Hypoglycemic effect of Lupinus mutabilis in healthy volunteers and subjects with dysglycemia

Abstract

Metabolic syndrome and type-2 diabetes are increasing health problems that negatively affect health care systems worldwide. There is a constant urge to develop new therapies with better effects, lower side effects at lower prices to treat these diseases. Lupinus species and their derivates are good candidates to be used as hypoglycaemic agents. A phase II clinical trial was conducted to assess the role of raw Lupinus mutabilis on blood glucose and insulin in normoglycemic and dysglycemic subjects. Results show that consumption of L. mutabilis by normal weight healthy young individuals did not change importantly blood glucose and insulin levels. On the other hand, consumption of similar doses of lupinus by dysglycemic individuals (fasting glucose > 100 mg/dL) decreased significantly blood glucose. Lupinus effects were greater in those subjects with higher basal glucose levels. Glucose lowering effects of lupinus were not observed after soy intake that was used as control. A statistically significant reduction in insulin levels was also observed in the lupinus group compared with the soy group after 60 minutes of treatment. Furthermore, only treatment with lupinus improved insulin resistance in dysglycemic subjects. These data demonstrate that lupinus consumption could be a feasible and low cost alternative to treat chronic hyperglycemic diseases.

Full text

425
Nutr Hosp. 2012;27(2):425-433
ISSN 0212-1611 • CODEN NUHOEQ
S.V.R. 318
Original
Hypoglycemic effect of Lupinus mutabilis in healthy volunteers
and subjects with dysglycemia
M. Fornasini1, J. Castro2, E. Villacrés3, L. Narváez1, M.ª P. Villamar1y M. E. Baldeón1
1Colegio de Ciencias de la Salud. Universidad San Francisco de Quito. 2Servicio de Diabetología. Unidad Municipal de Salud
Norte Patronato San José. 3Instituto Nacional Autónomo de Investigaciones Agropecuarias, INIAP. Quito. Ecuador.
EFECTO HIPOGLICEMIANTE DE LUPINUS
MUTABILIS EN VOLUNTARIOS SANOS Y SUJETOS
CON DISGLICEMIA
Resumen
La diabetes tipo 2 y el síndrome metabólico son proble-
mas de salud en crecimiento que afectan a los sistemas de
salud en todo el mundo. Hay una necesidad urgente de
desarrollar terapias nuevas con mejores efectos, con
menos efectos adversos y de bajo costo para tratar estas
patologías. Las especies de Lupinus y sus derivados son
buenos candidatos para ser utilizados como agentes hipo-
glicemiantes. Se realizó un estudio clínico de fase II para
analizar el efecto de Lupinus mutabilis crudo sobre los
niveles de glucosa e insulina en la sangre de sujetos nor-
males y con disglicemia. Los resultados del estudio
demuestran que el consumo de L mutabilis por sujetos
sanos, jóvenes de peso normal, no altera importante-
mente los niveles sanguíneos de glucosa o insulina. Por
otro lado, la ingesta de dosis similares por individuos con
disglicemia (glucosa en ayunas > 100 mg/dL) disminuyó
significativamente los niveles de glucosa. Los efectos del
lupinus fueron más evidentes en aquellos sujetos con los
niveles basales de glucosa más altos. Los efectos hipoglice-
miantes de lupinus no se observaron después del consumo
de soya que se utilizó como control. Se observó también
una disminución estadísticamente significativa en los
niveles de insulina sanguínea luego de 60 minutos en el
grupo de voluntarios que consumió lupinus pero no en
aquellos que consumieron soya. Además solamente el tra-
tamiento con lupinus mejoró la resistencia a la insulina en
los sujetos con disglicemia. Estos datos demuestran que el
consumo de lupinus podría ser una alternativa factible y
de bajo costo para el tratamiento de enfermedades cróni-
cas con hiperglicemia.
(Nutr Hosp. 2012;27:425-433)
DOI:10.3305/nh.2012.27.2.5412
Palabras clave: Lupinus mutabilis. Hipoglicemia. Diabe-
tes. Ecuador. Alcaloides.
Abstract
Metabolic syndrome and type-2 diabetes are increasing
health problems that negatively affect health care systems
worldwide. There is a constant urge to develop new thera-
pies with better effects, lower side effects at lower prices
to treat these diseases. Lupinus species and their derivates
are good candidates to be used as hypoglycaemic agents. A
phase II clinical trial was conducted to assess the role of raw
Lupinus mutabilis on blood glucose and insulin in normo-
glycemic and dysglycemic subjects. Results show that
consumption of L. mutabilis by normal weight healthy
young individuals did not change importantly blood glucose
and insulin levels. On the other hand, consumption of
similar doses of lupinus by dysglycemic individuals (fasting
glucose > 100 mg/dL) decreased significantly blood
glucose. Lupinus effects were greater in those subjects
with higher basal glucose levels. Glucose lowering effects
of lupinus were not observed after soy intake that was
used as control. A statistically significant reduction in
insulin levels was also observed in the lupinus group
compared with the soy group after 60 minutes of treat-
ment. Furthermore, only treatment with lupinus
improved insulin resistance in dysglycemic subjects.
