EJINME-03078; No of Pages 6

European Journal of Internal Medicine xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

European Journal of Internal Medicine

journal homepage: www.elsevier.com/locate/ejim

Original Article

The association between mean platelet volume and cardiovascular risk factors
Nakarin Sansanayudh a,⁎, Dittapol Muntham b, Sukit Yamwong c, Piyamitr Sritara c,
Tawatchai Akrawichien d, Ammarin Thakkinstian e
a
Cardiology Unit, Department of Internal Medicine, Phramongkutklao Hospital, Bangkok, Thailand
b
Section for Clinical Epidemiology and Biostatistics, Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand
c
Division of Cardiology, Department of Medicine, Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Thailand
d
Electricity Generating Authority of Thailand, Thailand
e
Section for Clinical Epidemiology and Biostatistics, Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand

a r t i c l e i n f o a b s t r a c t

Article history: Background: Mean platelet volume (MPV) correlates with platelet activation and has recently emerged as a
Received 12 July 2015 potential marker of cardiovascular diseases. Previous publications also suggest possible association between
Received in revised form 21 October 2015 MPV and some cardiovascular risk factors but the evidences are still conflicting and inconclusive.
Accepted 30 November 2015 Objective: To study the association between MPV and cardiovascular risk factors.
Available online xxxx
Methods: This is a cross-sectional study using data from the second survey of the Electricity Generating
Authority of Thailand (EGAT) cohort. All participants of the survey who had results of MPV were included.
Keywords:
Mean platelet volume
Exclusion criteria included subjects with known hematologic disease or subjects with hematocrit b 30% or plate-
Cardiovascular risk factors let count b 140,000/mm3. The details of cardiovascular risk factors were documented and the association
Hypertension between MPV and risk factors was analyzed using fractional polynomial regression analysis.
Diabetes Results: There were 2727 subjects with MPV results. After excluding those who had hematologic disease, 2642
Metabolic syndrome subjects were included for analysis. Univariate analysis revealed that gender, diabetes, serum triglyceride,
Smoking hypertension, and prehypertension were associated with MPV. Hematocrit, platelet count and fasting plasma
glucose were inversely correlated with MPV. After adjusting with other variables, the risk factors that remained
significantly associated with MPV included female gender, diabetes, metabolic syndrome, serum triglyceride,
hypertension, and prehypertension. Platelet count and hematocrit were found to have significant inverse
correlation with MPV.
Conclusion: After adjusting for other cardiovascular risk factors, the independent factors remain associated with
MPV included female gender, diabetes, metabolic syndrome, serum triglyceride, hypertension and
prehypertension. MPV has significant, but inverse association with platelet count and hematocrit.
© 2015 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.

1. Background compared to patients with low MPV. The pathophysiologic mechanisms

Mean platelet volume (MPV) is a quantitative measurement of
average size of platelets [1]. Larger platelets are metabolically and
enzymatically more active and have higher homeostasis property than
smaller platelets [2]. MPV correlates with platelet function and
activation and has recently emerged as a potential marker of cardiovas-
cular diseases.
Pooling results from previous publications suggested the association
between MPV and coronary artery disease (CAD) [3], i.e., patients with
CAD had significantly larger MPV compared to patients without CAD,
and patients with high MPV also had higher chance of having CAD
⁎ Corresponding author. Tel.: +66 23547600; fax: +66 26444755.
E-mail addresses: dr_nakarin@hotmail.com (N. Sansanayudh),
dittapol.mun@mahidol.ac.th (D. Muntham), sukit.yam@mahidol.ac.th (S. Yamwong),
piyamitr.sri@mahidol.ac.th (P. Sritara), Tawatchai.a@egat.co.th (T. Akrawichien),
ammarin.tha@mahidol.ac.th (A. Thakkinstian).
of the association can be explained by the direct effects of higher platelet
activity in patients with high MPV. In addition, previous publications
also found the associations between MPV and risk factors of cardiovas-
cular disease (CVD), in which diabetes and metabolic syndromes were
the most common risk factors that had been studied, but there was
less evidence on other individual risks such as age, gender, smoking,
dyslipidaemia, body mass index (BMI), and waist circumference.
Among those evidences, effects of MPV on these cardiovascular risk
factors were also inconclusive and the findings were inconsistent,
which might be due to limitation of previous studies. Most studies had
only small sample sizes and thus less power to detect the association
which made them unable to adjust for the effects of other risk factors.
Few large studies have been conducted in specific group of patients
(e.g. obese, diabetes, pre-diabetes, etc.) and the results might not be
able to be applied to a more general population.
We therefore conducted this cross-sectional study based on a cohort
study of healthy volunteers which aimed to assess the association

http://dx.doi.org/10.1016/j.ejim.2015.11.028
0953-6205/© 2015 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.

