Nutrition 28 (2012) 148153

Contents lists available at ScienceDirect

Nutrition
journal homepage: www.nutritionjrnl.com

Applied nutritional investigation

Vegetarianism produces subclinical malnutrition, hyperhomocysteinemia

and atherogenesis
Yves Ingenbleek M.D. a, Kilmer S. McCully M.D. b, c, *
a

Laboratory of Nutrition, Faculty of Pharmacy, University Louis Pasteur, Strasbourg, France

Pathology and Laboratory Medicine Service, VA Boston Healthcare System, West Roxbury, Massachusetts, USA
c
Department of Pathology, Harvard Medical School, Boston, Massachusetts, USA
b

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 30 December 2010
Accepted 27 April 2011

Objective: To explain why vegetarian subjects develop morbidity and mortality from cardiovascular
diseases unrelated to vitamin B status and Framingham criteria.
Methods: A study of 24 rural male subjects 18 to 30 y old and 15 urban male controls was conducted in the Sahel region of Chad. Food consumption was determined from a dietary questionnaire, and overall health status was assessed by body weight, body mass index, serum albumin,
plasma transthyretin, urinary nitrogen, and creatinine. Plasma lipids, vitamins B6, B9 and B12,
homocysteine, and related sulfur amino acids were measured as selected cardiovascular disease
risk factors.
Results: Body weight, body mass index, blood, and urinary markers of protein status were significantly lower, with an estimated 10% decrease of lean body mass in the study group compared
with urban controls. Neither lipid fractions nor plasma levels of vitamins B6, B9, and B12 were
signicantly different between the two groups. Although the mean consumption of sulfur amino
acids (10.4 mg$kg
1$d
1) by rural subjects was signicantly below the recommended dietary
allowances (13 mg$kg
1$d
1), plasma methionine values were similar in the two groups. In
contrast, homocysteine concentration was signicantly increased (18.6 mmol/L, P < 0.001), and the
levels of cysteine and glutathione were signicantly decreased in the study group, demonstrating
inhibition of the trans-sulfuration pathway. The strong negative correlation (r
0.71) between
transthyretin and homocysteine implicated lean body mass as a critical determinant of
hyperhomocysteinemia.
Conclusion: The low dietary intake of protein and sulfur amino acids by a plant-eating population
leads to subclinical protein malnutrition, explaining the origin of hyperhomocysteinemia and the
increased vulnerability of these vegetarian subjects to cardiovascular diseases.
Published by Elsevier Inc.

Keywords:
Protein malnutrition
Lean body mass
Transthyretin
Homocysteine
Sulfur deciency
Cardiovascular disease

Introduction
According to World Health Organization estimates [1], 16.7
million deaths are attributable to cardiovascular disease (CVD)
each year, accounting for 30% of all deaths worldwide and
constituting an international pandemic [2]. More than half of
these deaths occur in developing countries where vegetarian
diets consisting of several varieties of legumes and starchy foods
have prevailed for millennia. The benecial health effects of such
regimens in the prevention of chronic disorders such as CVD,
This work was funded by University Louis Pasteur, Strasbourg, France.
* Corresponding author. Tel.: 857-203-5990; fax: 857-203-5623.
E-mail address: kilmer.mccully@va.gov (K. S. McCully).
0899-9007/$ - see front matter Published by Elsevier Inc.
doi:10.1016/j.nut.2011.04.009

diabetes, and cancer have been documented by many investigators [3,4], explaining their growing popularity in Westernized
countries during the past several decades. It is estimated that
about 2% to 5% of individuals living in developed countries have
adopted vegetarianism as a lifestyle choice [5,6].
Plant-based regimens, however, do not optimally fulll the
nutritional requirements of population groups living in developing countries, as illustrated by the increasing incidence of
hyperhomocysteinemia (HHcy) states [7,8], now recognized as
a critical factor in occurrence of CVD and stroke [9]. HHcy was
originally associated with the pathogenesis of atherosclerosis by
the study of the vascular pathology of children with inherited
disorders of methionine (Met) metabolism [10]. Dietary insufciency or malabsorption of any of three water-soluble B vitamins