These data demonstrate that lupinus consumption could
be a feasible and low cost alternative to treat chronic
hyperglycemic diseases.
(Nutr Hosp. 2012;27:425-433)
DOI:10.3305/nh.2012.27.2.5412
Key words: Lupinus mutabilis. Hypoglycemia. Diabetes.
Ecuador. Alkaloids.
Correspondence: Manuel E. Baldeón.
Colegio de Ciencias de la Salud.
Universidad San Francisco de Quito.
Hospital de los Valles.
Edificio de Consultorios Médicos.
Planta Baja.
Vía Interoceánica, km. 12 1/2 y Av. Florencia.
Sector La Primavera, Cumbayá.
Quito. Ecuador.
E-mail: manuelb@usfq.edu.ec
Recibido: 2-VIII-2011.
Aceptado: 27-IX-2011.

Introduction
Type-2 diabetes is a highly prevalent disease around
the world. It was reported that 135 million people were
affected in 2005 and it has been estimated that approxi-
mately 217 million will be affected with the disease by
2030.1Unfortunately, most of the new cases are
expected to come from developing countries.2For
instance, diabetes was the first cause of death in
Ecuador in 2009.3Treatment of diabetes is expensive,
for the year 2000 the total cost to treat the disease was
$132 billion. Considering the limited resources dedi-
cated for heath care in developing countries most
people will not be able to afford current treatments. It is
estimated that approximately 30% of hospital bed-days
in Latin America are used for diabetes related condi-
tions whose cost are an important burden for health
care.2Although there are several drugs that improve the
hyperglycaemic state characteristic of the disease,
there is a constant urge to develop new therapies with
higher efficacy, lower side effects at lower prices to
treat diabetes.4
Lupinus mutabilis a sweet fabaceae is an endemic
legume species of South America. In the Andean
region L. mutabilis beans are known as “chochos”,
“tarwi” or Andean lupin; in Spain the beans are
called “altramuz”. The beans are a traditional food in
Ecuador, Perú and Bolivia. Lupinus mutabilis is rich
in quinolizidine alkaloids and require cooking and
removal of these alkaloids before consumption. The
alkaloid fraction from L.mutabilis has more than 20
different quinolizidine alkaloids and lupanine, 13-
hydroxylupanine, 4-hydroxylupanine, sparteine and
tetrahydrorhombifoline are the most abundant.5
Proteins and alkaloids from lupin are good candi-
dates to be used as hypoglycaemic agents.6,7 Admin-
istration of conglutin-γ, a lupin seed protein, to rats
subjected to glucose overload produces a significant
reduction of plasma glucose with respect to control.6
In vitro studies demonstrate that quinolizidine alka-
loids from lupinus have secretagogue effects in
pancreatic isolated rat islets. Also, studies in vivo in
streptozotocin-induced diabetic rats demonstrated
that 2-thiionospartine a quinolizidine alkaloid from
lupinus lowered plasma glucose concentrations.8
Furthermore, the intravenous administration of the
alkaloid sparteine to non-insulin dependent diabetic
subjects decreased plasma glucose concentrations
and increased insulin concentrations.9Taken
together these data demonstrate that alkaloids from
lupinus could be used as hypoglycaemic agents to
treat type-2 diabetes.
One of the main problems with the use of alkaloids
from lupinus spp is their toxicity. The main side
effects after a toxic dose of alkaloid consumptions
are neurologic such as loss of motor coordination and
muscle control. The lethal dose of lupinus spp alka-
loids in humans is approximately 30 mg/kg of body
weight.10 There are limited data regarding the toxic
dose of these alkaloids in humans. However, in the
study by Paolisso et al., described above the authors
used a dose of 240 mg of sparteine sulphate in moder-
ately overweight individuals without apparent acute
toxicity.9The therapeutic dose for sparteine sulfate
ranges from 75 to 600 mg/day.11 Secondary data indi-
cate that toxic doses of quinolizidinic alkaloids for
human adults start at 25 mg/kg of body weight.12
Considering the hypoglycaemic effects of lupinus
from alkaloids as well as their potential toxicity
shown in previous work, the objective of the present
study was to evaluate the effect of raw Lupinus muta-
bilis consumption on plasma glucose and insulin
concentrations in healthy volunteers as well as in
dysglycemic individuals.
Subjects and methods
The Human Subjects Protection Committee at the
Universidad San Francisco de Quito, Quito Ecuador,
approved this study. Each participant signed an
Informed Consent form after receiving explanation of
the study and its possible consequences.
Subjects and study design
A phase II clinical trial was conducted in two
different populations. Young healthy normal weight
individuals constituted the first group (group-one).