Please cite this article as: Sansanayudh N, et al, The association between mean platelet volume and cardiovascular risk factors, Eur J Intern Med
(2015), http://dx.doi.org/10.1016/j.ejim.2015.11.028
2 N. Sansanayudh et al. / European Journal of Internal Medicine xxx (2015) xxx–xxx

between MPV and all traditional cardiovascular risk factors. Magnitude Institutional Review Board approval was obtained from the Faculty
of MPV effects on these risks were then compared. of Medicine Ramathibodi Hospital, prior to the conduct of the study.
All participants gave written informed consent before enrollment.
2. Methods
2.2. Statistical analysis
This is a cross-sectional study using data from the second survey of
the Electricity Generating Authority of Thailand (EGAT) cohort that Data were described using mean ± standard deviation (SD) or
was conducted in 1997. The EGAT cohort is a prospective, ongoing median and range where appropriate for continuous data, and using
cohort of middle-income workers. The cohort was started in 1985 by frequency and percentage for categorical data. Mean MPVs were com-
recruiting the employees of EGAT and had the survey follow up in pared between risk factor groups using t-test or one-way analysis of
1997, 2002, 2007 and 2012. The cohort primarily aims to study risk variance, where appropriate.
factors and incidence of cardiovascular disease. A simple linear regression was performed by regress MPV on each
Participants in the cohort were included in this study if they had individual categorical risk factor. The MPV was transformed to natural
complete blood count (CBC) results. The exclusion criteria included logarithm scale to meet linear regression assumption of approximated
known history of hematologic disease, hematocrit b30%, and platelet normality of residuals. For those continuous variables (i.e., age, BMI,
count b140,000/mm3. total cholesterol, triglyceride, hematocrit, and platelet count), a
During the survey in 1997, data from all participants were collected fractional polynomial regression was applied because its relationship
using self-administered questionnaires, which were consisted of with log (MPV) was non-linear. Each continuous variable was fitted
general demographic data, risk/preventive behaviors, underlying using a set of power of −2, −1, −0.5, 0, 0.5, 1, 2, 3; and the best fitting
disease, family history of illness, and use of medications. In addition, degree of 2 was selected. For instance, degrees of − 2 and − 2 with a
physical examination, electrocardiography, chest radiography, and scale of 10-years for age; 1 and 2 with scale 10 for BMI; 2 and 2 with
blood tests were also performed at the survey camp sites. Anthropomet- scale 10 for hematocrit; 1 and 3 with scale 100 for platelet count, 0.5
ric measurements were measured by trained nurses using a standard and 2 with scale 100 for cholesterol; and − 1 and 3 with scale 1000
protocol [4]. Physical examination was performed by cardiologists. for triglyceride. Those risk factors whose p values were less than 0.10
The completeness of each procedure, including the questionnaire, was plus traditional cardiovascular risk factors (i.e., age, BMI, diabetes,
checked at the end of the survey process for every participant. All hypertension, total cholesterol, and triglyceride) regardless of their p
participants were contacted and informed in advance to fast overnight values from the univariate analysis were simultaneously considered in
for 12 h before receiving blood examinations including CBC, fasting a multivariate fractional polynomial model. Normality of residual was
plasma glucose (FPG), and lipid profile. The blood samples were then checked. All analyses were performed using STATA version 13
transferred to a central laboratory for analysis. All samples were (Stata Corp., College Station, Texas, USA). A p-value of less than 0.05
processed approximately 6–8 h after venupuncture. Previous publica- was considered statistically significant.
tions found a substantial difference in MPV values measured by differ-
ent analyzers [5,6]. Therefore, all blood samples for CBC in our study
3. Results
were collected in ethylenediaminetetraacetic acid (EDTA) tubes and
were analyzed using a light scattering method by a single automated
Among 2967 participants of EGAT 1997, 240 participants did not
machine (Technicon H-1).
perform blood collections for laboratory analysis, thus they had no
result of MPV and other laboratory parameters. These participants
2.1. Risk factors and measurements were excluded leaving 2727 participants included in our study. Of
these, 85 subjects were excluded due to hematocrit b 30%, or platelet
The details of cardiovascular risk factors (i.e., age, gender, BMI, waist count b140,000/mm3, leaving 2642 participants included for analysis.
circumference, waist-to-hip ratio, smoking, hypertension, diabetes, and
metabolic syndrome) were collected at the survey camp. Table 1
Diabetes was diagnosed if the participants had history of diabetes, Baseline characteristics of the studied participants.
had fasting plasma glucose ≥ 126 mg/dL, or taking anti-diabetic
Characteristics N = 2642
medications. Pre-diabetes was defined as participants who were not di-
abetes but had fasting plasma glucose between 100 and 125 mg/dL. Age, years, mean (SD) 54.34 (4.90)
Sex, number (%)
Hypertension was diagnosed if the participants had history of Male 1984 (75.1)
hypertension, had systolic blood pressure (SBP) ≥ 140 mmHg or diastol- Female 658 (24.9)
ic blood pressure (DBP) ≥ 90 mmHg, or taking antihypertensive medica- Smoking, number (%) 560 (21.60)
tions. Blood pressure (BP) was measured using automated BP machines Diabetes, number (%) 298 (11.28)
Hypertension, number (%) 1267 (47.96)
(Omron ®) which were calibrated before the survey. BP measurement
Metabolic syndrome, number (%) 892 (33.76)
protocol followed the recommended standard. After resting for at Body weight, kg, mean (SD) 65.65 (10.46)
least 5 min, BP was measured at least twice in each participant and Height, cm, mean (SD) 162.70 (7.25)
the average between readings was used. BMI, kg/m2, mean (SD) 24.76 (3.35)
Metabolic syndrome was diagnosed using modified NCEP Waist circumference, cm 88.00 (9.59)
Systolic blood pressure, mmHg, mean (SD) 135.76 (21.39)
criteria [7]. The participants were classified as having metabolic syn- Diastolic blood pressure, mmHg, mean (SD) 81.41 (13.11)
drome if they had at least three of the following criteria: 1) waist Total cholesterol, mg/dL, mean (SD) 239.51 (41.10)
circumference ≥ 80 cm in women or ≥ 90 cm in men; 2) FPG ≥ 100 Triglyceride, mg/dL, mean (SD) 135 (90)
mg/dL; 3) triglyceride ≥ 150 mg/dL; 4) HDL b 50 mg/dL in women or HDL, mg/dL, mean (SD) 52.72 (11.10)
LDL, mg/dL, mean (SD) 155.85 (39.98)
HDL b 40 mg/dL in men and 5) SPB ≥ 130 or DBP ≥ 85 mmHg.
Fasting blood glucose, mg/dL), mean (SD) 95.71 (28.93)
Age in year was calculated by subtracting survey date with date Creatinine, mg/dL, mean (SD) 1.15 (0.37)
of birth. BMI was calculated as weight (kg)/height (m2), where Hematocrit, %, mean (SD) 44.88 (4.28)
weight and height were measured using the standard techniques. White blood cell, 103 mg/dL, mean (SD) 7.16 (1.82)
Smoking was collected by self-reporting, which categorized as smoking Platelet count, 109/L, mean (SD) 255.44 (59.30)
MPV, fL, mean (SD) 7.24 (0.81)
(i.e., current plus ex-smoke) and non-smoking.