Y. Ingenbleek, K. S. McCully / Nutrition 28 (2012) 148153

(B6, pyridoxal-50 -phosphate; B9, folates; or B12, cobalamins)

regulating the Methomocysteine (Hcy) cycle are regarded as
key determinants of acquired HHcy [11]. In particular, cobalamin
deciency constitutes a worldwide problem [12] and is currently
regarded as the principal causal factor of HHcy in vegetarian
populations [7,8] because vitamin B12 is almost totally absent
from plant foods. In contrast, folates are abundant in plant foods,
and pyridoxal-50 -phosphate is ubiquitous in unprocessed foods
from plant and animal sources. Nevertheless, some studies
undertaken in developing areas have shown that HHcy status
may develop despite normal B12 status [13], prompting the
search for alternative explanations.
A possible causal factor for HHcy in vegetarian populations is
related to the observation that most plant foods contain lower
concentrations of protein and essential amino acids, especially
lysine and Met, than animal foods [14]. White beans, for
example, contain about half the sulfur amino acids (SAAs; 422
mg/100 g of edible portion) contained in bovine meat (930 mg/
100 g) [15]. As a result, the Met provided by vegetarian diets does
not meet the tissue requirements of mammalian species [15].
Clinical investigations conducted in subclinically malnourished
patients [16,17] and recent experimental studies of the mouse
[18] and rat [19] models have revealed that restriction of dietary
Met intake inhibits the trans-sulfuration cascade and leads to
HHcy, irrespective of cobalamin status. The present eld study
was undertaken to investigate the importance of dietary vitamin
B12 and Met deciency in the development of HHcy. The results
explain how the altered body composition of vegetarian subjects
affects SAA metabolism and produces HHcy, a risk factor for CVD
in vegetarian populations.

149

Dietary and health assessments

A questionnaire assessing detailed information regarding family status,
occupation, and dietary habits was recorded from the 24 subjects who successfully completed the screening protocol. During the week before collection of
blood and urine samples, the typical daily portion size was provided for each food
item and expressed in grams. The results were compared with values reported in
food composition tables [21,22]. The data allowed the calculation of daily
consumption of energy, ber, carbohydrate, lipid, protein, and SAAs after
adjustment for body weight (BW). Body mass index (BMI) was calculated as BW
(kilograms) divided by height (meters) squared.
Venous blood samples obtained after overnight fast were centrifuged to yield
plasma and serum that were frozen at
20 C and shipped on dry ice. A 24-h
urine sample was collected into sterile polyvinylchloride asks under night and
day supervision of eld staff, and, after volume measuring, an aliquot was stored
at
20 C for determination of urinary nitrogen by the Kjeldhal method and
 method, allowing estimates of total body nitrogen turncreatinine by the Jaffe
over rate and muscle mass, respectively.
Specimen analysis
Serum cholesterol was determined by an enzymatic photometric assay
(CHOD-PAP method) using a Hitachi 917 analyzer (Roche, Basel, Switzerland).
Serum high-density lipoprotein (HDL) and low-density lipoprotein (LDL) were
determined by homogeneous enzymatic colorimetric assays (Roche). The
between-run CVs were 1.4% for cholesterol, 1.9% for LDL, and 2.3% for HDL. Serum
triacylglycerols were assayed by the GPO-PAP method (Advia Trig, Bayer,
Leverkusen, Germany), and the CV was 1.4%.
Plasma Hcy was determined using a reverse-phase high-performance liquid
chromatographic method [23]. The detection limit was 0.5 mmol/L, and the
between-run CV was 5.5%. Plasma cysteine (Cys) was determined after deproteinization of plasma by sulfosalicylic acid and chromatography on a Beckman
121 MB Amino Acid Analyzer (Beckman-Coulter, Villepinte, France) using a single
column and a sequence of lithium citrate buffers of increasing pH. Glutathione
(GSH) was determined by a high-performance liquid chromatographic method
[24]. Plasma vitamin B12 and folates were determined by chemoluminescence
(ACS 180, Bayer Vital). The between-run CVs for vitamin B12 and folates were
6.9% and 5.8%, respectively. Plasma pyridoxine was determined by high-performance liquid chromatography by uorometric detection, with a CV of 5.1%.