Most participants in group-one were third year medical
students. The second group was formed by volunteers
with a glucose level at initial screening greater than 100
mg/dL (group-two). Most of the volunteers in group-
two were either first-degree relatives of patients with
type-2 diabetes or overweight/obese subjects. Most
participants in group-two were recruited at the
Endocrinology Unit in a health care center (Unidad
Municipal de Salud del Norte). Within each of the indi-
cated populations a control group that received raw soy
was conformed for comparison. Eligible subjects were
randomly assigned to the groups in a proportion of
three (lupinus) to one (soy).
Source of Lupinus mutabilis and Soy
Lupinus mutabilis is the most abundant variety of
lupinus and a popular legume consumed in Ecuador.
Lupinus mutabilis line 450 was harvested between
September and October of 2007 from Cotopaxi, which
is an Andean province of Ecuador.12 The L. mutabilis
was hand harvested when the beans had approximately
17% of humidity. Beans were then dried over a cement
platform for approximately 48 hours until the beans
were 13% water. The dried beans were then placed in
sacs and stored in dry environment at room tempera-
ture 17-20º C. After milling to a 300 micron powder
426 M. Fornasini et al.
Nutr Hosp. 2012;27(2):425-433

(done on site at CC laboratory plant, Ambato Ecuador)
L. mutabilis was encapsulated in gelatin capsules by
CC-laboratories. No fillers or binders were added to the
lupinus powder. Dry soybeans were purchased in a
local supermarket and were processed in a similar
fashion than L. mutabilis beans.
Dose of Lupinus mutabilis
Raw L. mutabilis or soy was administered in
gelatin capsules containing 400 mg of lupine powder
seeds or milled soy. Capsules were identical in
appearance and were kindly prepared by (CC-Labo-
ratorios, Quito- Ecuador). Similar amounts of L.
mutabilis or soy per kilogram of body weight were
administered to volunteers. To estimate the dose of L.
mutabilis, the therapeutic (75 to 600 mg/day) and the
toxic (25 mg/kg) doses of alkaloids present in L.
mutabilis were consi dered.11,10 Thus, it was estimated
that 3.1 mg/kg of body weight was safe to carryout
the trial. The amount of grain containing 3.125 mg of
alkaloid was 98.4 mg.
The amount of alkaloid present in raw L. mutabilis
was measured as previously described and is indicated
in table I.13
Anthropometric measurements
A standardized clinic weight scale was used to deter-
mine both height and weight of each participant.
Weight and height were taken with light clothing on,
but shoes off.14
Glucose, insulin and homeostasis model assessment-
estimatedinsulin resistance (HOMA) determinations
Serum glucose was measured using a glucose
oxidase method (Roche-Diagnostics) using a Hitachi
Roche-917 full-automated analyzer system.
Serum insulin was determined using an electro-
chemiluminicense immunoassay (ECLIA) following
the manufacturer’s instructions (Roche/Hitachi,
Quito-Ecuador) and chemiluminescent emission
was measured using a fully automated analyzer
system ELECSYS 2010 (Roche Diagnostics, Quito-
Ecuador).
Both glucose and insulin determinations were
carried out in NetLab Laboratory. NetLab maintains an
internal and external quality control system (College of
American Pathologists, Brazilian Society of Clinical
Pathologists, SBOC).
The HOMA 2 values were calculated using a modi-
fied formula of Wallace et al, fasting plasma glucose
(mg/dL) multiplied by fasting serum insulin (mU/ml)
and divided by 405.15 The same calculation was done
using serum glucose and insulin at 60 and 90 minutes
after lupinus or soy intake.
Blood samples
Qualified personnel took blood samples in the
morning between 8 am and 10 am after 12 hours of
fasting. Samples were centrifuged within 2 hours
and processed immediately. Two additional samples
were collected at 60 and 90 minutes after treat-
ments. Hypoglycemic effects of lupine alkaloids
have been observed within two hours of alkaloid
administration.9
Treatments
Based on the reported toxic doses of L mutabilis
alkaloids (25 mg/kg) a dose of 3.1 mg/kg of body
weight was considered safe and consequently was used
in this study. Control group received soy in an amount
that matched L mutabilis intake.
Statistics
Descriptive statistics such as means and standard
deviation and percentages were calculated. Due to the
nature of the data non-parametric statistics were used.
Kruskal Wallis was used to test statistical significance
for difference between groups. Paired Wilcoxon
Signed Ranks Test was used to test for differences
within treatment groups.
Results
Demographics and general
characteristics of volunteers
Data indicate that there were not statistically signifi-
cant differences within normoglycemic and dysg-
lycemic groups in the demographic/anthropometric
variables. However, there were statistically significant
Hypoglycemic effect of Lupinus
mutabilis in human volunteers
427Nutr Hosp. 2012;27(2):425-433
Table I
Alkaloid content of Lupinus mutabilis
Type of sample Alkaloid content
Raw grain with skin 3.26 ± 0.135 Content in
percentage (% W/W*)
Cooked grain with skin 0.75 ± 0.02
Cooked grain without skin 0.3034 ± 0.008
Water used to cook raw grain 2.09 (% V/V**)
*W/W: Weight/weight.
