New Concepts in Diabetes and Its Treatment - part 9 pdf

cause of the Charcot foot and most patients have a dense neuropathy, but
good circulation. Early animal experiments suggested that walking on an
insensitive limb could lead to joint destruction. Excessive and repetitive stress
to bones leads to microfractures, which render the bone more brittle and could
lead to bone and joint destruction. However, the degree of bone destruction
often seen in the absence of major injury has suggested an underlying bone
abnormality. Diabetic neuropathy leads to an increase in bone blood flow,
which may promote osteoclastic activity and bone resorption. Indeed, a small
study has demonstrated increased serum markers of osteoclastic action in
patients with acute Charcot that was not accompanied by a concomitant
increase in markers of osteoblastic activity. Furthermore, lower limb bone
mineral density has been found to be lower in patients with a Charcot foot,
when compared to neuropathic controls. A full understanding of the patho-
logical process leading to the often dramatic and progressive destruction seen
in this condition has not yet been arrived at, and as it is rare and usually
presents late, the opportunities for further studies are limited.
In the early acute stages, when bone turnover is high, treatment involves
rest and immobilization of the foot (usually with a total contact cast) in an
attempt to reduce the metabolic activity within the bone. There is now some
evidence that bisphosphonate drugs given during this acute phase may shorten
the duration of the acute phase presumably by reducing the bone turnover
directly, and hence slowing down the process which weakens the bone and
renders it susceptible to further fracture and fragmentation.
Whenever unilateral swelling ofthe foot is present insomeone with diabetic
neuropathy, Charcot neuroarthropathy must be considered. Plain X-ray is
usually adequate to make or exclude the diagnosis, and while the radiological
appearances can be similar to those of osteomyelitis, in the absence of a source
of infection (such as an overlying ulcer), neuroarthropathy is nearly always
the cause.
Conclusion
Although the roles of peripheral neuropathy and peripheral vascular dis-

ease are now well established as the main aetiological factors in diabetic foot
ulceration, there is much work to be done in both the way in which ulcers
develop and the interactions of the main risk factors with each other and with
all the other risk factors discussed in this chapter. However, this complexity
should not deter the clinician, as it is now very clear that simple clinical tests
will identify patients at risk of ulceration and amputation, and appropriate,
but simple education about footcare can greatly reduce the likelihood of
216Shaw/Boulton
developing diabetic foot problems. Foot ulcers can be difficult to heal, but
treatment is likely to be successful in the vast majority of cases when pressure
is removed, callus and necrotic tissue debrided, infection controlled, and a
good circulation is maintained.
Suggested Reading
Boulton AJM, Connor H, Cavanagh PR (eds): The Foot in Diabetes, ed 2. Chichester, Wiley, 1994.
Caputo GM, Cavanagh PR, Ulbrecht JS, Gibbons GW, Karchmer AW: Assessment and management of
foot disease in patients with diabetes. N Engl J Med 1994;331:854–860.
Schapper NC, Bakker K (eds): The diabetic foot. Diabet Med 1996;13(suppl 1):1–64.
Shaw JE, Boulton AJ: The pathogenesis of diabetic foot problems: An overview. Diabetes 1997;46(suppl 2):
58–61.
Dr J. Shaw, Department of Medicine (M7), Manchester Royal Infirmary,
Manchester M13 9WL (UK)
Tel. +44 161 276 4452, Fax +44 161 274 4740, E-Mail
217Foot Problems in Diabetes
Chapter XVI
Belfiore F, Mogensen CE (eds): New Concepts in Diabetes and Its Treatment.
Basel, Karger, 2000, pp 218–228

Erectile Dysfunction in Diabetes and
Its Treatment
M. Tagliabue, G.M. Molinatti

Dipartimento di Medicina Interna, Universita
`
degli Studi di Torino,
Ospedale Molinette, Torino, Italy
Introduction
The existence of sexual disorders in diabetes mellitus has long been recog-
nized. In the pre-insulin era, impotence was considered one of the commonest
symptoms of diabetes, being present in both severe and milder forms of the
disease. However, only now are sexual function problems receiving their right-
ful attention, as the medical professional has moved from a mere ‘survivalist’
approach to diabetes and its more invalidating complications towards care for
the diabetic individual in all his or her complexity.
Today, diabetology is no longer satisfied to keep diabetics in reasonably
good health, but also addresses everything that may affect the individual’s
quality of life and, from this standpoint, sexuality cannot fail to occupy a
role of primary importance. In the diabetic male, it is sexuality in the narrowest
sense, namely what is conventionally designated ‘potentia coeundi’, that is
compromised and it is precisely in relation to that situation that a comprehen-
sive overview of this condition can provide the diabetic patient with the answer
he seeks.
Epidemiology
Erectile dysfunction is 3 times more common among diabetics than in
the healthy control population. However, the complication is still considered
occult since it is often unreported by patients. In the various studies published,
the incidence of this dysfunction in diabetics varies from 28 to 59%. The
predictive factors are: age, duration of the disease, degree of metabolic com-
218
pensation, the presence of microvascular complications (especially retinopathy)
and neuropathy, high blood pressure and the drugs taken for that condition,
smoking and alcohol abuse. The age factor is particularly important. The very

earliest epidemiological studies showed that the incidence of impotence in
diabetic males rose from 1.5% in the under-40s to 25% in the 40–60 age group.
More recently, others have reported incidences rising from 15% in the under-
40 group to 55% at 60. The trend revealed by the Wisconsin Epidemiologic
Study in particular is highly significant (p=0.0001), rising from 1.1% in the
21–30 age group to 47.1% among insulin-dependent diabetics over 43 years
old. Klein himself confirms that significance when he analyses the duration
of diabetes: men with a more than 35-year history of the disease are 7.2 times
more likely to present this complication than those with only a 10- to 14-year
history. However, the link between erectile dysfunction and disease duration
is not inevitable, since the erectile disorder occasionally appears before the
clinical onset of the diabetes. In addition, diabetic erectile dysfunction is also
related to HbA
1c
levels which indicate the ability to metabolize glucose, the
risk of impotence tripling in those worst affected (HbA
1c
?9.8%). That in-
creased risk is explained when we consider the treatment needed to control
diabetes – restrictive diets and drugs to lower blood sugar and/or insulin –
that are most aggressive in the most metabolically compromised patients. Both
diabetic retinopathy, especially if severe, and neuropathy, both peripheral and
autonomic, are related to a higher incidence of erectile dysfunction which is
5.3 times more likely to occur in such patients. High blood pressure is an
additional risk factor especially when treated by certain drugs such as -
blockers, methyldopa and particularly diuretics. Finally, excessive drinking
and smoking intensify the risk of erectile dysfunction in diabetics.
A recent Italian study of 9,868 diabetic patients reported a 35.8% incidence
of erectile dysfunction and confirms the reported literature data: incidence
increasing with age, duration of diabetes, severity of failure to metabolize