Materials and methods

Statistical analysis
Study participants and controls
The study was conducted in 2009 in the Sahel region of southern Chad,
about 400 km from the capital city of NDjamena. The Sahel belt is a shrubby
countryside, which extends across Africa from Senegal to Somalia, and is
locally populated by the Sara ethnic group. The men are physically active
farmers engaged in sowing, irrigation from well water, and harvesting. In this
semiarid area, the staple foods are cassava, sweet potatoes, beans, millets, and
ground nuts. Cassava leaves, cabbages, and carrots provide adequate carotenes,
folates, and pyridoxine, but contain little protein or SAAs. The lack of satisfactory grazing land explains why livestock are undernourished and provide
few dairy foods. Meat is rarely eaten, but poultry and eggs are occasionally
consumed.
The study was approved by the Ministry of Public Health and by regional
health authorities. From 47 candidates, 24 men 18 to 30 y old in satisfactory
health were recruited from three remote rural hamlets and examined for health
status, dietary habits, and anthropomorphic parameters. Women were excluded
because of the difculty in collecting valid urine specimens in an animist population. Alcoholism (1 of 47), smoking (3 of 47), medications (4 of 47), microalbuminuria (2 of 47), and cytokine-induced overproduction of acute-phase
reactants (13 of 47) were exclusion criteria. In our clinical experience, the
combined measurement of C-reactive protein (CRP; >10 mg/L) and of a1-acid
glycoprotein (>800 mg/L) allows the identication of nearly all viral, bacterial,
fungal, and parasitic infestations [20], notably hepatitis B, human immunodeciency virus, tuberculosis, and malaria, which have a signicant prevalence in
this area. CRP and a1-acid glycoprotein inammatory markers were determined
using immunoassays with, in the case of CRP, high-sensitivity Denka reagents.
The minimally detectable CRP threshold was 0.03 mg/L and the interassay
coefcient of variation (CV) was 4%. Our 24 rural participants and 15 control
subjects had a1-acid glycoprotein values below 700 mg/L. Their high-sensitivity
CRP values were 0.34  0.41 and 0.23  0.32 mg/L, respectively, with no statistically signicant difference between groups (P > 0.05). Identication of microalbuminuria was accomplished by testing a morning urine specimen. The control
participants consisted of 15 healthy men 18 to 29 y old, belonging to the same
Sara ethnic group but residing in an upper-class urban environment. The diet of
the control group was similar to that of the study group except for occasional
consumption of bovine meat, canned or powdered milk, and smoked sh.

All values are presented as mean  standard deviation. For normal data,
differences between study and control participants were compared using the
Student t test. For non-normal data such as values for plasma Hcy and vitamins
B6, B9, and B12, median values were calculated. The Mann-Whitney test, a nonparametric procedure, was used to compare the median values between the
study and control groups. Linear regression models were applied to determine
the relation between variables. Spearman rank correlation coefcients (r) evaluated the relations and degree of signicance (P) between the different variables.
Differences with P  0.05 were considered statistically signicant.

Results
Food consumption differed between the study and control
groups, as assessed from the food questionnaire (Table 1).
Rural vegetarian participants consumed a comparable energy
intake, but they consumed less protein and fat and more
carbohydrate and ber than the urban omnivorous participants. The control group consumed a well-balanced dietary
regimen with 54.1% of total energy as carbohydrate, 32.6% as
fat, and 13.3% as protein, close to the recommended dietary
allowances for each nutrient class, in contrast with the
signicantly (P < 0.05) larger proportion of carbohydrate
(63.1%) but less fat (25.7%) and less protein (11.2%) consumed
by the study group. The consumption of SAAs (10.4
mg$kg
1$d
1) by the study group was signicantly (P < 0.001)
lower than the control group (14.6 mg$kg
1$d
1).
Measurements of anthropometric and biochemical markers
of health status of the study group were signicantly lower than
those of the control group (Table 2). The study group was characterized by normal height but signicantly (P < 0.05) lower BW,
BMI, and serum albumin (SA) and much lower plasma

150

Y. Ingenbleek, K. S. McCully / Nutrition 28 (2012) 148153

Table 1
Characteristics of food consumption
Nutrient classes

Energy (MJ/d)
Fiber (g/d)
Carbohydrate (g/d)
Total energy (%)
Fat (g/d)
Total energy (%)
Protein (g/d)
Total energy (%)
SAA (mg$kg
1$d
1)

Study group

Control group

(n 24)

(n 15)

7.33 (6.18.4)
73.4 (6297)
265.3 (174389)
63.1
52.1 (3483)
25.7
49.7 (2451)
11.2
10.4 (912)

7.22 (6.48.9)
49.4 (3873)
220.5 (160291)
54.1
62.4 (4884)
32.6
63.4 (2959)
13.3
14.6 (1121)

NS
<0.01
<0.05
<0.05
<0.01
<0.001

SAA, sulfur amino acid

Intakes of dietary nutrients were calculated from the analysis of a 1-wk dietary
questionnaire for rural vegetarian participants in the study group and urban
omnivorous participants in the control group. Intakes of dietary SAAs were
calculated from tables of composition of dietary proteins.

transthyretin (TTR; P < 0.001) compared with the control group.