** V/V: Volume/volume.

differences between normoglycemic and dysglycemic
individuals in all the variables indicated in table II.
Effect of lupinus mutabilis on blood glucose
and insulin in healthy volunteers
Twenty-one healthy volunteers were randomly
assigned to consume lupinus (N = 16) or soy (N = 5).
Subjects were admitted at Hospital de los Valles after
an overnight fasting. After a short interview and
anthropometric measurements, a blood sample was
drawn from volunteers. Subsequently, a calculated
dose (refer to subjects and methods) of L mutabilis or
soy was administered to each participant. After treat-
ments two blood samples at 60 and 90 minutes were
obtained. Blood was processed as indicated before and
serum glucose and insulin were measured. Table III
shows the mean concentrations of these parameters
before and after treatment with L mutabilis or soy. Data
in table III indicates that consumption of L mutabilis or
soy did not induced statistical significant changes in
blood glucose and insulin after 60 and 90 minutes
neither within nor between treatment groups of healthy
volunteers.
Subsequently, the potential hypoglycemic effects of
L mutabilis were evaluated in volunteers with
increased concentrations of fasting glucose that were
not receiving any hypoglycemic treatment.
Effect of Lupinus mutabilis on blood glucose
and insulin in dysglycemic volunteers
Volunteers for this phase of the study were first-
degree relatives of patients with type-2 diabetes or
overweight/obese subjects without a previous diag-
nosis of diabetes. Most participants in this group were
recruited in an outpatient endocrinology unit. Twenty-
four volunteers were randomly allocated to consume L
mutabilis (N = 17) or soy (N = 7) and were treated simi-
larly than the healthy volunteer group that was normo-
glycemic described above.
Comparisons within groups indicate that consump-
tion of L mutabilis decreased blood glucose concentra-
tions after 60 minutes of ingestion but the decrease was
not statistically significant table IV. However, L muta-
bilis ingestion produced a statistical significant
decrease of serum insulin after 90 minutes of treatment.
No other statistical significant changes were observed
after L mutabilis treatment. On the other hand soy
intake did not cause statistical significant changes in
either glucose or insulin during the study table IV.
A close analysis of the data indicated that the greatest
variability in glucose and insulin concentrations was
observed in individuals that consumed L mutabilis and
whose basal glucose concentration was greater than 100
mg/dL. The statistical analysis of the variations in blood
glucose and insulin in subjects with basal glucose
greater than 100 mg/dL is shown in table V.
428 M. Fornasini et al.
Nutr Hosp. 2012;27(2):425-433
Table II
Demographics compared by group
Age Gender Weight Height BMI WC
Treatments (years) Female
μ
U ± SD
μ
U ± SD
μ
U ± SD
μ
U ± SD
(%) (kg) (m) (cm)
Normoglycemic L. mutabilis 21.8 ± 2,6 4/16 (25) 64.8 ± 11.2 1.71 ± 0.1 22.1 ± 2.9 77.5 ± 8.1
Soy 23.8 ± 0.8 2/5 (40) 68.2 ± 21.0 1.73 ± 0.1 22.5 ± 4.4 79 ± 14.6
Dysglycemic L. mutabilis 55.8 ± 12.0 12/17 (70.6) 74.9 ± 17 1.54 ± 0.01 31.5 ± 6.2 98 ± 13
Soy 58.4 ± 4.9 5/7 (71.4) 87.3 ± 17 1.56 ± 0.01 36.1 ± 6.5 111.2 ± 18.4
Table III
Comparison of blood glucose and insulin concentrations in normoglycemic volunteers treated with lupinus mutabilis or soy
Glucose Glucose Glucose
0 min 60 min 90 min P-value P-value
Treatments μ± SD μ± SD μ± SD 0 vs. 60 min 0 vs. 90 min
(mg/dL) (mg/dL) (mg/dL)
Lupinus (N = 16) 82.8 ± 6.2 81.3 ± 6.6 80.3 ± 5.2 0.36 0.08
Soy (N = 5) 83.4 ± 5.8 80.0 ± 9.3 81.4 ± 9.7 0.35 0.89
Insulin Insulin Insulin
0 min 60 min 90 min P-value P-value
μU ± SD μU ± SD μU ± SD 0 vs. 60 min 0 vs. 90 min
(mg/dL) (mg/dL) (mg/dL)
Lupinus (N = 16) 7.3 ± 3.7 7.8 ± 3.6 6.5 ± 4 0.44 0.37
Soy (N = 5) 10.4 ± 9.7 10.5 ± 9.6 6.8 ± 6.8 1.00 0.14

Table V shows that in subjects with basal glucose
≥ 100 mg/dL that consumed L mutabilis had statisti-
cally significant lower concentrations of glucose after
60 and 90 minutes after treatment. In addition, insulin
concentrations were also significantly lower than basal
levels after 90 minutes of L mutabilis intake. On the
other hand, consumption of soy did not result in signifi-
cant changes in either, glucose or insulin levels over
the study time figure 1 panels C and D.