glucose, complexity of diabetic therapy, diabetic complications (angiopathy,
retinopathy, kidney disease, neuropathy) as well as cardiovascular disease and
the use of certain drugs in its treatment and finally habitual smoking.
Apart from erectile dysfunction, diabetes can also produce problems with
ejaculation, especially retrograde ejaculation as the so-called ‘dry orgasm’ (a
dysfunction of the autonomic and somatic nervous system) which occurs in
1–4% of male diabetics, most particularly in those with the longest history of
the disease and those who are most metabolically compromised.
Ejaculation without orgasm and indeed failure to achieve ejaculation
(reflecting a compromised sympathetic nervous system) are also commoner
among diabetics than in the general population, accounting for 8% of ejacula-
tory disorders. By contrast, the incidence of premature ejaculation is almost
219Treatment of Erectile Dysfunction in Diabetes
Table 1. Possible causes of erectile dysfunctions in diabetics
Psychological causes
Poor glycaemic control
Vascular alterations (macro-/microangiopathies and venous)
Autonomic neuropathy
Endothelial alterations (reduced NO secretion)
Concomitant pathologies and related drugs
identical in the diabetic and the healthy populations, a finding that confirms
the psychosomatic pathogenesis of that disorder.
Be that as it may, the organic character of diabetes and its complications
should not lead us to forget the influence of psychological factors which may
at times be preponderant.
Etiopathogenesis
Diabetic erectile dysfunction is often a complex problem given its psy-
chogenic and organic components, the latter linked to the failure to metabolize
glucose and the related organic complications, not to mention the pathological
conditions known to be caused by the drugs used to treat those complications

(table 1). However, many studies have confirmed that erectile dysfunction is
primarily organic in origin, since the dysfunction is rarely reversible. In mon-
itoring the nocturnal erections associated with REM sleep, researchers have
found fewer REM sleep-erections in diabetic males, a finding which supports
the view that impotence in diabetics is more likely to be organic than psycho-
logical in origin. Nevertheless, the role of hormonal abnormalities in the
physiopathology of organic erectile dysfunction remains controversial. It is
therefore vital to examine the fundamental psychological and organic factors
involved in the sexual function of diabetics.
The Psychological Factor
There is substantial evidence to suggest that erectile dysfunction in dia-
betes is often psychological in origin. The main contributory factors are:
awareness of suffering from a chronic condition, relationship problems and
the fear of failure during sexual intercourse as a result of that situation. It is
not clear whether such psychological factors are greater in the diabetics affected
than in the general population of people with erectile dysfunctions, though
diabetics with theproblem appear more stressed than thoseunaffected. Accord-
220Tagliabue/Molinatti
ing to some authors, diabetic patients are more fearful of developing erectile
dysfunctions as a complication of their condition than of going blind, but
however great their concern, they are unlikely to discuss it with their doctor.
In fact only 50% of erectile dysfunction sufferers report it to their physician.
Impaired Glucose Metabolism
It is important for the correct management of diabetic erectile dysfunction
to repair any severe metabolic disorder the patient may be suffering from. Pa-
tients may recover normal erectile function, as soon as insulin injections to cor-
rect their impaired glucose metabolism are started. The impaired delivery of
oxygentothe tissue,causedbytheformation ofglycosylatedhaemoglobin which
has a greater affinity foroxygen increases vascular permeability,depositinglipo-
protein on the vessel wall and this may be the cause of the vascular damage.

Vascular Alterations
Vascular disorders cause impotence in 18% of diabetic males. Haemo-
dynamic disturbances in diabetics may be either arterial (macro- and/or micro-
angiopathies) or venous given the direct communication between the two
vascular systems.
Macroangiopathy causes major arterosclerotic obstructions of the large,
medium-sized and small arterial blood vessels which cut off the blood flow
to the corpora cavernosa. Many authors claim that the primary cause of
impotence in diabetics is of vascular origin and atherosclerosis is, in fact, the
earliest lesion on the peripheral arteries of the penis. Histological findings of
pathological alterations to the small arteries are also reported to occur before
any neurological damage. Later, neurological damage caused by the same
atherosclerotic processes appears on the vasa nervorum. Such vascular altera-
tions are the result of proliferating endothelial and intimal cells, fragmentation
of the endothelium, calcium deposits, and perivascular fibrosis. Perineural
fibrosis may occur without causing any direct damage to the nerve fibres.
There is an equally close link between ischaemic heart disease and diabetic
erectile dysfunction, both being caused by the ischaemic vasculopathy affecting
both areas. Diabetic microangiopathies produce alterations and irregularities
in the local microvascular blood flow. Those alterations concern: endothelial
cell metabolism and function; the basal membrane of vessel walls, which
are thickened; oxygen transportation; the characteristics of blood flow and
haemostasis.
In the venous system the uncontrolled blood flow and the failure of the
arteriovenous anastomoses may also contribute to the erectile dysfunction.
Finally, venous occlusions in diabetes may be due to a structural alteration
in the fibroelastic components of the trabeculae.
221Treatment of Erectile Dysfunction in Diabetes
Neurological and Endothelial Alterations
The penile nerve system is both autonomic and somatic and the relaxation