Despite an apparently satisfactory state of health, plasma TTR
values of all rural subjects were 240 mg/L (Fig. 1), indicating
that an element of protein malnutrition was present in all
members of the study group. The lower excretion of urinary
nitrogen and creatinine by the study group reected lower
protein consumption and decreased lean body mass (LBM) and
muscle mass, respectively, compared with the control group.
There was no difference in serum levels of triacylglycerols,
cholesterol, LDL cholesterol, or HDL cholesterol between the two
groups (Table 2). There was no difference in plasma levels of
vitamins B6 or B9 between the two groups, and the lower plasma
level of vitamin B12 of the study group (174 pmol/L) was of
borderline signicance (P 0.067) compared with the control
group (269 pmol/L). In the study group, only 4 of 24 participants
had plasma vitamin B12 levels below the lower limit of normal

Table 2
Health and plasma indices
Indices

Overall health status

Height (m)
Weight (kg)
Body mass index (kg/m2)
Serum albumin (g/L)
Plasma transthyretin (mg/L)
Urinary nitrogen (g/d)
Urinary creatinine (g/d)
Serum lipids
Triacylglycerols (mmol/L)
Total cholesterol (mmol/L)
HDL cholesterol (mmol/L)
LDL cholesterol (mmol/L)
Plasma vitamins
Pyridoxine (B6, nmol/L)
Folate (B9, nmol/L)
Cobalamin (B12, pmol/L)
Plasma SAAs
Methionine (mmol/L)
Homocysteine (mmol/L)
Cysteine (mmol/L)
Glutathione (mmol/L)
Taurine (mmol/L)

Study group

Control group

(n 24)

(n 15)

1.68
61.7
19.3
3.8
178.1
9.6
1.18
0.85
3.98
1.21
2.27













0.3
3.6
0.8
0.6
27.5
0.9
0.37
0.55
0.59
0.41
0.32

41.9  7.4
16.9  2.7
174  28
29.3
18.6
14.5
2.7
32







3
2.9
8.3
0.7
6

1.73
69.3
22.1
4.3
292.9
11.2
1.43
1.08
4.31
1.49
2.12













0.3
1.9
0.6
0.4
10.8
0.9
0.26
0.41
0.92
0.30
0.25

NS
<0.05
<0.05
<0.05
<0.001
<0.05
<0.05
NS
NS
NS
NS

52.4  13.8
14.4  2.6
269  36

NS
NS
<0.06







NS
<0.001
<0.001
<0.01
<0.01

28.0
10.8
43.4
4.3
58

4
0.7
4.5
0.7
9

HDL, high-density lipoprotein; LDL, low-density lipoprotein; SAA sulfur amino

acid
Parameters of general health status and serum lipids, plasma vitamins, and
plasma SAAs were determined after 1 wk of observation.

Fig. 1. Relation of plasma homocysteine and transthyretin. Regression line analysis

plots total plasma homocysteine and plasma transthyretin values in participants
with protein malnutrition (diamonds) and in controls (circles). Decreased plasma
transthyretin values are negatively correlated (r
0.71) with increased plasma
homocysteine values in the study group, indicating that a decrease of lean body
mass is a critical determinant of hyperhomocysteinemia.

(140 pmol/L). A previous study undertaken in the same Chadian area [17] in a larger group of 60 rural participants did
demonstrate a weak inverse correlation between B12 and Hcy
concentrations in the 20 subjects most severely protein depleted,
as identied by the lowest TTR plasma values. It is therefore
likely that the HHcy status of some of our rural subjects in the
present survey might have resulted from combined B12 and
protein deciencies. The correlation of B12 deciency with HHcy
could well reach statistical signicance if a larger group of
subjects were studied. Analysis of plasma SAAs showed no
difference in concentration of Met between the study and control
groups (Table 2). In contrast, plasma levels of Hcy were signicantly higher (18.6 mmol/L, P < 0.001) in the study group than in
the control group (mean average 10.8 mmol/L, with no value
11.7 mmol/L). Plasma levels of Cys (P < 0.001) and GSH (P <
0.01) were signicantly lower in the study group compared with
the control group.
The Spearman correlation coefcients among plasma Hcy,
nutritional indicators, and dietary variables (Table 3) indicated

Table 3
Spearman correlation coefcients between total homocysteine and nutritional
indicators and signicance
Nutritional indicator

Study group

Control group

(n 24)

Body mass index

Serum albumin
Plasma transthyretin
Urinary nitrogen
Urinary creatinine
Triacylglycerols
Total cholesterol
HDL cholesterol
LDL cholesterol
Pyridoxine (B6)
Folate (B9)
Cobalamin (B12)
Methionine
Cysteine
Glutathione
Taurine

(n 15)