Differences between treatment groups
Comparison between treatment groups that consumed
L mutabilis with those that consumed soy showed that
there were not statistically significant differences in the
levels of blood glucose before and after treatments.
However, it is notorious that in severe disglycemic indi-
viduals that consumed L mutabilis there were statistically
significant decreases in glucose levels which lasted up to
90 minutes figure 1 panel A whereas in the group that
consumed soy there were not statistically significant
decreases in blood glucose figure 1 panel C.
Regarding insulin concentrations, data show that
after 60 minutes of treatments there was a statistically
significant decrease in insulin concentration in the L
mutabilis compared with the soy group (p = 0.039),
figure 1 panels B vs. D. Panels B and D also show that
in volunteers with dysglycemia, insulin concentrations
decreased with L mutabilis consumption whereas
insulin concentration in the soy group increased.
Furthermore, there was a progressive inhibitory effect
of L mutabilis in serum insulin that was dependent on
basal glucose concentrations. The higher the basal
blood glucose the greater the L mutabilis inhibitory
effect on insulin level at 60 and 90 minutes post treat-
ment figure 1 panels B vs. D.
In order to assess changes in insulin resistance in
treatment groups the Homeostatic model assessment
(HOMA) was calculated. HOMA indicates the degree
of insulin resistance (IR) based on the levels of fasting
glucose and insulin.15 It is well known that only fasting
glucose and insulin concentrations are used to estimate
HOMA. However, in order to evaluate the potential
effect of L mutabilis consumption on insulin resistance
the HOMA-IR was also calculated at 60 and 90
Hypoglycemic effect of Lupinus
mutabilis in human volunteers
429Nutr Hosp. 2012;27(2):425-433
Table IV
Comparison of blood glucose and insulin concentrations in dysglycemic volunteers treated with lupinus mutabilis or soy
Glucose Glucose Glucose
0 min 60 min 90 min P-value P-value
Treatments μ± DE μ± DE μ± DE 0 vs. 60 min 0 vs. 90 min
(mg/dL) (mg/dL) (mg/dL)
Lupinus (N = 17) 107.5 ± 14.4 103.4 ± 7.2 105.3 ± 6.8 0.08 0.51
Soy (N = 7) 102.3 ± 6.4 100.6 ± 4.8 102.9 ± 7.0 0.35 0.92
Insulin Insulin Insulin
0 min 60 min 90 min P-value P-value
μU ± DE μU ± DE μU ± DE 0 vs. 60 min 0 vs. 90 min
(mg/dL) (mg/dL) (mg/dL)
Lupinus (N = 17) 13.3 ± 16.0 9.9 ± 8.4 9.3 ± 7.7 0.12 0.00
Soy (N = 7) 15.7 ± 8.9 18.4 ± 9.2 11.7 ± 7.0 0.06 0.18
Table V
Comparison of blood glucose and insulin concentrations in dysglycemic volunteers treated with lupinus mutabilis or soy
with basal glucose levels ≥ 100 mg/dL
Glucose Glucose Glucose
0 min 60 min 90 min P-value P-value
Treatments μ± SD μ± SD μ± SD 0 vs. 60 min 0 vs. 90 min
(mg/dL) (mg/dL) (mg/dL)
Lupinus (N = 12) 114.2 ± 11.6 105.4 ± 5.6 107.8 ± 5.6 0.00 0.03
Soy (N = 5) 105.2 ± 5.0 102.4 ± 4.3 104.2 ± 6.7 0.28 0.46
Insulin Insulin Insulin
0 min 60 min 90 min P-value P-value
μU ± SD μU ± SD μU ± SD 0 vs. 60 min 0 vs. 90 min
(mg/dL) (mg/dL) (mg/dL)
Lupinus (N = 14) 15.1 ± 18.6 11.0 ± 9.6 10.0 ± 8.9 0.15 0.00
Soy (N = 5) 12.2 ± 5.5 14.4 ± 5.2 8.4 ± 4.8 0.23 0.23

minutes after treatments. Values of HOMA-IR greater
than 2.5 are abnormal and indicate glucose intolerance
or type-2 diabetes.16,17 Data indicate that in subjects
with normoglycemia the HOMA values were main-
tained in the normal range after L mutabilis or soy
consumption, figure 2 panels A and C. On the other
hand, in individuals with dysglycemia that consumed L
mutabilis there was a decrease in HOMA-IR 60 and 90
min after its intake figure 2 panel B. In the group with
dysglycemia that consume L. mutabilis the mean value
of HOMA-IR went from a frank abnormal basal value
to normal values at 60 and 90 minutes of treatment
figure 2 panel B. Furthermore, subjects with dysg-
lycemia that consumed soy showed initially an
increase after 60 minutes and subsequently a decrease
after 90 minutes in HOMA-IR values figure 2 panel D.
In this last group, HOMA-IR was always maintained in
the abnormal range.