Please cite this article as: Sansanayudh N, et al, The association between mean platelet volume and cardiovascular risk factors, Eur J Intern Med
(2015), http://dx.doi.org/10.1016/j.ejim.2015.11.028
 N. Sansanayudh et al. / European Journal of Internal Medicine xxx (2015) xxx–xxx 3

The baseline data of the studied participants are shown in Table 1. MPV (7.26 ± 0.83 fL) compared to subjects without metabolic
The mean age and BMI of the studied participants were 54.34 ± syndrome (7.23 ± 0.79 fL, p = 0.412).
4.9 years and 24.76 ± 3.35 kg/m2, respectively. Approximately three Fractional polynomial was performed by assessing the relationship
quarters were male and a half were smokers/ex-smokers. Prevalence between individual risk factors and log of MPV, see Table 3. Seven
of diabetes, metabolic syndrome, and hypertension were 11.28%, variables that might be associated with log (MPV) were identified
33.76% and 47.96%, respectively. Mean MPV was 7.24 ± 0.81 fL. from univariate analysis which included gender, smoking status, diabe-
The associations between MPV and risk factors were described and tes, cholesterol, triglyceride, hematocrit, and platelet count. Therefore,
compared, see Table 2. This suggested trends of associations between these variables plus 3 traditional risks (i.e., age, BMI, and hypertension)
MPV and gender, smoking, and diabetes, but no trends for other risk were simultaneously included in multivariate fractional polynomial
factors. MPV was higher in female than in male (7.40 ± 0.82 fL vs. regression model. The final model suggested 4 cardiovascular risk
7.19 ± 0.79 fL, p b 0.001), and higher in non-smokers than in smokers factors (i.e., gender, diabetes, hypertension, and serum triglyceride)
(7.28 ± 0.82 fL vs. 7.19 ± 0.80 fL, p = 0.008). In addition, there was a plus 2 hematologic factors (i.e., hematocrit and platelet count) were
trend towards higher MPV in diabetes and impaired fasting glucose independently associated with MPV (see Table 3), whereas age, BMI,
compared to normoglycemia (7.32 ± 0.79 fL vs. 7.27 ± 0.94 fL vs. smoking, and serum cholesterol were not significantly associated with
7.22 ± 0.78 fL). Subjects with metabolic syndrome had slightly higher MPV after adjusting for other risk factors.
We found association between female gender and MPV. Female
Table 2 participants had 0.034 higher log (MPV) (i.e., 1.034 (95% CI: 1.022,
Mean MPV by cardiovascular risk factor groups. 1.046) fL) higher MPV than males. Participants with diabetes had 0.02
Risk factors N Mean MPV (SD) P value larger log(MPV) (i.e., 1.02 (95% CI: 1.008, 1.033) fL) higher MPV than
non-diabetic participants. Likewise, prehypertension and hypertensive
Age
b50 512 7.28 (0.82) 0.091
participants had 0.011 and 0.017 larger log (MPV) (i.e., 1.011 (95% CI:
50–51.99 429 7.18 (0.74) 1.00, 1.022) and 1.017 (95% CI: 1.006, 1.028) fL greater MPV when
52–54.99 523 7.29 (0.89) compared to normotensive subjects. Our study also found the
55–58.99 725 7.23 (0.78) significant association between serum triglyceride and MPV. Every
≥60 453 7.20 (0.79)
100 mg/dL increase in serum triglyceride is associated with 1.043
Sex (95% CI: 1.007, 1.079) fL increase in MPV.
Male 1984 7.19 (0.79) b0.001 Furthermore, the MPV also inversely associated with hematocrit and
Female 658 7.40 (0.82) platelet count. Every 10% increase in hematocrit is associated with 0.975
Smoking (95% CI: 0.961, 0.989) fL decrease in MPV. Likewise, every 100,000/mm3
Non-smoker 1287 7.28 (0.81) 0.024
Ex-smoker 746 7.19(0.78)
increase in platelet count results in 0.898 (95% CI: 0.884, 0.911) fL
Current smoker 560 7.21 (0.82) reduction of MPV.
The effect of MPV on metabolic syndrome was tested in multivariate
Diabetes fractional polynomial regression by including metabolic syndrome
Diabetes 298 7.32 (0.79) 0.107
without considering diabetes, hypertension, and BMI. After adjusting
Pre-diabetes 382 7.27 (0.94)
Non-diabetes 1939 7.22 (0.78) for confounders, participants with metabolic syndrome had higher
log(MPV) approximately 0.0128, which was equivalent to 1.013 (95%
Hypertension CI: 1.005, 1.021) fL larger MPV than participants without metabolic
Hypertension 1267 7.25 (0.78) 0.775 syndrome, see supplement Table 1.
Pre-hypertension 817 7.24 (0.85)
No hypertension 541 7.22 (0.78)
Male 1984 7.19 (0.79) b0.001 4. Discussion

Metabolic syndrome We have identified the cardiovascular risk factors that were inde-
Yes 892 7.26 (0.83) 0.412
pendently associated with MPV, which included sex, prehypertension,
No 1611 7.23 (0.79)
hypertension, diabetes, serum triglyceride, and metabolic syndrome.
BMI Our study also suggested an inverse association between MPV and
Underweight 55 7.21 (0.89) 0.895 hematologic factors including hematocrit and platelet count. However,
Normal 732 7.22 (0.78) we found no effects of other cardiovascular risk factors such as age,
Overweight 693 7.24 (0.78)
Obese 1137 7.25 (0.84)
BMI, smoking status, and hypercholesterolemia on MPV.
Diabetes is an important and well established risk factor for
Alcohol cardiovascular disease. Higher platelet reactivity has been one of the
Ex-alcohol/Current alcohol 1225 7.21 (0.80) 0.123 fundamental determinants for increased risk and worse outcome in
Non-alcohol 1357 7.26 (0.81)
diabetic patients [8] which might be explained by altered platelet
Total cholesterol morphology, metabolism and function [9]. This was supported by the
b200 382 7.28 (0.82) 0.290 results of our study that found MPV to be 1.020 fL larger in diabetes
≥200/med 2245 7.23 (0.80) than non-diabetic subjects. Our finding was consistent with most of
previous studies [10–12].
LDL
b130 588 7.24 (0.80) 0.911
Our study also detected association between metabolic syndrome
≥ 130/med 1984 7.23 (0.81) and MPV, in which its magnitude of association was as close as effect
of diabetes, i.e., 1.0129 fL higher in metabolic syndrome than non-
Triglyceride metabolic syndrome. This finding is consistent with findings from
b150 1316 7.25 (0.79) 0.448
some previous studies [13,14] but not for other studies [11,15,16].
≥150/med 1308 7.22 (0.81)
The possible mechanisms of association between MPV and
HDL metabolic syndrome remain to be elucidated. Metabolic syndrome is a
b40/med 589 7.28 (0.79) 0.385 syndrome characterized by insulin resistance. Two small studies report-
40–59 1377 7.22 (0.79) ed an association between MPV and insulin resistance measured by
≥60 661 7.23 (0.84)
HOMA-IR [17,18]. Our study found the trend of MPV increment with