0.139

0.273

0.709

0.357

0.285
0.215
0.188

0.315
0.209
0.189
0.237

0.175
0.260

0.668

0.319

0.182

NS
NS
<0.001
0.003
0.004
NS
NS
NS
NS
NS
NS
0.061
NS
<0.001
0.013
0.043

0.047
0.029

0.275
0.058
0.067
0.086
0.127

0.075
0.286
0.279
0.189

0.067

0.023
0.215
0.181
0.098

NS
NS
0.004
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS

HDL, high-density lipoprotein; LDL, low-density lipoprotein

Y. Ingenbleek, K. S. McCully / Nutrition 28 (2012) 148153

a signicant association of decreased plasma TTR, urinary

nitrogen, and creatinine with Hcy in the study group but no
association, except for plasma TTR, in the control group. Correlations between BMI and SA with Hcy were not signicant for
either group. There was no correlation of serum triacylglycerols,
cholesterol, HDL, or LDL with Hcy in either group.
The data in Figure 1 show a strongly negative correlation (r

0.71) between plasma TTR and Hcy in the study group. The data
in Figure 2 show a strongly negative correlation between plasma
Cys and Hcy (r
0.67) in the study group but not in the control
group.

Discussion
The blood lipid proles of rural subjects were within normal
ranges (Table 2), the likely result of low fat and high ber intakes
and strenuous physical activity, ruling out these parameters as
causal risk factors for CVD. The normal concentrations measured
for vitamins B6, B9, and B12 (Table 2) excluded deciencies of
these micronutrients as determinants of HHcy. In contrast, the
consumption of a protein-decient diet caused signicant
alterations in body composition (Table 2). All members of the
study group had signicantly lower BW, BMI, SA, and plasma TTR
levels (Table 2, Fig. 1) than the control group, despite an apparently good state of health. Recent observations have demonstrated that LBM constitutes a critical determinant of Hcy status
in healthy subjects [25] and in diseased patients [26]. There exist
close relations between the evolutionary patterns of TTR and
LBM values in health and disease [26]. The Foundation for Blood
Research (Scarborough, ME, USA) recently published TTR results
in 68 720 healthy US citizens 0 to 100 y, constituting reference
data for evaluating the relation of LBM uctuations with TTR in
relation to sex and age [26]. Decreased SA was a less sensitive
index of protein malnutrition than TTR [27]. Nevertheless,
because LBM represents about two-thirds of BW [26], the
concomitant lowering of BW, BMI, SA, and TTR concentrations
and of urinary nitrogen and creatinine excretions (Table 2) were
characteristic biomarkers of subclinical protein malnutrition,
reecting an average decrease of LBM by 10% in the study group.
Calculation of the heightcreatinine index was established by

Fig. 2. Relation of plasma homocysteine and cysteine. Regression line analysis plots
total plasma homocysteine and plasma cysteine values in participants with protein
malnutrition (diamonds) and in controls (circles). Decreased plasma cysteine values
are negatively correlated (r
0.67) with increased plasma homocysteine values in
the study group, indicating that downregulation of cystathionine-b-synthase
activity is dependent on the bioavailability of methionine.

151

comparison of 24-h creatinine excretion of the rural participants

versus 24-h excretion of control subjects of the same height. The
normal heightcreatinine index is 1.0, and this ratio progressively decreases as the muscle mass decreases toward sarcopenia. In our study, the rural group had a heightcreatinine index of
0.83 (17% lower than controls). A decrease of LBM by about 10%
suggested that muscle mass was more severely protein depleted
than the visceral organs in the study group.
The consumption of 10.4 mg$kg
1$d
1 of the two SAAs
(approximately two-thirds Met and one-third Cys [15]) by the
study group was below the recognized recommended dietary
allowance of 13 mg$kg
1$d
1 [21] for adequate nutrition.
Despite the low consumption of SAAs, plasma Met levels were
the same in the two groups, contrasting with increased Hcy (18.6
mmol/L) and decreased Cys and GSH concentrations in the study
group (Table 2). These data indicate that maintenance of Met
plasma concentration is mandatory for the synthesis and functioning of many Met-dependent metabolic processes described
in detail elsewhere [15,28]. These changes in SAA metabolism
demonstrate downregulation of the trans-sulfuration pathway at
the cystathionine-b-synthase (CbS) level, producing increased
Hcy in biological uids and increased methylation of Hcy to Met.
This remarkable adaptive response was rst observed during
a eld study conducted in 1986 in Senegal [16] and was later
conrmed in clinical studies [17] and animal experiments [18,19].
These observations show that dietary Met deprivation causes
post-transcriptional downregulation of CbS activity without
a decrease in hepatic S-adenosylmethionine [18], the classic
allosteric regulator of this enzyme [29]. The strong negative
correlation (r
0.71) found between TTR and Hcy, also reported in previous studies [15,26,28], indicates that decreased LBM
is the primary determinant of HHcy in malnourished individuals
[26]. Because the concentrations of nitrogen and sulfur are
closely correlated in body tissues [15,26], a decrease in LBM
causes concomitant shrinking in total body nitrogen and in total
body sulfur that is proportional to the severity of protein
malnutrition. The data are consistent with mouse [18] and rat
[19] experiments in displaying a BW loss of 33% and 44%,
respectively, at the end of the Met-deprivation period. We
postulate that depletion of total body sulfur and of total body
Met, its principal constituent, activates an adaptive feedback
loop downregulating CbS activity [15].
The signicantly low plasma Cys levels in the study group
(33.4% of control value; Table 2, Fig. 2) may be explained by lower
dietary SAAs (Table 1), a decreased release of Cys from protein
catabolism in malnutrition [30], and a decreased conversion of
Met to Cys because of decreased CbS activity. The signicantly
low plasma GSH level in the study group (62.8% of control value;
Table 2) is in part explained by the almost complete absence of
GSH from cereals, roots, and dairy products [31] and by the
restriction of Met and Cys intake, reecting primary substrate
limitations [32]. Cys undergoes reversible GSH conversion and is
catabolized to hydrogen sulde (H2S) using alternate reactions
catalyzed by two pyridoxal-50 -phosphatedependent enzymes,
CbS and cystathionine-g-lyase [33]. Our data show that the most
decreased Cys and GSH values were found in subjects with the
lowest TTR concentrations and decreased LBM, reecting the
most marked protein-depleted states. Cys is a limiting substrate
for the production of GSH and H2S, suggesting that decreased
availability of Cys, GSH, and H2S causes a loss of their reducing
properties in proportion to the decrease in LBM. Besides nitrogen
oxide and carbon monoxide, H2S has been recently recognized as
the third gaseous signaling molecule that mediates vasodilation
effects [34]. The reducing property of H2S functions as a major