Some patients that received L mutabilis experi-
enced dizziness, hypotension and blurred vision. Ten
out of 33 subjects that consumed L mutabilis reported
transitory dizziness, 2 were part of group with
normoglycemia and 8 were subjects from the group
with dysglycemia. None of the volunteers that
consumed soy reported any side effects. Two of the
individuals with dysglycemia that reported dizziness
also had hypotension and one of them had also
blurred vision.
Discussion
Present data show that consumption of L mutabilis
by normal weight healthy young individuals did not
significantly change blood glucose and insulin levels.
However, consumption of similar doses of L mutabilis
by individuals with dysglycemia (fasting glucose > 100
mg/dL) decreased significantly blood glucose and
insulin levels. L mutabilis effect was greater in those
subjects with higher basal glucose levels. Glucose
lowering effects of L mutabilis were not observed in a
control group receiving soy. In addition, a statistically
significant reduction in insulin levels was observed in
the L mutabilis group compared with the soy group after
60 minutes of treatment. Furthermore, only treatment
with L mutabilis improved insulin resistance in subjects
with dysglycemia whereas soy treatment did not. These
data demonstrate that L mutabilis consumption has
important metabolic effects.
Previous in vitro and in vivo work indicates that
conglutin-γand alkaloids from lupinus spp cause
important metabolic effects, which are in agreement
430 M. Fornasini et al.
Nutr Hosp. 2012;27(2):425-433
Fig. 1.—Levels of blood glucose and insulin in volunteers treated with lupinus mutabilis or soy. Panel A glucose and b insulin, lupinus
group. Panel C glucose and D insulin, soy group. NG = normoglycemic; DG = dysglycemic (do not use color, IR).
GLUCOSE INSULIN
A
C D
B
150
100
50
0
120
100
80
60
40
20
0
20.0
15.0
10.0
5.0
0.0
50.0
40.0
30.0
20.0
10.0
0.0
NG
DG ≥ 100
DG ≥ 110
DG ≥ 120
NG
DG ≥ 100
DG ≥ 110
NG
DG ≥ 100
DG ≥ 110
NG
DG ≥ 100
DG ≥ 110
DG ≥ 120
06090
min min min
06090
min min min
06090
min min min
06090
min min min

with the present report. Conglutin-γ, a protein from
lupinus has shown specific binding to insulin and
hypoglycaemic effects in rats subjected to glucose
overload.6The mechanisms of action of this protein for
its hypoglycaemic effects have not been determined.
Garcia Lopez et al., reported that quinolizidinic alkaloids
from lupinus spp induced insulin release from cultured
pancreatic islets from normal rats.7This secretagogue
effect was greater in islets cultured with high glucose
concentrations than in islets cultured with low glucose.
However, some alkaloids, like 2-thionosparteine,
induced insulin secretion even in those islets cultured
in medium with low glucose. On the other hand, alka-
loids such as lupanine, the most abundant alkaloid in L
mutabilis, only induced insulin release by the islets that
were cultured in high glucose concentrations.7In the
present study, raw L mutabilis, with its entire alkaloid
content decreased blood glucose levels only in individ-
uals with high levels of glucose. This observation was
evident in a volunteer with fasting glucose of 143
mg/dL whose glucose dropped to 112 mg/dL after 90
min of lupinus consumption. This represents a decrease
of 22% in blood glucose.
In an in vivo model of diabetes induced by streptozo-
tocin the intra-peritoneal injection of quinolizidine
alkaloids provoked a decrease in blood glucose
concentrations.8In that study not all alkaloids showed
the same hypoglycemic effects. Thus, in diabetic
animals treated only with lupanine there was not a
decrease in blood glucose while in those animals
treated with 2-thionosparteine there was a significant
decrease in blood glucose at 90 and 120 min after alka-
loid administration.8This decrease in blood glucose
was similar to the observed after treatment with gliben-
clamide, a commonly drug used to decrease blood
glucose. However, administration of quinolizidine alka-
loids to non-diabetic rats did not change blood glucose
concentrations.8Regarding insulin, only administration 2-
thionosparteine significantly increased plasma insulin
levels in non-diabetic animals. In that study alkaloid
treatment increased blood insulin levels although the
increase was not statistically significant.8
Clinical studies in healthy volunteers or with indivi -
duals with type-2 diabetes have shown that intravenous
administration of lupinus spp alkaloids decrease blood
glucose and increase insulin. Thus, administration of
sparteine sulphate (240 mg i.v.) decreased blood glucose
after 20 minutes of infusion to normal subjects.11
Glucose decrease was accompanied with an increment
of serum insulin that also occurred 20 minutes after
infusion. These results do not agree with the animal
model of diabetes discussed previously where after
alkaloid administration there were no changes in either
glucose or insulin levels in animals that did not develop
diabetes.8Present results also indicate that oral admin-
istration of raw L mutabilis did not affect significantly
glucose or insulin levels in normal healthy volunteers.