Please cite this article as: Sansanayudh N, et al, The association between mean platelet volume and cardiovascular risk factors, Eur J Intern Med
(2015), http://dx.doi.org/10.1016/j.ejim.2015.11.028
4 N. Sansanayudh et al. / European Journal of Internal Medicine xxx (2015) xxx–xxx

numbers of MS criteria (data not shown). An increasing number of MS
features were associated with a more insulin resistance state. This
finding provides further support to insulin resistance as an important
mechanism linking MPV and metabolic syndrome.
However, in our study, we did not find the association between
waist circumference or BMI and MPV. This might indicate that the
association between MS and MPV might not be exerted only through
the insulin resistance, but might involve diabetes and hypertension,
the two components of MS which were correlated with MPV in our
analysis.
Our study suggested similar effects of hypertension on MPV as
diabetes, i.e., subjects with hypertension had approximately 1.0169 fL
larger MPV than normotensive subjects. This corresponded with previ-
ous publications [19–21].
In addition, we also could identify the effect of pre-hypertension on
MPV, although less is known about prehypertension effect. This was

Table 3
Association between MPV and cardiovascular risk factors: univariate and multivariate fractional polynomial regression models.

Risk factors Univariate analysis Multivariate analysis

Coefficient 95% CI P value Coefficient 95% CI P value

Sex
Female 0.0302 0.0208, 0.0396 b0.001 0.0336 0.0218 0.0453 b0.001

Age
Age1 −4.0159 −23.0479, 15.0161 0.679 −0.2487 −17.0375 16.54 0.977
Age2 3.7372 −12.2498, 19.7243 0.647 1.0252 −13.0951 15.1456 0.887

BMI
BMI1 −0.0277 −0.1335, 0.0780 0.607 0.0541 −0.0452 0.1534 0.285
BMI2 0.0079 −0.0125, 0.0282 0.448 −0.007 −0.0261 0.0121 0.473

Smoking
Ex-smoke −0.0131 −0.0229, −0.0034 0.008 0.0004 −0.0091 0.0099 0.932
Current-smoke −0.0098 −0.0206, 0.0009 0.072 0.0106 0.0002 0.0209 0.046

Diabetes
DM 0.0134 0.0003, 0.0266 0.045 0.0186 0.0067 0.0306 0.002
Pre DM 0.0053 −0.0065, 0.0171 0.379 −0.0016 −0.0122 0.0089 0.762

Hypertension
HT 0.0041 −0.0067, 0.0150 0.456 0.0182 0.0079 0.0285 0.001
Pre HT 0.0024 −0.0093, 0.0141 0.690 0.0105 0.0001 0.021 0.048

Total cholesterol (TC)

TC1 −0.1833 −0.3260, −0.0406 0.012 0.0231 −0.1118 0.1581 0.737
TC2 0.0116 0.0028, 0.0204 0.009 −0.0013 −0.0097 0.0071 0.760

Triglyceride (TG)
TG1 0.0012 0.00004, 0.0024 0.043 0.0001 −0.0011 0.0012 0.887
TG2 0.0464 0.0095, 0.0834 0.014 0.0431 0.0094 0.0767 0.012

Hematocrit (Hct)
Hct1 −0.0577 −0.0888, −0.0266 b0.001 −0.0411 −0.0697 −0.0124 0.005
Hct2 0.0273 0.0118, 0.0428 0.001 0.0185 0.0043 0.0328 0.011

Platelet count (Plt)

Plt1 −0.1025 −0.1185, −0.0864 b0.001 −0.1102 −0.1254 −0.0951
Plt2 0.0015 0.0010, 0.0021 b0.001 0.0013 0.0008 0.0018
Waist 0.0597 −0.2261, 0.3455 0.682
Waist1 0.0240 −0.0927, 0.1407 0.687
Waist2 0.0357 −0.1338, 0.2053 0.679