152

Y. Ingenbleek, K. S. McCully / Nutrition 28 (2012) 148153

scavenging agent against HHcy-induced accumulation of reactive oxygen species [35].

Of additional importance is the discovery that Cys and GSH
participate in the non-enzymatic conversion of elemental sulfur
(S8) into H2S [36]. This converting process may be impeded in the
case of insufcient dietary S8 intake and underlying malnutrition, causing decreased Cys and GSH bioavailability. Pedologists
and agronomists have demonstrated for many decades that S8 is
an essential element in plant nutrition, because deciency of S8
produces growth retardation, decreased chlorophyll synthesis,
and decreases in fertility and crop yields [37]. Because plant
foods supply inadequate SAAs for mammalian tissue requirements [15,28], consumption of foods with imbalanced nitrogen:sulfur ratios grown in S8-decient agricultural soils will
exacerbate protein malnutrition [38]. S8 deciency is a worldwide soil problem affecting more than half the countries
producing mainly foodstuffs of plant origin, including rice,
soybeans, wheat, maize, and oilseeds [37].
In contrast to the status of plant agronomy, the mammalian
requirements for elemental S8 are presently unknown. Only very
recently has the essentiality of sulfur has been recognized in
human nutrition, causing Hcy elevation in decient subjects [39].
The Indian subcontinent, notably the northern region extending
from Pakistan to Bangladesh, is one of the areas with soil that is
most severely decient in S8 [40], contributing to the public
health problem of CVD. More than 40% of Indian patients with
HHcy-induced CVD are refractory to combined administration of
folates and cobalamins [41], supporting the concept of S8 deciency as the key etiologic factor. The concept of sulfur deciency
may explain why many clinical trials using B vitamin interventions have proved ineffective in the prevention of recurrent
vascular disease in patients with advanced CVD, cerebrovascular
disease, renal failure, and inammatory disorders [42]. Many
other countries in southeastern Asia with vegetarian lifestyles,
such as China, Taiwan, Indonesia, Philippines, Pakistan, and Sri
Lanka, are also affected by increasing CVD morbidity and
mortality [79,37]. These developments call attention to the
need to redene CVD risk factors in South Asia [43] and explain
why cardiologists from ve of these countries, Bangladesh, India,
Pakistan, Nepal, and Sri Lanka, have exhorted their governments
to focus more attention on CVD epidemiology [44].
We assume that our Chadian data are representative of the
health status of most vegetarian African populations [13,16,17].
The rst African report showing that HHcy is signicantly associated with ischemic stroke came from a study carried out in the
Sahel region of Nigeria, close to the Chadian border [45]. In that
study, 50 adult patients had a clinical diagnosis of cerebral
infarction conrmed by computerized tomographic scanning. At
the onset of thrombovascular injury, HHcy (20.8 mmol/L) was
found in all patients, revealing a slightly higher concentration
than that found in our apparently healthy Sara farmers (18.6
mmol/L).
The impact of our observational study is limited because of
the modest number of participants in the study and control
groups, resulting from the difculty in selecting valid candidates
who fulll the requirements of eld study conditions. Additional
larger investigations in vegetarian populations with protein
malnutrition and dietary S8 deciency in other regions of the
world are needed to conrm the present ndings.
In conclusion, our investigation shows that a vegetarian
population consuming a diet decient in SAAs is characterized
by normal B vitamin status, normal serum lipids, subclinical
protein malnutrition, and decreased LBM, leading to an
adaptive downregulation of the trans-sulfuration pathway,