The limited changes observed in blood glucose and
insulin could be due to the fact that participating volun-
teers were young healthy individuals. In normal
Hypoglycemic effect of Lupinus
mutabilis in human volunteers
431Nutr Hosp. 2012;27(2):425-433
Fig. 2.—Homeostatic model asessment (HOMA) in volunteers treated with lupinus or soy. Panel A normoglycemic subjects, lupinus
group. Panel B dysglycemic subjects, lupinus group. Panel C normoglycemic subjects, soy group. Panel D dysglycemic subjects, soy
group.
A
CD
B
1,600
1,500
1,400
1,300
1,200
1,100
2,500
2,000
1,500
1,000
0.500
0.000
HOMA0 HOMA 60 HOMA 90
HOMA0 HOMA 60 HOMA 90 HOMA0 HOMA 60 HOMA 90
5,000
4,000
3,000
2,000
1,000
0,000
5,000
4,000
3,000
2,000
1,000
0,000
HOMA0 HOMA 60 HOMA 90

subjects blood glucose concentrations are finely regu-
lated. Close regulation of blood glucose maintains a
physiologic constant range of glucose between 70 a
100 mg/dL.
On the other hand, Paolisso G. et al, have reported
that intravenous administration of sparteine sulphate
(240 mg i.v.) to subjects with type-2 diabetes
decreased blood glucose and increased insulin.9In
that study, changes in glucose and insulin were
evident after 40 minutes of intravenous alkaloid infu-
sion. In the present study, there was also a decrease in
blood glucose in volunteers with dysglycemia after 60
minutes of lupinus consumption. Regarding insulin,
in contrast to the observations of Paolisso G. et al, in
the present study there was not an increase in insulin
levels; on the contrary there was a statistical signifi-
cant decrease of the hormone after 90 minutes of
treatment. A potential explanation of these differ-
ences could be the fact that in the present study insulin
was measured only at two time points 60 and 90
minutes after L mutabilis intake while Paolisso et al
measured insulin every 10 minutes during one hour.9
This allowed them to measure the putative peak of
insulin after the release from the pancreatic β-cell.9In
addition, the type of treatments used in the present
and the indicated studies could explain the different
results on the levels of insulin. In the indicated studies
with normal individuals as well as with patients with
type-2 diabetes, purified lupinus spp alkaloids were
used intravenously while in the present study raw L
mutabilis beans were orally administered.11,9 Purified
alkaloids and raw lupinus could have a distinct phar-
macokinetics. Raw lupinus beans contain several
alkaloids and micro and macronutrients that could
modulate: a. the secretagoge activity of alkaloids; b.
the potential effect on insulin receptor; and c. insulin
half-life. Insulin is metabolized in the liver after its
release from the pancreas in a relative short time,
approximately 3.5 minutes. Insulin that reaches the
periphery, muscle and adipose tissue, after engaging
its receptor is internalized by the cells and subse-
quently degraded.18 During this process target cells
are stimulated to allow glucose uptake through
glucose transporter GLUT-4. Both events insulin
uptake and glucose transport inside the cells could
explain present observations of glucose and insulin
decrease after L mutabilis consumption. In addition,
the improvement in HOMA-IR observed in subjects
with dysglycemia that consumed L mutabilis could
also be the result of a decrease in blood glucose and
insulin levels. Future pharmacokinetics studies on L
mutabilis and its components such as proteins and
alkaloids, could clarify their effect on pancreatic β-
cells as well as insulin target cells.
There were no severe side effects after raw L muta-
bilis consumption both in normal or volunteers with
dysglycemia. However, some individuals with dysg-
lycemia that consumed lupinus showed transient dizzi-
ness, hypotension and blurred vision. These side
effects have been reported in individuals with poly-
morphism for the enzyme “debrisoquin hydroxylase”
that is responsible of the alkaloid sparteine metabo-
lism.19 Regardless the capacities to metabolize lupinus
alkaloids, 75% of these are excreted unchanged in the
urine within hours of consumption.20 Raw lupinus
consumption without medical supervision must be
discouraged to avoid potential toxicity and undesirable
side effects.
Further studies are needed to determine if lower
doses of raw L mutabilis decrease blood glucose and
insulin. At the moment, there is evidence that the effect
lasts at least 90 minutes which was the longest time
measured in the present study. In addition studies that
compare the effect of raw lupinus with the putative
effect of isolated proteins and alkaloids are also
needed.
Acknowledgements
We would like to thank all participating volunteers.
We also thank Maria Fernanda Loaiza, Cristina
Chavez, Sara Cifuentes, Jorge Reyes, Paulina Armas,
Patricio Rojas. Carlos Cobo (CC-Laboratories) for
their technical assistance. This work received partial
financial support from Chancellor Grants Universidad
San Francisco de Quito.
References
1. Smyth S, Heron A. Diabetes and obesity: the twin epidemics.
Nat Med 2006; 12 (1): 75-80.
