post-
menopausal women. A long term of tight glycemic con-
trol has proven beneficial effects on diabetic chronic
complications [24]. Based on our observation, these
beneficial effects may be extended to diabetes-related
bone mass loss and fractures. In addition, we did not find
the occurrence of osteoporosis increased due to thiazo-
lidinediones use for the treatment of T2DM.
Our findings were observed after other risk factors for
osteoporosis were considered, e.g., active smokers, age, years
since menopause, fasting insulin, the BMI, total calcium
723
intake, alcohol intake, energy expended from physical activity
(MET), and family history. However, genes associated with
the development of osteoporosis and insulin-like growth
factor 1 levels that are thought to be decreased in diabetes and
be associated with osteoporosis were not included in this
investigation [25–27]. In addition, BMD assessments with
dual energy X-ray absorptiometry were not performed at other
sites of the body, such as humeri, ribs, hands, and feet. Fur-
thermore, genetic determinants of homocysteine levels (a
novel polymorphic site in MTHFR (G1793A) that may
influence the homocysteine levels [28]) were not carried out.
Lastly, the study sample was relatively small, and these data
may be influenced by the patient selection (e.g., based on only
one hospital). Future prospective studies are warranted on a
larger scale, which includes more potentially confounding
factors. This may clarify the homocysteine–osteoporosis
relationship observed in this cross-sectional study, thus pro-
viding a potential value for guiding management of osteo-
porosis in T2DM.
In conclusion, our study indicates that the plasma
homocysteine level is related to and is possibly a useful
indicator of osteoporosis in postmenopausal women with
T2DM. If prospective studies could further define an
increase of plasma homocysteine as a causal factor of
osteoporosis in populations with type 2 diabetes, treatment
of elevated homocysteine concentration with supplemen-
tation of folate or methylcobalamin as an alternative way
of mitigating osteoporosis may be justifiable for some
conditions [29–31].
Acknowledgments We would like to thank National Natural Sci-
ence Foundation (81070655) of China and Jiangsu Provincial Natural
Science Foundation (BK2009441) of China for supporting this pro-
ject. This work would not have been possible without the funding.
Conflict of interest All authors have no conflicts of interest.
References
1. Yang W, Lu J, Weng J, Jia W, Ji L, Xiao J, China National
Diabetes and Metabolic Disorders Study Group (2010) Preva-
lence of diabetes among men and women in China. N Engl J Med
362:1090–1101
2. Tyagi N, Vacek TP, Fleming JT, Vacek JC, Tyagi SC (2011)
Hyperhomocysteinemia decreases bone blood flow. Vasc Health
Risk Manag 7:31–35
3. Starkebaum G, Harlan JM (1986) Endothelial cell injury due to
copper-oxidized hydrogen peroxide generation from homocyste-
ine. Clin Invest 77:1370–1376
4. Stamler JS, Osborne JA, Jaraki O, Rabbani LE, Mullins M, Singel
D (1993) Adverse vascular effects of homocysteine are modu-
lated by endothelial-derived relaxing factor and related oxides of
nitrogen. Clin Invest 91:308–318
5. Abdella NA, Mojiminiyi OA, Akanji AO, Moussa MA (2002)
Associations of plasma homocysteine concentration in subjects
with type 2 diabetes mellitus. Acta Diabetol 39:183–190
123
724
6. Jacques PF, Bostom AG, Wilson PW, Rich S, Rosenberg IH,
Selhub J (2001) Determinants of plasma total homocysteine
concentration in the Framingham Offspring cohort. Am J Clin
Nutr 73:613–621
7. Chico A, Perez A, Cordoba A, Arcelus R, Carreras G, de Leiva A
(1998) Plasma homocysteine is related to albumin excretion rate in
patients with diabetes mellitus: a new link between diabetic
nephropathy and cardiovascular disease? Diabetologia 41:684–693
8. Mudd SH, Pool JR (1975) Labile methyl balance for normal
humans on various dietary regimens. Metabolism 24:721–733
9. Baines M, Kredan MB, Usher J, Davison A, Higgins G, Taylor W
(2007) The association of homocysteine and its determinants:
MTHFR genotype, folate, vitamin B12 and vitamin B6 with bone
mineral density in postmenopausal British women. Bone
40:730–736
10. Ganji V, Kafai MR (2009) Demographic, lifestyle, and health
characteristics and serum B vitamin status are determinants of
plasma total homocysteine concentration in the post-folic acid
fortification period, 1999–2004. J Nutr 139:345–352
11. Isidro ML, Ruano B (2010) Bone disease in diabetes. Curr Dia-
betes Rev 6:144–155
12. Viégas M, Costa C, Lopes A, Griz L, Medeiro MA, Bandeira F
(2011) Prevalence of osteoporosis and vertebral fractures in
postmenopausal women with type 2 diabetes mellitus and their
relationship with duration of the disease and chronic complica-
tions. J Diabetes Complicat 25:216–221
13. WHO: World Health Organization (2006) Diabetes mellitus:
report of a WHO Study Group. World Health Org, Geneva
14. WHO: World Health Organization (2003) Prevention and man-
agement of osteoporosis. World Health Organ Tech Rep Ser
921:1–164
15. Chinese Nutrition Society (2008) Chinese dietary guidelines.
Tibet People’s Publishing Company Press, China
16. Leboff MS, Narweker R, LaCroix A, Wu L, Jackson R, Lee J
(2009) Homocysteine levels and risk of hip fracture in post-
menopausal women. J Clin Endocrinol Metab 94:1207–1213
17. Refsum H, Smith AD, Ueland PM, Nexo E, Clarke R, McPartlin J
(2004) Facts and recommendations about total homocysteine
determinations: an expert opinion. Clin Chem 50:1482–1483
18. Hao L, Ma J, Zhu J, Stampfer MJ, Tian Y, Willett WC (2007)
Vitamin B-12 deficiency is prevalent in 35- to 64-year-old Chi-
nese adults. J Nutr 137:1278–1285
19. Bonassi MR, Chada BE, Eduardo FC, Marcelo LE, Rodrigues
DLG (2007) Effects of estrogen and estrogen-progestogen
123
J Bone Miner Metab (2014) 32:718–724
therapy on homocysteine levels and their correlation with carotid
vascular resistance. Gynecol Endocrinol 23:619–624
20. Viégas M, Costa C, Lopes A, Griz L, Medeiro MA, Bandeira F
(2011) Prevalence of osteoporosis and vertebral fractures in
postmenopausal women with type 2 diabetes mellitus and their
relationship with duration of the disease and chronic complica-
tions. J Diabetes Complicat 25:216–221
21. Kuyumcu ME, Yesil Y, Oztürk ZA, Cınar E, Kızılarslanoglu C,
Halil M (2012) The association between homocysteine (hcy) and
serum natural antioxidants in elderly bone mineral densitometry
(BMD). Arch Gerontol Geriatr 55:739–743
22. Robert RM, Paul FJ, Jacob S, Katherine LT, Elizabeth JS, Kerry
EB (2004) Homocysteine as a predictive factor for hip fracture in
older persons. N Engl J Med 350:2042–2049
23. Yamamoto M, Yamaguchi T, Yamauchi M, Sugimoto T (2009)
Low serum level of the endogenous secretory receptor for
advanced glycation end products (esRAGE) is a risk factor for
prevalent vertebral fractures independent of bone mineral density
in patients with type 2 diabetes. Diabetes Care 32:2263–2268
24. Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA (2008)
10-year follow-up of intensive glucose control in type 2 diabetes.
N Eng J Med 359:1577–1589
25. Hong XM, Hsu YH, Henry TD, Tang GF, Liu X (2007) Asso-
ciation of the methylenetetrahydrofolate reductase C677T poly-
morphism and fracture risk in Chinese postmenopausal women.
Bone 40:737–742
26. Li JB, Wang CY, Chen JW, Feng ZQ, Ma HT (2004) Expression
of liver IGF-1 gene and its serum IGF-1 level in rats with dia-
betes. World J Gastroenterol 10:255–259
27. Raisz LG (2005) Pathogenesis of osteoporosis concepts, conflicts,
and prospects. J Clin Invest 115:3318–3325
28. Mao R, Fan Y, Chen F, Fu S (2008) Genetic polymorphism of
MTHFR G1793A in Chinese populations. Eur J Epidemiol
23:363–368
29. Chan E, Fonseca VA (2003) Management of hyperhomocy-
steinemia. Metab Syndr Relat Disord 1:159–170
30. House AA, Eliasziw M, Cattran DC, Churchill DN, Oliver MJ,
Fine A (2010) Effect of B-vitamin therapy on progression of
diabetic nephropathy: a randomized controlled trial. JAMA
303:1603–1609
31. Li JB, Wang CY, Chen JW, Li XL, Feng ZQ, Ma HT (2010) The
preventive efficacy of methylcobalamin on rat peripheral neu-
ropathy influenced by diabetes via neural IGF-1 levels. Nutr
Neurosci 13:79–86
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.