HHcy, and oxidative stress, thereby increasing the vulnerability

to CVD.
Acknowledgments
The authors thank the members of the study and control
groups for their participation and the eld workers for their
assistance in the completion of the family and dietary questionnaires and for procurement of the blood and urine specimens. K. Braunschweig, Ph.D., D. Mirallet, Ph.D., and G. Verbiest
contributed to the measurement of laboratory parameters and to
the statistical analysis of the data and received no additional
compensation for their services.

References
[1] World Health Organization. The world health report 2001. Health systems:
improving performance. Geneva: World Health Organization; 2001.
[2] McDermott MM. The international pandemic of chronic cardiovascular
disease. JAMA 2007;297:12535.
[3] Walter P. Effects of vegetarian diets on aging and longevity. Nutr Rev
1997;55:S618.
[4] Key TJ, Appleby PN, Rosell MS. Health effects of vegetarian and vegan diets.
Proc Nutr Soc 2006;65:3541.
[5] American Dietetic Association. Position of the American Dietetic Association and Dieticians of Canada: vegetarian diets. ADA Rep 2003;6:106.
[6] Reddy S. Vegetarian diets. In: Caballero B, Allen I, Prentice A, editors. The
Encyclopedia of human nutrition. 2nd ed. London: Academic Press; 2005.
p. 1317.
[7] Hung CJ, Huang PC, Lu SC, Li YH, Huang HB, Lin BF, et al. Plasma homocysteine levels in Taiwanese vegetarians are higher than those of omnivores. J Nutr 2002;132:1528.
[8] Yajnik CS, Deshpande SS, Lubree HG, Naik SS, Bhat DS, Uradey BS, et al.
Vitamin B12 deciency and hyperhomocysteinemia in rural and urban
Indians. J Assoc Phys India 2006;54:77582.
[9] Christopher R, Nagaraja D, Shankar SK. Homocysteine and cerebral stroke
in developing countries. Curr Med Chem 2007;14:2393401.
[10] McCully KS. Vascular pathology of homocysteinemia: Implications for the
pathogenesis of arteriosclerosis. Am J Pathol 1969;56:11128.
[11] McCully KS. Homocysteine and vascular disease. Nat Med 1996;2:3869.
[12] Stabler S, Allen R. Vitamin B12 deciency as a worldwide problem. Annu
Rev Nutr 2004;24:299326.
[13] Glew RH, Williams M, Conn CA, Cadena SM, Crossey M, Okolo SN,
VanderJagt DJ. Cardiovascular disease risk factors and diet of Fulani
pastoralists of northern Nigeria. Am J Clin Nutr 2001;74:7306.
[14] Young VR, Pellett PL. Plant proteins in relation to human protein and amino
acid nutrition. Am J Clin Nutr 1994;59:S120312.
[15] Ingenbleek Y. The nutritional relationship linking sulfur to nitrogen in
living organisms. J Nutr 2006;136:S164151.
[16] Ingenbleek Y, Barclay D, Dirren H. Nutritional signicance of alterations in
serum amino acid patterns in goitrous patients. Am J Clin Nutr
1986;43:31019.
[17] Ingenbleek Y, Hardillier E, Jung L. Subclinical malnutrition is a determinant
of hyperhomocysteinemia. Nutrition 2002;18:406.
[18] Tang B, Mustafa A, Gupta S, Melnyk S, James SJ, Kruger WD. Methioninedecient diet induces post-transcriptional down-regulation of cystathionine b-synthase. Nutrition 2010;26:11705.
[19] Elshorbagy AK, Valdivia-Garcia M, Refsum H, Smith AD, Mattocks DA,
Perrone CE. Sulfur amino acids in methionine-restricted rats: hyperhomocysteinemia. Nutrition 2010;26:12014.
[20] Ingenbleek Y, Carpentier YA. A prognostic inammatory and nutritional
index scoring critically ill patients. Int J Vitam Nutr Res 1985;55:91101.
[21] Food and Agriculture Organization. Amino-acid content of foods and biological data on proteins. Rome: Food and Agriculture Organization; 1970.
[22] McCance RA, Widdowson EM. The composition of foods. 5th ed. Cambridge: Royal Society of Chemistry; 1991.
[23] Araki A, Sako Y. Determination of free and total homocysteine in human
plasma by high-performance liquid chromatography with uorescence
detection. J Chromatogr 1987;422:4352.
[24] Toyooka T, Uchiyama S, Saito Y, Imai K. Simultaneous determination of
thiols and disuldes by high-performance liquid chromatography with
uorescence detection. Anal Chim Acta 1988;205:2941.
[25] Battezzatti A, Bertoli S, San Romerio A, Testolin G. Body composition: an
important determinant of homocysteine and methionine concentrations in
healthy individuals. Nutr Metab Cardiovasc Dis 2007;17:52534.
[26] Ingenbleek Y. Plasma transthyretin reects the uctuations of lean body
mass in health and disease. In: Richardson SJ, Cody V, editors. Recent