of the smooth muscle tissue of the corpus cavernosum results from the inter-
action of three systems: adrenergic, cholinergic and VIPergic. Other factors
like nitric oxide (NO) initially called endothelium-derived relaxing factor
(EDRF) and produced by constitutive nitric oxide synthetase (cNOS) which
increases the concentration of intracellular cyclic guanosine monophosphate
(cGMP) may well be involved in the relaxation of the smooth muscle in the
corpus cavernosum. As we know, the earliest studies into NO production by
cNOS were carried out on bioptic samples taken from the corpus cavernosum
tissue of impotent diabetics who presented reduced acetylcholine vasodilation.
The reduced NO production and consequent reduction in intracellular cGMP
probably leads to an increase in intracytosolic calcium that is responsible for
the contraction in smooth muscle cells. In diabetics there may well also be a
reduction in the noradrenaline, VIP and acetylcholine content of the corpus
cavernosum and both the cholinergic fibres and their ability to synthesize
acetylcholine may be reduced over time, as may the VIPergic pathways. Cho-
linergic stimulation is certainly known to increase NO production. This would
compromise both the neurogenic and the endothelial mechanisms dependent
on the relaxation of the cavernosal smooth muscle. Increased levels of endoth-
elin-1, a powerful vasoconstrictor released by the endothelial cells, have also
been found in patients with erectile dysfunction, especially those with diabetes.
That finding suggests that endothelial dysfunction may contribute to erectile
dysfunction and that in the absence of any significant vascular element the
increase in plasmatic endothelin-1 may be related to early atherosclerosis. The
autocrine role of this peptide, which causes the smooth muscle cells of the
corpus cavernosum to proliferate and/or contract, has been confirmed in
experimentally induced diabetes mellitus.
Somatic and autonomic neuropathy (bladder dysfunction) is often associ-
ated with impotence in diabetes and is responsible for 67% of cases, according to
recent statistics. The disease causes axonaldegeneration of the nerves in the penis
(and other parts of the body) together with thickening of the basal membrane.

The biochemical abnormalities encountered in diabetic patients can be
somewhat improved iftheir glucose metabolism problem is carefully controlled.
Researchers have also found a lack of coordination in the electrical activity
of the corpus cavernosum in diabetics with a consequent loss of the diminished
or absent activity that is normal in the tumescent or erectile phase.
Hormonal Alterations
Total basal testosterone levels have been found to be normal or low and
researchers have also documented a diminished response in terms of absolute
222Tagliabue/Molinatti
testosterone increases after HCG stimulation. Some reports describe a decrease
in the free fraction of testosterone and estradiol, which they attribute to a
marked increase and/or enhanced binding capacity in SHBG and/or inappro-
priate gonadotropin secretion. In addition, there also appears to be an altera-
tion in gonadic response to tropine stimulation with a tendential rise in basal
LH and a more protracted increase after GnRH stimulation.
Theevidenceoncirculatinggonadotropinandprolactinisconflicting.Some
report an increase in urinary LH among diabetics with primary organic impo-
tence, as well as a reduction in levels of free testosterone. Diabetes and obesity
often go together and increased aromatization of testerostene in the adipose
tissue produces an increase in oestrogen levels thatcontributes to erectile failure.
In actual fact, there is increasing doubt about the role of steroids and
other hormones in the aetiopathology of sexual disorders in the diabetic male
and the variations that may be found do not appear to play a particularly
important part in the genesis of this complication.
Spermatogenesis
The most frequently encountered alteration in spermatogenesis is reduced
spermatozoa motility which appears to be closely linked to the metabolic
disorders and the presence of autonomic neuropathy. Functional damage to
the seminal vesicles is probably a major contributory factor. Studies conducted
on rats with streptozotocin-induced diabetes revealed an alteration in the

animals’ sexual behaviour and a reduction in the weight of their secondary
sex glands, in their production of androgens and in spermatogenesis as a result
of altered gonadotropin pulsatility.
Diagnosis
Erectile dysfunction is diagnosed in diabetics in much the same way as in
their healthy counterparts and diagnosis therefore includes anamnestic assess-
ment, objective examination, laboratory tests and instrumental investigation
(table 2). Both the rigidity and the duration of penile erections are affected (re-
duced arterial blood flow and altered control of the autonomic nervous system
over the penile circulation) while, at least initially, the libido remains unaffected.
Anamnesis
Anamnesis is the first step. In diabetics, erectile dysfunction usually arises
insidiously, evolving slowly but inexorably. Physicians should pay particular
223Treatment of Erectile Dysfunction in Diabetes
Table 2. Main diagnostic procedures for erectile dysfunction in diabetics
Anamnesis (physiology, pathology, pharmacology, sexual)
Clinical examination
Blood chemical analyses (HbA
1c
, T, PRL, TSH)
Instrumental investigations (penile biothesiometry and cardiovascular tests)
Vascular evaluation (office-intracavernosal injection test)
attention to their patient’s smoking and drinking habits and to the presence
of any concomitant pathologies (high blood pressure and dyslipidaemias are
common) and the drugs they are being treated with, since these often have a
negative impact on sexual activity itself. In the case of diabetes mellitus, it is
essential to know type, duration and treatment, how far glucose metabolism
is compromised and whether or not the patient presents with micro/macroangi-
opathic and/or neuropathic complications.
In order to obtain a full diagnostic picture, the investigation of organic

factors must be accompanied by investigation of learning, intrapsychic, dyadic,
systemic and sociocultural factors.
The sexual anamnesis will cover aspects like the presence or absence of
sexual desire, the presence of spontaneous erections on awakening and/or in
response to visual stimuli and/or erotic thoughts and/or physical stimulation
by a partner, as well as the quality and frequency of sexual intercourse, the
presence of any significant changes in recent months and the description of
the sexual intercourse itself.
Instrumental Investigations
The instrumental investigations indicated include those used to examine
the vascular system (ultrasound scans or basal and dynamic Doppler echo-
sonography) which usually involves the intracavernosal injection of prosta-
glandin (PGE
1
) and penobrachial plethysmography. Others examine
neurological aspects (vibration and heat perception thresholds, autonomic
cardiovascular tests, peroneal motor conduction velocity and sural sensitivity
tests; possibly also sacral evoked potential tests). Then there are urological
tests (cavernosography, cavernosometry, bulbocavernosus reflex). A new min-
imally invasive test has recently been proposed and has been tested on type
2 diabetics with no vascular disease in whom basal penile tumescence was
assessed using the Rigiscan technique and found to be related to autonomic
nerve damage.
224Tagliabue/Molinatti
An easy-to-use protocol for the diagnosis of level I erectile dysfunctions
in diabetics was proposed by the Italian Diabetology Society’s ‘Diabetic Neu-
ropathy’ Study Group in 1996:
Anamnestic screening
Targeted questionnaire
Penile biothesiometry