HDL
HDL1 −0.0003 −0.0011, 0.0005 0.438
HDL2 0.0001 −0.0002, 0.0005 0.428

LDL
LDL1 −0.0051 −0.0102, 0.00003 0.051
LDL2 0.0058 0.0010, 0.0106 0.019

Metabolic syndrome
MS ≥ 3 0.0037 −0.0052, 0.0125 0.415

Age1, (Age / 10)−2 − 0.0339; Age2, (Age / 10)−2 × ln(Age / 10) − 0. 0573.
BMI1, (BMI / 10) − 2.4757; BMI2, (BMI / 10)2 − 6.1292.
TC, total cholesterol, TC1, (TC97 / 100)0.5 − 1.5476; TC2, (TC97 / 100)2 − 5.7367.
HCT, Hematocrit, Hct1, (Hct / 10)2 − 20.1445; Hct2, (Hct / 10)2 × ln(Hct / 10) − 30.2463.
Plt, Platelet count, Plt1, (Plt / 100) − 2.5543; Plt2, (Plt / 100)3 − 16.6665.
WBC1, (WBC)−2 − 1.9482; WBC2, (WBC)−2 × ln(WBC) + 0.650.
Waist1, (Waist / 100)−2 − 1.2913; Waist2, (Waist / 100)−2 × ln(Waist / 100) + 0.1651.
TG, Triglyceride, TG1, (TG / 1000)−1 − 6.1655; TG2, (TG / 1000)−3 − 0.0043.
HDL, High-density lipoprotein, HDL1, (HDL / 10)3 − 146.5798; HDL2, (HDL / 10)3 × ln(HDL / 10) − 243.6924.
LDL, Low-density lipoprotein, LDL1, (LDL / 100)3 − 3.7856; LDL2, (LDL / 100)3 × ln(LDL / 100) − 1.6798.