Y. Ingenbleek, K. S. McCully / Nutrition 28 (2012) 148153

[27]
[28]

[29]

[30]

[31]

[32]

[33]

[34]
[35]

advances in transthyretin evolution, structure and biological functions.

Berlin: Springer Verlag; 2009. p. 32957.
Ingenbleek Y, De Visscher M, De Nayer P. Measurement of prealbumin as
index of proteincalorie malnutrition. Lancet 1972;300:1069.
Ingenbleek Y, Young VR. The essentiality of sulfur is closely related to
nitrogen metabolism: a clue to hyperhomocysteinemia. Nutr Res Rev
2004;17:13553.
Finkelstein JD, Kyle WE, Martin JJ, Pick AM. Activation of cystathionine
synthase by adenosylmethionine and adenosylethionine. Biochem Biophys
Res Commun 1975;66:817.
Jahoor F, Badaloo A, Reid M, Forrester T. Sulfur amino acid metabolism in
children with severe childhood undernutrition: cysteine kinetics. Am J Clin
Nutr 2006;84:13939.
Jones DP, Coates RJ, Flagg EW, Eley JW, Block G, Greensberg RS, et al.
Glutathione in foods listed in the National Cancer Institutes health
habits and history food frequency questionnaire. Nutr Cancer
1992;17:5775.
Lyons J, Rauh-Pfeiffer A, Yu YM, Lu XM, Zurakowski D, Tompkins RG, et al.
Blood glutathione synthesis rates in healthy adults receiving a sulfur amino
acid-free diet. Proc Natl Acad Sci U S A 2000;97:50716.
Stipanuk MH, Ueki I. Dealing with methionine/homocysteine sulfur:
cysteine metabolism to taurine and inorganic sulfur. J Inherit Metab Dis
2011;34:1732.
Wang R. Twos company, threes a crowd: can H2S be the third endogenous
gaseous transmitter? FASEB J 2002;16:17928.
Chang L, Geng B, Yu F, Zhao J, Jiang H, Du J, Tang C. Hydrogen sulde
inhibits myocardial injury induced by homocysteine in rats. Amino Acids
2008;34:57385.

153

[36] Searcy DG, Lee SH. Sulfur reduction by human erythrocytes. J Exp Zool
1998;282:31022.
[37] Jez JM, editor. Sulfur: a missing link between soils, crops, and nutrition.
Agriculture monograph no. 50. Madison, WI: American Society of
Agronomy, Crop Science Society of America, Soil Science Society of America; 2008.
[38] Das SK, Chhabra P, Chatterjee SR, Abrol YP, Deb DL. Inuence of sulphur
fertilization on yield of maize and protein quality of cereals. Fertil News
1975;20:302.
[39] Ingenbleek Y. Hyperhomocysteinemia is a biomarker of sulfur-deciency in
human morbidities. Open Clin Chem J 2009;2:4960.
[40] Khurana MPS, Sadana US, Bijay-Singh. Sulfur nutrition of crops in the IndoGangetic plains of South Asia. In: Jez JM, editor. Sulfur: a missing link
between soils, crops and nutrition. Madison, WI: American Society of
Agronomy, Crop Science Society of America, Soil Science Society of
America; 2008. p. 1124.
[41] Ghosh K, Khare A, Shetty S. Fasting plasma homocysteine levels in Indian
patients with acute myocardial infarction from Western India. Indian Heart
J 2007;59:2425.
[42] McCully KS. Homocysteine, vitamins, and vascular disease prevention. Am J
Clin Nutr 2007;74:S15638.
[43] Ahmad U, Frossard PM. Coronary heart diseases in South Asia: need to
redene risk. Int J Cardiol 2006;107:28990.
[44] Srivastava R. South Asias governments exhorted to focus on CVD. Lancet
2002;359:858.
[45] Alkali NH, Watt H, Bwala SA, Gadzama A. Association of plasma homocysteine and ischaemic stroke in a Nigerian population. Pak J Med Sci
2006;22:4058.