Cardiovascular tests
PGE
1
drug stimulation
Patients =40 years old: 5 g
Patients ?40 years old: 10 g
If the erectile response is
absent: repeat after 1 week with 10 or 20 g PGE
1
persistently absent: refer to your andrology unit for further investigation
(Rigiscan, Ultrasound, Doppler ultrasonography, invasive vascular tests)
If the erectile response is
present: oral and/or intracavernosal and/or psychological treatment.
The protocol proposed by Japanese authors is more complex. In it, vascu-
lar investigations, Rigiscan and audiovisual sexual stimulation precede intra-
cavernosal drug stimulation. If there is no erectile response, nocturnal penile
erectile tumescence is monitored, which, if found, allows us to label our patient
as ‘not suffering from any organic disorder’.
Treatment
The treatment for erectile dysfunction in diabetics (table 3) is primarily
based on rectifying the glucose-metabolizing disorder by diet and/or drug
treatment and by persuading the patient to abstain from risk factors like
smoking and alcohol abuse. In addition, every possible effort will be made to
find substitutes for any drugs with a negative impact on sexual function.
Psychological Treatment
Psychotherapy can help to minimize anxiety and modify the couple’s
sexual habits in a helpful fashion. Even so, ‘psychological problems’ appear
to be no more common among diabetics than among the general population.
Drug Treatment
Treatment with 

2
-antagonists (yohimbine) or 
1
-antagonists (doxazosin,
terazosin, etc.) can enhance penile vasodilation but cannot alone induce and
225Treatment of Erectile Dysfunction in Diabetes
Table 3. Treatment of erectile
dysfunction
Psycho- or behavioural therapy
Drug treatment
Yohimbine

1
-antagonists
Local nitroderivatives
Sildenafil
Endourethral alprostadil
Intracavernosal injection
Papaverine
Alprostadil
Phentolamine
Moxisylyte
External mechanical support
Vacuum device
Revascularization surgery
Venous
Arterial
Penile prosthesis
maintain erection. These drugs are indicated in patients with high blood
pressure and certain prostate pathologies. Moderate success has been obtained

with topical nitroderivatives that are rapidly absorbed through the skin (nitro-
glycerin). These act directly by stimulating the release of the adenylase cyclase
which relaxes the smooth muscle tissue.
Sildenafil, a selective inhibitor of type V cyclic phosphodiesterase in the
corpus cavernosus, prevents the breakdown of cGMP and therefore acts on
the NO mechanism (NO/cGMP) that plays a dominant part in the relaxation
of smooth muscle tissue and hence penile erection. This is a rapidly absorbed
drug that is taken 60 min before intercourse and has an effect lasting about
4 h. It does not trigger erection as such but improves its quality by promoting
the smooth muscle relaxation initiated by NO release. It is important to
remember that Sildenafil is contraindicated in men taking nitrates, given the
risk of significant reductions in blood pressure caused by intensification of
the vasodilatory effects of such drugs. Caution must also be exercised in a
number of other clinical conditions (kidney or liver problems) and/or when
the patient is taking other drugs which might interfere with absorption kinetics.
In doses of 25–100 mg, Sildenafil has proved effective in 59% of diabetic
patients with erectile dysfunction (placebo 15%). It is well tolerated and the
226Tagliabue/Molinatti
side effects reported in this population are: headache, dyspepsia and flushing.
Other side effects described are temporary and relate to the perception of
colours, sensitivity to light and fuzzy vision.
Alprostadil (asyntheticpreparation of PGE
1
), also appears as a compound
that is incorporated into pellets for intraurethral application (MUSE: Medical
Urethral System for Erection). Patients have to be taught how to use the
applicator which delivers a microsuppository into the urethra at an average
dose of 500–1,000 g. The side effects are penile pain, stranguria and slight
bleeding. MUSE is contraindicated without the use of a condom in intercourse
with a partner who is pregnant or liable to conceive. It is not widely used

in Italy and may be contraindicated in diabetics with their greater risk of
infections.
Intracavernosal Injection
Intracavernosal pharmacoprosthesis is currently one of the most widely
used treatments and enjoys good patient compliance, especially among dia-
betics who get a satisfactory response in 66% of cases compared to only 23%
in nondiabetics with erectile dysfunction. Vasodilatory substances either alone
or in combination (papaverine, alprostadil, phentolamine and atropine) are
inoculated directly into the corpus cavernosum. The patient has to be trained
in the inoculation procedure and doses must be carefully selected in order to
avoid prolonged erections or priapism. Erection occurs about 10 min after
inoculation of the substance. The incidence of complications depends on the
type of drug used. Pain on the inoculation site is rare, but penile fibrosis or
priapism may develop over time. The most commonly used drug is alprostadil
because of its minimal side effects and the dose varies from 5 to 20 g. This
prostaglandin produces cAMP which acts with cGMP to relax the smooth
musculature. Both are metabolized by the type V phosphodiesterase found in
both the penile smooth muscle tissue and the eyes. Moxisylyte is a selective

1
drug used in 10- to 20-g doses. Its minimal side effects include penile pain
and prolonged erections. Vasointestinal peptide combined with phentolamine
is now in the advanced research stage. Its side effects are flushing and tachycar-
dia and, rarely, penile pain.
External Mechanical Support
A patient who prefers not to use drugs can employ a so-called ‘vacuum
device’. This is a cylinder into which the penis is inserted. The cylinder is
pressed against the abdominal wall and a mechanical or electric pump is
activated to create a vacuum which draws blood into the corpus cavernosum
thereby producing a penile erection. The erection is maintained by a rubber