Please cite this article as: Sansanayudh N, et al, The association between mean platelet volume and cardiovascular risk factors, Eur J Intern Med
(2015), http://dx.doi.org/10.1016/j.ejim.2015.11.028
 N. Sansanayudh et al. / European Journal of Internal Medicine xxx (2015) xxx–xxx
inconsistent with previous findings from two studies [22] [23], where
prehypertension effect was disappeared after adjusting for confounders.
However, both studies had small number of subjects (n = 122 [22] and
62 [23]), thus might be lack of power in a multivariate analysis.
We found no association between MPV and age which was in
agreement with some previous studies [24,25]. However, one large
study reported the association between age and MPV only in men, but
not in women [26]. The large study of MPV and age, adjusting for
other variables and including a wide range of ages, should be encour-
aged to confirm the association between MPV and age.
Gender influence on MPV is another topic for debate. Shimodaira
et al., reported larger MPV in male (10.05 ± 0.73 vs. 9.94 ± 0.73, p b
0.001) [27], whereas most previous studies found no difference in
MPV between male and female [25,28,29]. Conversely in our study,
MPV was significantly larger in female even after adjusting for other
variables.
One possible explanation might be the racial difference in
population studied. Maluf et al., reported substantial difference in
MPV among races [25]. The different studies of association between
MPV and risk factors such as sex and age were conducted in different
ethnic groups. The future study of reference interval and influence of
other risk factors on MPV in different ethnicity is required.
Our study found no correlation between BMI and MPV. This finding
was in contrast with previous studies [26,30] which found positive
effects of BMI on MPV; and decreasing of MPV when weight decreased.
The association between MPV and smoking was detected in univar-
iate analysis but disappeared after adjusting with other variables. De-
spite a report of an increase in MPV over time in passive smoker who
were exposed to moderate to high intensity of cigarette smoke in one
small experimental study [31], most evidence form observation studies
[32,33] showed no association between smoking status and MPV which
is consistent with the results from our study.
Few previous studies reported the association between MPV and
dyslipidaemia, i.e., MPV was significantly larger in patients with hyper-
cholesterolemia [34,35] and low HDL [36]. In terms of changes in MPV
after lipid management, the results were conflicting. While cholesterol
reducing diet resulted in an increase in MPV [37], LDL apheresis as
well as atorvastatin and rosuvastatin has been reported to significantly
lower MPV [34,38,39] but the degree of change in MPV after receiving
these treatments was not correlated with changes in cholesterol level
[34,38,39].
However, all previous studies were small in sample sizes and did not
control for other variables. At present, our study is the largest study of
MPV and dyslipidaemia and we did not observe significant association
between MPV and serum cholesterol, LDL or HDL. Interestingly, we
found the unexpected association between MPV and serum triglyceride
which was significant even after adjusting with other cardiovascular
risk factors. This finding should further be explored and confirmed in fu-
ture research.
In steady state, there is a usual inverse correlation between platelet
count and volume in order to maintain the constant platelet mass
(platelet count × platelet volume). In our study, we observed the
inverse correlation between platelet count and MPV which is in agree-
ment with previous publications [24,40].
There are some strengths in our study. First, the sample size is ade-
quate for detecting the significance of the association between MPV
and cardiovascular risk factors and also sufficient for multivariate
adjustment. Second, using the cross-sectional data from a cardiovas-
cular disease cohort, all cardiovascular risk factors were systemati-
cally collected and computerized, and the classifications of each
risk factor were very solid and verified by investigator team. Third,
all important cardiovascular risk factors were considered and evalu-
ated, not just focusing on any single factor but also simultaneous
adjustment was performed in order to identify the independent
effect of each risk factor on MPV. Fourth, the study was conducted
in healthy volunteers and made the results more generalizable
Please cite this article as: Sansanayudh N, et al, The association between mean platelet volume and cardiovascular risk factors, Eur J Intern Med
(2015), http://dx.doi.org/10.1016/j.ejim.2015.11.028