2. Yach D, Stuckler D, Brownell KD. Epidemiologic and
economic consequences of the global epidemics of obesity and
diabetes. Nat Med 2006; 12 (1): 62-66.
3. INEC [database on the internet]. Anuario de Nacimientos y
Defunciones 2009. [cited 2010 Sept 20]. Available from:
http://www.inec.gov.ec/web/guest/publicaciones/anuarios/inv
_soc/nac_def
4. Pearson ER. Pharmacogenetics and future strategies in treating
hyperglycaemia in diabetes. Front Biosci 2009; 14: 4348-62.
5. Hatzold T, Elmadfa I, Gross R, Wink M, Hartmann T, and
Witte L. Quinolizidine alkaloids in seeds of Lupinus mutabilis.
J Agric Food Chem 1983; 31 (5): 934-8.
6. Chiara Magni C, Fabio Sessa F, Elena Accardo E, Marco
Vanoni M, Paolo Morazzoni P, Alessio Scarafoni A et al.
Conglutin, a lupin seed protein, binds insulin in vitro and
reduces plasma glucose levels of hyperglycemic rats. Review
Article J Nutr Biochem 2004, 15 (11): 646-650.
7. García López PM, de la Mora PG, Wysocka W, Maiztegui B,
Alzugaray ME, Del Zotto H et al. Alkaloids isolated from
Lupinus species enhance insulin secretion. Eur J Pharmacol
2004; 504 (1-2): 139-42.
8. Keskin M, Kurtoglu S, Kendirci M, Atabek E, Yaz C.
Homeostasis Model Assessment Is More Reliable Than the
Fasting Glucose/Insulin Ratio and Quantitative Insulin
Sensitivity Check Index for Assessing Insulin Resistance
Among Obese Children and Adolescents. Pediatrics 2004;
115 (4): e500-3.
9. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher
DF, Turner RC. Homeostasis model assessment: insulin resis-
tance and β-cell function from fasting plasma glucose and insulin
concentrations in man. Diabetologia. 1985; 28 (7): 412-9.
432 M. Fornasini et al.
Nutr Hosp. 2012;27(2):425-433

10. Bobkiewicz-Kozłowska T, Dworacka M, Kuczy ski S, Abram-
czyk M, Kolano R, Wysocka W, Garcia Lopez PM, Winiarska
H. Hypoglycaemic effect of quinolizidine alkaloids—lupanine
and 2-thionosparteine on non-diabetic and streptozotocin-
induced diabetic rats. Eur J Pharmacol 2007; 565 (1-3): 240-4.
11. Paolisso G, Sgambato S, Passariello N, Pizza G, Torella R,
Tesauro P et al. Plasma glucose lowering effect of sparteine
sulphate infusion in non-insulin dependent (type 2) diabetic
subjects. Eur J Clin Pharmacol 1988; 34 (3): 227-32.
12. Australia New Zealand Food Authority. Lupin Alkaloids In
Food A Toxicological Review And Risk Assessment Technical
Report Series No. 3. 2001 November [cited 2010 September
29]. Avalaible from:http://www.foodstandards.gov.au/_src
files/TR3.pdf
13. Sgambato S, Paolisso G, Passariello N, Varricchio M,
D’Onofrio F. Effect of sparteine sulphate upon basal and
nutrient-induced insulin and glucagon secretion in normal man.
Eur J Clin Pharmacol 1987; 32 (5): 477-80.
14. Jarrín P. Caracterización y tratamiento del agua del desamar-
gado del chocho (Lupinus mutabilis Sweet) proveniente de la
planta piloto de la Estación de Santa Catalina [dissertation].
INIAP; 2003.
15. Von Baer, D, Reimerdes, E. and Feldheim, W. (1979). Meth-
oden zur Bestimmung der Chinolizidin-Alkaloid In: Lupinus
mutabilis. I. Schnellmethoden. Z. Lebenzm. Unters, Forsch,
169, 27-31.
16. Yepez R, Carrasco F, Baldeón ME. Prevalence of overweight
and obesity in Ecuadorian adolescent students in the urban area.
Arch Latinoam Nutr 2008; 58 (2): 139-143.
17. Wallace TM, Levy JC, Matthews DR. Use and abuse of HOMA
modeling. Diabetes Care 2004; 27 (6): 1487-95.
18. Kolman P, Pica A, Carvou N, Boyde A, Cockcroft S, Loesch A
et al. Insulin uptake across the luminal membrane of the rat
proximal tubule in vivo and in vitro. Am J Physiol Renal Phy -
siol 2009; 296 (5): F1227-37.
19. Relling MV. Polymorphic drug metabolism. Clin Pharm 1989;
8 (12): 852-63.
20. Petterson DS, Greirson BN, Allen DG, Harris DJ, Power BM,
Dusci LJ et al. Disposition of lupanine and 13-hydroxylupanine
in man. Xenobiotica 1994; 24 (9): 933-41.
Hypoglycemic effect of Lupinus
mutabilis in human volunteers
433Nutr Hosp. 2012;27(2):425-433