ring around the base of the penis for as long as the ring remains in situ.
227Treatment of Erectile Dysfunction in Diabetes
Patient compliance with this system is generally good. However the device
should not be used for more than 30 min at a time, since it can create a feeling
of chill in the penis. Furthermore, ejaculation is generally impeded and this
may make for a less satisfying orgasm. The vacuum device can be used to
potentiate the effect of drug treatment.
Revascularization Surgery
Where erectile dysfunction is caused by vascular disease, both venous and
arterial revascularization procedures are a possibility.
Penile Prothesis
Diabetic patients with irreversible penile dysfunctions are candidates for
penile implants, which, however, expose them to the risk of local infections.
Suggested Reading
Bancroft J, Gutierrez P: Erectile dysfunction in men with and without diabetes mellitus: A comparative
study. Diabet Med 1996;13:84–89.
Dunsmuir WD, Holmes SAV: The aetiology and management of erectile, ejaculatory, and fertility problems
in men with diabetes mellitus. Diabet Med 1996;13:700–708.
Feldman HA, Goldstein I, Hatzichristou DG, Krane RJ, McKinlay JB: Impotence and its medical and
psychosocial correlates: Results of the Massachusetts male aging study. J Urol 1994;151:54–61.
Klein R, Klein BEK, Lee KE, Moss SE, Cruickshanks KJ: Prevalence of self-reported erectile dysfunction
in people with long-term IDDM. Diabetes Care 1996;19:135–141.
Saenz de Tejada I, Goldstein I, Azadzoi K, Krane RJ, Cohen RA: Impaired neurogenic and endothelium-
mediated relaxation of penile smooth muscle from diabetic men with impotence. N Engl J Med
1989;320:1025–1030.
Takanami M, Nagao K, Ishii N, Miura K, Shirai M: Is diabetic neuropathy responsible for diabetic
impotence? Urol Int 1997;58:181–185.
Prof. G.M. Molinatti, Dipartimento di Medicina Interna, Universita
`
degli Studi di Torino,

Ospedale Molinette, I–10126 Torino (Italy)
Tel. +39 (0)11 6635318, Fax +39 (0)11 6634751, E-Mail
228Tagliabue/Molinatti
Chapter XVII
Belfiore F, Mogensen CE (eds): New Concepts in Diabetes and Its Treatment.
Basel, Karger, 2000, pp 229–240

Multifactorial Intervention in
Type 2 Diabetes mellitus
Peter Gæde, Oluf Pedersen
Steno Diabetes Centre, Copenhagen, Denmark
Introduction
The prevalence of type 2 diabetes mellitus is rapidly increasing. Patients
with type 2 diabetes suffer from micro- as well as macrovascular complications,
the latter causing the excess mortality seen in these patients compared to the
background population.
Several risk factors for the outcome of type 2 diabetes have been identified
in prospective epidemiological studies. However, until recently the treatment of
type 2 diabetes has been empirical rather than evidence based from randomized
intervention studies. Although the diagnosis of diabetes is based on blood
glucose levels, it is important to realize that patients with type 2 diabetes
mellitus share many clinical features with the metabolic syndrome such as
dyslipidaemia, hypertension, hyperinsulinaemia and an increased risk of
cardiovascular disease. In cardiovascular medicine a multifactorial treatment
approach of several risk factorsfor cardiovascular disease isgenerally accepted.
We suggest a similar approach in the treatment of type 2 diabetes mellitus
based on the results from several intervention studies in patients with this
disease.
Evidence for the Treatment Effect of Hyperglycaemia
The importance of strict glycaemic control for the development and pro-

gression of microvascular complications was definitely confirmed in the United
Kingdom Prospective Diabetes Study (UKPDS). 3,867 patients with newly
diagnosed type 2 diabetes were randomized to intensive therapy with oral
hypoglycaemic agents or insulin, or to conventional therapy. Patients were
229
followed for a median of 10 years. Mean HbA
1c
in the intensive group over
this period was 7.0% as compared to a mean value of 7.9% in the conventional
group. A significant risk reduction of 25% for microvascular events with inten-
sive therapy was found while only a borderline significant effect of 16%
(p>0.05) was seen on myocardial infarctions and no effect was seen on mortal-
ity. The number of patients needed to treat over a 5-year period to prevent
one complication was 39.2.
Even though no overall effect on cardiovascular disease was seen in the
UKPDS, it is still important to remember that hyperglycaemia is one of
the most important predictors for cardiovascular disease in epidemiological
studies. There is therefore every reason to believe that intervention against
hyperglycaemia also has beneficial effects on this complication in type 2 dia-
betes mellitus, however the final proof from a randomized intervention study
has still to come.
Treatment Approach for Hyperglycaemia
Treatment goals for different risk factors are shown in table 1. Tradition-
ally, lifestyle modification is the first and basic treatment approach, since
this is cheap and without major side effects. Even though pharmacological
treatment may be needed, it is important to emphasize the necessity of this
basic treatment to patients. Since deterioration of -cell function is inevitable,
regular controls are necessary in order to change treatment whenever needed.
Diet
An immediate effect on dieting on blood glucose levels is seen in most

studies. However, the UKPDS demonstrated that less than half of patients
with a normalization of fasting blood glucose after 3 months of dieting alone
were able to maintain this effect after 1 year. Therefore, in order to reach the
recommended treatment goals, supplemental pharmacological treatment is
needed in most cases.
Exercise
Even 30 min of moderate exercise with an intensity of 50–80% of V
O
2
max
3–4 times a week has beneficial effects both on carbohydrate metabolism and
on insulin sensitivity. However, many patients suffer from complications, e.g.
cardiovascular complications, arthroses or peripheral neuropathy, which
should be taken into account when recommending type of exercise.
Oral Hypoglycaemic Agents
Different classes of oral hypoglycaemic agents with different mechanism
of action exist. Thus, drugs can be combined. It is important to realize that
230Gæde/Pedersen
Table 1. Treatment goals for behaviour modification, and clinical as well
as biochemical variables in type 2 diabetes mellitus
Behaviour modification
Diet
Carbohydrates Percent of total energy intake 55%
Proteins Percent of total energy intake 15%
Fat Percent of total energy intake 30%
Fat composition
Saturated fatty acids 1/3
Monounsaturated fatty acids 1/3
Polyunsaturated fatty acids 1/3
Exercise Moderate, 30 min/day