5
compared to the studies conducted in specific groups of patients
(e.g. obese, pre-hypertension, diabetes, etc.).
Our study also had some limitations. First, due to the nature of the
on-camp site survey at the participants' working place, MPV could not
be analyzed immediately. The blood samples were taken in the morning
and all specimens were transported back to the central laboratory
center after lunch time when the survey finished each day. MPV was
reported to increase over time in EDTA [41,42]. However, the increase
in MPV occurs mostly within 60 min and little change occurs after 2 h.
The MPV was analyzed using a single automated machine at 6–8 h
after venupuncture in all blood samples in our study with little varia-
tion. Therefore, the change in MPV over time should have very little
effect on the internal validity of our findings. Second, the study was a
cross-sectional design, therefore, the causal relationship between
cardiovascular risk factors and MPV could not be identified. Third, we
did not measure HOMA-IR thus a direct relationship between insulin
sensitivity and MPV could not be demonstrated. Fourth, our study
population, although large in numbers, include participants from a
middle-class workers cohort. It is possible that this cohort population
may differ from the general population regarding their lifestyle and in-
cidence of cardiovascular risk factors.
5. Conclusion
After adjusting for other cardiovascular risk factors, the independent
factors remain associated with MPV included gender, diabetes,
metabolic syndrome, serum triglyceride, prehypertension and hyper-
tension. MPV has significant, but inverse association with platelet
count and hematocrit level.
Conflict of interest
The authors state that they have no conflicts of interest.
Acknowledgements
We would like to thank Nisakrorn Thongmung for data retrieval and
completeness of data, which include re-checking with source docu-
ments. This study was supported by the project for Higher Education
Research Promotion and National Research University Development,
Office of the Higher Education Commission, the Thailand Research
Fund and Health Systems Research Institute.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.ejim.2015.11.028.
References
[1] Bath PM, Butterworth RJ. Platelet size: measurement, physiology and vascular
disease. Blood Coagul Fibrinolysis 1996;7(2):157–61.
[2] Martin JF, et al. The biological significance of platelet volume: its relationship to
bleeding time, platelet thromboxane B2 production and megakaryocyte nuclear
DNA concentration. Thromb Res 1983;32(5):443–60.
[3] Sansanayudh N, et al. Mean platelet volume and coronary artery disease: a system-
atic review and meta-analysis. Int J Cardiol 2014;175(3):433–40.
[4] Vathesatogkit P, et al. Cohort profile: the electricity generating authority of Thailand
study. Int J Epidemiol 2012;41(2):359–65.
[5] Van den Bossche J, et al. Reference intervals for a complete blood count determined
on different automated haematology analysers: Abx Pentra 120 Retic, Coulter Gen-S,
Sysmex SE 9500, Abbott Cell Dyn 4000 and Bayer Advia 120. Clin Chem Lab Med
2002;40(1):69–73.
[6] Qiao R, et al. Complete blood count reference intervals and age- and sex-related
trends of North China Han population. Clin Chem Lab Med 2014;52(7):1025–32.
[7] Health, N.H.E.S.O.M.o.P. Report of 4th national health examination survey in Thailand
2008–2009. National Health Examination Survey Office: Nonthaburi; 2010.
[8] Rollini F, et al. Platelet function profiles in patients with diabetes mellitus. J
Cardiovasc Transl Res 2013;6(3):329–45.
[9] Mazzanti L, Mutus B. Diabetes-induced alterations in platelet metabolism. Clin
Biochem 1997;30(7):509–15.
6 N. Sansanayudh et al. / European Journal of Internal Medicine xxx (2015) xxx–xxx
[10] Papanas N, et al. Mean platelet volume in patients with type 2 diabetes mellitus.
Platelets 2004;15(8):475–8.
[11] Shah B, et al. The relationship between diabetes, metabolic syndrome, and platelet
activity as measured by mean platelet volume: the National Health and Nutrition
Examination Survey, 1999–2004. Diabetes Care 2012;35(5):1074–8.
[12] Turgutalp K, et al. Mean platelet volume and related factors in patients at different
stages of diabetic nephropathy: a preliminary study. Clin Appl Thromb Hemost
2014;20(2):190–5.
[13] Tavil Y, et al. Mean platelet volume in patients with metabolic syndrome and its re-
lationship with coronary artery disease. Thromb Res 2007;120(2):245–50.
[14] Furman-Niedziejko A, et al. Relationship between abdominal obesity, platelet blood
count and mean platelet volume in patients with metabolic syndrome. Folia Med
Cracov 2014;54(2):55–64.