Smoking No smoking
Clinical
Body mass index =25 kg/m
2
Systolic blood pressure =130 mm Hg
Diastolic blood pressure =85 mm Hg
Biochemical
HbA
1c
(normal range 5.2–6.4%) in the normal range
Fasting total cholesterol =5.0 mmol/l
Fasting LDL cholesterol =2.6 mmol/l
Fasting HDL cholesterol ?1.1 mmol/l
Fasting triglycerides =1.7 mmol/l
Urinary albumin excretion rate =30 mg/24 h
5–10% of patients experience oral hypoglycaemic drug failure per year, in
which case insulin treatment must be initiated to maintain satisfactory gly-
caemic control. In a substudy in the UKPDS, the effect of metformin versus
conventional treatment and sulphonylureas or insulin in obese patients was
examined. Patients randomized to metformin had lower risks for all-cause
mortality, strokes and any diabetes-related endpoint than patients on conven-
tional treatment, sulphonylureas or insulin and metformin was found to cause
less weight gain and fewer hypoglycaemic attacks than sulphonylureas or
insulin. In another substudy in patients with secondary failure to sulphonylu-
reas, addition of metformin doubled the risk of mortality compared to patients
continuing sulphonylureas alone. These results are difficult to interpret because
of the study design. However, until results from a randomized study specifically
designed to answer the question about the effect of metformin on mortality
in secondary failure to sulphonylureas, according to the American Diabetes
Association no restrictions in the use of metformin in combination with

sulphonylureas is justified.
231Multifactorial Intervention in Type 2 Diabetes mellitus
Insulin
Insulin treatment can be given as monotherapy or as combination therapy
with insulin and oral hypoglycaemic agents. From studies comparing several
insulin regimens, bedtime NPH insulin in combination with metformin twice
daily and self-adjustment of insulin dose based on patients’ own fasting blood
glucose measurements seems to be an easy and safe way of obtaining optimal
glycaemic control. This combined insulin and metformin treatment regimen
is associated with less weight gain than monotherapy with insulin. Whereas
hypoglycaemia is frequent in type 1 diabetes mellitus, this is not the case in
insulin resistant type 2 diabetes mellitus, even with large insulin doses and
HbA
1c
values near the normal range. It has been discussed whether insulin
per se is atherogenous, since hyperinsulinaemia is a risk factor for cardiovas-
cular disease in epidemiological studies, however this effect has not been
demonstrated in clinical studies.
Evidence for the Treatment Effect of Hypertension
The effect of antihypertensive treatment in type 2 diabetes was examined
in a substudy in the UKPDS. 1,148 hypertensive patients were randomized
either to tight blood pressure control aiming for a blood pressure =150/85
mm Hg with the use of angiotensin-converting enzyme (ACE) inhibitors or
-blockers or to less tight blood pressure control aiming at a blood pressure
=180/105 mm Hg and if possible avoiding the previous mentioned drugs. The
mean blood pressure during follow-up was 144/82 mm Hg in thegroup assigned
tight blood pressure control compared to 154/87 mm Hg in the group assigned
to less tight control. Significant risk reductions were found for deaths related
to diabetes (32%), strokes (44%) and microvascular disease (37%). The effect
of blood pressure control was independent of blood glucose control. The

number of patients needed to treat over a 5-year period to prevent one diabetes-
related death was 30.0 while it was 12.2 for microvascular disease. Thus based
upon the results from intervention trials it seems that blood pressure control
is more effective than blood glucose control in type 2 diabetes mellitus. Another
important message from this study is that in order to obtain the specified
blood pressure goal it is in most cases necessary with two or more different
antihypertensive drugs.
The results from the Hypertensive Optimal Treatment (HOT) randomized
trial also underline the importance of tight blood pressure control. 18,790
patients with hypertension were randomly assigned to three groups with target
diastolic blood pressure of O90, O85 and O80 mm Hg. A long-acting calcium
antagonist, felodopine, was the primary drug with addition of an ACE inhib-
232Gæde/Pedersen
itor, -blocker or diuretic if blood pressure goals were not obtained. In a
subgroup analysis in the hypertensive diabetic patients, significant risk reduc-
tions in the group assigned to the lowest blood pressure group compared to the
highest was found for cardiovascular mortality (67%) and major cardiovascular
events while (51%) only a borderline significant reduction was found for all-
cause mortality (44%). The number needed to treat to prevent one complication
is in accordance with the numbers found in the UKPDS (table 2).
A positive treatment effect for the elderly hypertensive type 2 diabetic
patients was recently demonstrated in a post-hoc analysis from The Systolic
Hypertension in Europe Trial. Of the originally enrolled 4,695 patients with
systolic hypertension (systolic blood pressure ?160 mm Hg and diastolic
blood pressure =95 mm Hg), all patients ?60 years of age, 492 had diabetes.
Patients were randomized to treatment with a calcium antagonist, nitrendipine,
or placebo. In the group receiving active treatment significant risk reductions
were obtained for total mortality (55%), mortality from cardiovascular disease
(76%), all cardiovascular events combined (69%), fatal and nonfatal strokes
(73%), and all cardiac events combined (63%) compared to the group receiving

placebo. The suspicion that calcium channel blockers may be harmful in
patients with hypertension and diabetes mellitus could clearly not be confirmed
in this study.
The question of which kind of antihypertensive drug treatment should
be used was also investigated in the Captopril Prevention Project (CAPPP).
10,985 patients with hypertension were randomly assigned the ACE inhibitor
captopril or conventional antihypertensive therapy with diuretics or -blockers.
The risk of a major cardiovascular event during an average follow-up of 6.1
years did not differ between treatment groups. An overall reduction in the
rate of fatal cardiovascular events was reported in the diabetic subpopulation
(n>572), however no data on blood pressure control and actual drug regimens
in this subgroup was reported making the results difficult to interpret. Similar,
the previously described UKPDS did not find any difference between the
treatment effect of ACE inhibitors or -blockers in the type 2 diabetic patients
randomized to tight blood pressure control.
Treatment Approach for Hypertension
Basic dietary advice recommending weight reduction in obese, a low
sodium intake and regular exercise are important in the treatment of hyperten-
sion and should be given for 2–3 months before initiating pharmacological
therapy.
Positive treatment effects have been demonstrated with the use of ACE
inhibitors, diuretics, calcium antagonists and -blockers. This clearly suggests
that it is the obtained blood pressure level rather than the type of drug
233Multifactorial Intervention in Type 2 Diabetes mellitus
Table 2. Number of patients needed to treat for 5 years to prevent one
complication in different trials (unless indicated with p value, only significant
results from the different studies are shown; numbers are based onextrapolation
of results from the original follow-up period to a 5-year period)
Trials Number needed
to treat

Hyperglycaemia
UKPDS [UKPDS Group, 1998a]
Any diabetes-related endpoint 39.2
Myocardial infarction (p>0.05) 92.4
Any microvascular endpoint 71.4
Hypertension
UKPDS [UKPDS Group, 1998b]
Any diabetes-related endpoint 12.2
Diabetes-related death 30.0
HOT Study (subgroup analysis) [Hansson et al., 1998]
Major cardiovascular event 16.0
Cardiovascular mortality 27.0
Dyslipidaemia
4S study (subgroup analysis) [Pyo
¨
ra
¨
la
¨
et al., 1997]
Major CHD event 4.8
CARE [Sacks et al., 1986]
Major CHD event (p>0.05) 12.4
Microalbuminuria
Ravid study [Ravid et al., 1993]
Progression to nephropathy 3.4
Aspirin
HOT Study [Hansson et al., 1998]
Major cardiovascular events 125.0
Myocardial infarction 153.8

US Male Physicians’ Health Study [Final report, 1989]
Myocardial infarction 108.2
Multifactorial intervention
Steno 2 Study [Gæde et al., 1999]
Any microvascular endpoint 4.7
Progression to nephropathy 5.4
234Gæde/Pedersen
used that is essential in treating hypertension in type 2 diabetic patients.
As mentioned above, most patients will need treatment with two or more
antihypertensive drugs in order to obtain a desired treatment response.
Evidence for the Treatment Effect of Dyslipidaemia
The term dyslipidaemia covers many patterns of lipid changes from the
normal values. The most common pattern in type 2 diabetic patients is
elevated serum triglyceride levels and decreased serum HDL cholesterol levels.
Evidence for a positive treatment effect of dyslipidaemia in type 2 diabetes
mellitus comes from post-hoc analyses of diabetic patients participating in
larger secondary intervention studies comprising patients with known cardio-
vascular disease. Since the risk of a first myocardial infarction in patients
with type 2 diabetes is the same as the risk for a re-infarction in a nondiabetic
subject, it seems reasonable to extrapolate from the overall results from these
studies.
The Scandinavian Simvastatin Survival Study (4S) included 4,444 patients
with a recent myocardial infarction or angina pectoris and an increased fasting
serum-cholesterol level in the range 5.5–8.0 mmol/l. Of these, 202 had diabetes.
Patients were randomized to treatment with placebo or simvastatin. The me-
dian follow-up time for the diabetic patients was 5.3 years. A significant
risk reduction of 55% with lipid-lowering drug therapy was seen for major
cardiovascular events, while no effect was seen on total and cardiovascular
mortality. The number needed to treat to prevent one major cardiovascular
event during a 5-year period was 4.8.

The Cholesterol and Recurrent Events (CARE) trial was also a secondary
intervention study including 4,159 patients with a fasting serum-cholesterol
level =6.2 mmol/l, thus examining the effect of cholesterol-lowering therapy
in patients with cardiovascular disease and a fasting serum cholesterol level
within the normal range. 586 diabetic patients participated in the study. A
borderline significant risk reduction of25% was found for majorcardiovascular
events (death from cardiovascular disease or nonfatal myocardial infarction,
coronary artery bypass grafting, or percutaneous transluminal coronary angi-
oplasty). No effect on mortality was found.
The Helsinki Heart Study was a primary intervention trial enrolling both
nondiabetic and diabeticpatients with highfasting serum non-HDL cholesterol
levels. The overall result was a significant 34% reduction in the risk for cardio-
vascular disease with treatment with a fibrate (gemfibrozil), however the 38%
risk reduction found in a subanalysis in the type 2 diabetic population was
not significant.
235Multifactorial Intervention in Type 2 Diabetes mellitus
Table 3. Various type of drug therapy of dyslipidaemia
LDL cholesterol Triglyceride lowering Combined
lowering dyslipidaemia
Drug of choice Statins Fibrates Statins
Combination Resins Nicotinic acid or statins Fibrates
It is important to repeat that the evidence for a positive treatment effect
of dyslipidaemia is based on post-hoc analyses from secondary intervention
studies in mixed populations. No primary intervention trials in type 2 diabetic
patients have been published.
Treatment Approach for Dyslipidaemia
Patients with dyslipidaemia should be examined in order to exclude sec-
ondary dyslipidaemia (e.g. renal disease, hypothyroidism), in which case the
underlying cause should be treated.
Behaviour Modification

Diet should be optimized with a reduction in dietary content of saturated
fatty acids. Weight loss and increased physical activity will lead to decreased
serum triglyceride and increased fasting serum HDL cholesterol levels. Max-
imal effect of this approach is a reduction in fasting serum LDL cholesterol
of 0.4–0.6 mmol/l.
Pharmacological Therapy
Glucose-lowering therapy reduces both serum levels of triglycerides and
to a lesser extent serum LDL cholesterol levels, thus optimizing glycaemic
control is prior to treatment with specific dyslipidaemic drugs. The serum
LDL cholesterol level for initiation of drug therapy is still debated, but the
more risk factors the lower level. In high-risk patients the limit is 2.6 mmol/l.
Various types of drug therapy for the different forms of dyslipidaemia are
shown in table 3.
Evidence for theTreatment Effect of Microalbuminuria
Microalbuminuria (urinary albumin excretion rate in the range 30–300 mg/
24 h) is an important risk factor for the development of both micro- and
macrovascular disease. It is known that both the treatment of hyperglycaemia
236Gæde/Pedersen
and hypertension can reduce the albumin excretion rate. However, treatment
with ACE inhibitors seems to reduce urinary albumin excretion rate indepen-
dently of the blood pressure-lowering effect of these drugs. 94 patients with
type 2 diabetes mellitus and microalbuminuria and a blood pressure =140/90
mm Hg were randomized to treatment with 10 mg enalapril daily or placebo.
Over the study period of 5 years, 6 patients in the ACE inhibitor group and
19 in the placebo group developed diabetic nephropathy. This difference could
not be attributed to differences in glycaemic control, body mass index, or
blood pressure values, which were similar in both groups throughout the study
period. Reciprocal plasma creatinine levels as a measure for renal function
decreased significantly in the placebo group as a sign of deterioration of kidney
function but remained stable in the enalapril group.