[15] Kutlucan A, et al. The relationship between mean platelet volume with metabolic
syndrome in obese individuals. Blood Coagul Fibrinolysis 2012;23(5):388–90.
[16] Zaccardi F, et al. Platelet mean volume, distribution width, and count in type 2 dia-
betes, impaired fasting glucose, and metabolic syndrome: a meta-analysis. Diabetes
Metab Res Rev 2014.
[17] Elsherbiny IA, Shoukry A, El Tahlawi MA. Mean platelet volume and its relation to in-
sulin resistance in non-diabetic patients with slow coronary flow. J Cardiol 2012;
59(2):176–81.
[18] Varol E, et al. Mean platelet volume is associated with insulin resistance in non-
obese, non-diabetic patients with coronary artery disease. J Cardiol 2010;56(2):
154–8.
[19] Cao X, et al. Increased platelet volume in a general population with prehypertension:
a cross-sectional study of 80 545 participants from China. Hypertens Res 2012;
35(9):903–8.
[20] Nadar SK, et al. Platelet indexes in relation to target organ damage in high-risk hy-
pertensive patients: a substudy of the Anglo-Scandinavian Cardiac Outcomes Trial
(ASCOT). J Am Coll Cardiol 2004;44(2):415–22.
[21] Ntaios G, et al. Hypertension is an independent predictor of mean platelet volume in
patients with acute ischaemic stroke. Intern Med J 2011;41(9):691–5.
[22] Varol E, et al. Mean platelet volume in patients with prehypertension and hyperten-
sion. Clin Hemorheol Microcirc 2010;45(1):67–72.
[23] Yazici M, et al. Lifestyle modification decreases the mean platelet volume in
prehypertensive patients. Platelets 2009;20(1):58–63.
[24] Demirin H, et al. Normal range of mean platelet volume in healthy subjects: insight
from a large epidemiologic study. Thromb Res 2011;128(4):358–60.
[25] Maluf CB, Barreto SM, Vidigal PG. Standardization and reference intervals of platelet
volume indices: Insight from the Brazilian longitudinal study of adult health (ELSA-
BRASIL). Platelets 2014:1–8.
Please cite this article as: Sansanayudh N, et al, The association between mean platelet volume and cardiovascular risk factors, Eur J Intern Med
(2015), http://dx.doi.org/10.1016/j.ejim.2015.11.028
[26] Santimone I, et al. White blood cell count, sex and age are major determinants of
heterogeneity of platelet indices in an adult general population: results from the
MOLI-SANI project. Haematologica 2011;96(8):1180–8.
[27] Shimodaira M, et al. Gender differences in the relationship between serum uric acid
and mean platelet volume in a Japanese general population. Platelets 2014;25(3):
202–6.
[28] Adibi P, et al. Population-based platelet reference values for an Iranian population.
Int J Lab Hematol 2007;29(3):195–9.
[29] Butkiewicz AM, et al. Platelet count, mean platelet volume and thrombocytopoietic
indices in healthy women and men. Thromb Res 2006;118(2):199–204.
[30] Coban E, et al. The mean platelet volume in patients with obesity. Int J Clin Pract
2005;59(8):981–2.
[31] Yarlioglues M, et al. The acute effects of passive smoking on mean platelet volume in
healthy volunteers. Angiology 2012;63(5):353–7.
[32] Biljak VR, et al. Platelet count, mean platelet volume and smoking status in stable
chronic obstructive pulmonary disease. Platelets 2011;22(6):466–70.
[33] Arslan E, Yakar T, Yavasoglu I. [The effect of smoking on mean platelet volume and
lipid profile in young male subjects]. Anadolu Kardiyol Derg 2008;8(6):422–5.
[34] Coban E, Afacan B. The effect of rosuvastatin treatment on the mean platelet volume
in patients with uncontrolled primary dyslipidemia with hypolipidemic diet treat-
ment. Platelets 2008;19(2):111–4.
[35] Icli A, et al. Increased mean platelet volume in familial hypercholesterolemia.
Angiology 2015.
[36] Varol E, et al. Mean platelet volume is elevated in patients with low high-density li-
poprotein cholesterol. Angiology 2014;65(8):733–6.
[37] Jenkins DJ, et al. Effect on hematologic risk factors for coronary heart disease of a
cholesterol reducing diet. Eur J Clin Nutr 2007;61(4):483–92.
[38] Blaha M, et al. The decrease of mean platelet volume after extracorporeal LDL-
cholesterol elimination. Atheroscler Suppl 2013;14(1):77–81.
[39] Sivri N, et al. Statins decrease mean platelet volume irrespective of cholesterol low-
ering effect. Kardiol Pol 2013;71(10):1042–7.
[40] Bessman JD, Williams LJ, Gilmer Jr PR. Mean platelet volume. The inverse relation of
platelet size and count in normal subjects, and an artifact of other particles. Am J Clin
Pathol 1981;76(3):289–93.
[41] Thompson CB, et al. The role of anticoagulation in the measurement of platelet vol-
umes. Am J Clin Pathol 1983;80(3):327–32.
[42] Bath PM. The routine measurement of platelet size using sodium citrate alone as the
anticoagulant. Thromb Haemost 1993;70(4):687–90.