Treatment Approach for Microalbuminuria
Since both the treatment of hyperglycaemia and hypertension decreases
urinary albumin excretion rate, these treatment modalities should be optim-
ized. In case of increased albumin excretion rate despite sufficient antihyperten-
sive treatment without an ACE inhibitor, treatment with this drug should be
initiated. Treatment with an ACE inhibitor should be started in a low dose
and gradually increased, especially in patients with normotension because of
the risk for orthostatic hypotension.
Evidence for the Treatment Effect of Aspirin
The beneficial effect of low-dose acetylsalicylic acid as secondary preven-
tion of cardiovascular disease is well established in both the diabetic and
nondiabetic population. The previously described HOT trial also examined
the effect of treatment with acetylsalicylic acid in patients with hypertension.
Of the 18,790 patients, half were randomized to 75 mg aspirin daily and the
other half to placebo. No specific data for the diabetic subgroup have been
published. For the whole study population a significant 15% reduction in the
risk for major cardiovascular events was seen, primarily due to a 36% reduction
in the risk for myocardial infarction. Fatal bleeds were equally common in
the two groups, but nonfatal bleeds (primarily gastrointestinal) were signifi-
cantly more frequent among patients receiving acetylsalicylic acid than in
those receiving placebo.
The US Physicians Health Study was a primary prevention trial in which
a low-dose aspirin regimen (375 mg every other day) was compared with
placebo in 22,071 male physicians. There was an overall significant 44% risk
reduction for myocardial infarction in the acetylsalicylic acid-treated group,
237Multifactorial Intervention in Type 2 Diabetes mellitus
whereas subgroup analysis in the diabetic physicians revealed a relative risk
of 0.39 for the diabetic men taking aspirin.
Who Should be Treated with Aspirin
Primary prevention: Consider aspirin therapy as a primary prevention

strategy in type 2 diabetic subjects with at least one of the following criteria:
family history of coronary artery disease; cigarette smoking; controlled hyper-
tension; micro- or macroalbuminuria.
Secondary prevention: Aspirin should be used as a secondary prevention
strategy in all diabetic patients with evidence of large vessel disease.
People with aspirin allergy, bleeding tendency, anticoagulant therapy, on-
going or recent gastrointestinal bleeding, and clinically active hepatic disease
are not to be treated with aspirin. Therapy should be given as enteric-coated
aspirin in doses of 75–325 mg/day.
Evidence for the Effect of Smoking Cessation
Smoking may be one of the most important risk factors for cardiovascular
disease in type 2 diabetes mellitus. This is based onseveral epidemiologicalstud-
ies,wheresmokingisfoundtobeariskfactorforbothall-causemortality,cardio-
vascular disease and stroke in both the diabetic and nondiabetic population.
The Multiple Risk Factor Intervention Trial (MRFIT) randomized 12,866
subjects to a control group or an intervention group with intervention against
several risk factors such as diet, smoking and blood pressure. Smoking cessa-
tion programmes were used with individual advice by a doctor. A 13% reduc-
tion in the number of smokers in the intervention group was seen after 6
years. However, no significant effects on total mortality or mortality from
cardiovascular disease was seen in the overall analyses. Despite these negative
results from one of the most successful smoking cessation intervention pro-
grammes, it is because of the epidemiological evidence recommended that all
diabetic patients quit smoking.
Smoking Cessation Programmes
Trying to convince patients to stop smoking is one of the hardest and
most disillusioning tasks for the health worker. Even though several smoking
cessation programmes and nicotine replacement therapy exist, the percentage
of smokers who quit smoking in these programmes is low. The average rate
of success is around 8% after 1 year in unselected smokers. This number can

be increased by selecting motivated patients (e.g. patients with myocardial
infarctions) but a 1-year stop rate of more than 25% is rarely seen.
238Gæde/Pedersen
Since most patients who actually quit smoking normally have had two
or three unsuccessful smoking cessation periods, it is important to keep on
motivating patients to try another programme even though they have already
failed once or more.
Evidence for the Treatment Effect of Multifactorial Intervention
Several treatment modalities in type 2 diabetes mellitus have been de-
scribed, some apparently more effective than treatment of hyperglycaemia,
thus providing a rationale for multifactorial intervention. The Steno 2 Study
compared the effect of intensive multifactorial intervention with standard
multifactorial intervention. The study population was 160 type 2 diabetic
patients with microalbuminuria. 80 of these patients were randomized to
intensive multifactorial intervention comprising behaviour modification (diet,
exercise, smoking cessation) and polypharmacological therapy targeting sev-
eral risk factors (hyperglycaemia, hypertension, dyslipidaemia, albumin excre-
tion rate). Patients were followed for a mean of 3.8 years yielding significant
risk reductions for both nephropathy (56%), retinopathy (40%), and neuro-
pathy (62%). Extrapolating these risk reductions, the number of patients
needed to treat for a 5-year period to prevent one microvascular complication
in this high-risk diabetic population was only 4.7. No differences between
groups were seen for major cardiovascular events, however follow-up time
may have been too short to detect such a difference because of the small
size of the two treatment groups.
Treatment Approach for Multifactorial Intervention
Multifactorial intervention is a mixture of thetreatment modalities already
described. It is based on the simultaneous treatment of several risk factors.
Frequent and thorough examinations are mandatory, since risk factors may
interact and change thus requiring a shift in treatment strategy. This is also

the case if there is lack of patient compliance, that may be caused by side
effects of the different drugs used.
Side Effects
Since intensive multifactorial intervention requires the use of polyphar-
macological therapy, it is crucial to inform patients about the risk of side
effects. The physician should consider patients’ age, kidney and liver function
and contraindications against the different drugs before starting pharmaco-
logical treatment. Drug interactions may present as a major problem. A
classic example is the masking of hypoglycaemic symptoms with -blocker
239Multifactorial Intervention in Type 2 Diabetes mellitus
treatment, but many drugs also cause increased or decreased response to
oral hypoglycaemic agents. Patients treated with anticoagulants should be
monitored closely when treatment with statins or fibrates, or -blockers is
initiated. Patient compliance decreases with the number of side effects experi-
enced, however informing patients and starting with low doses may diminish
this problem.
Suggested Reading
Final report on aspirin component of the ongoing Physicians’ Health Study. N Engl J Med 1989;321:
1825–1828.
Frick MH, Elo O, Haapa K, et al: Helsinki Heart Study: Primary-prevention trial with gemfibrozil in
middle-aged men with dyslipidemia. N Engl J Med 1987;317:1237–1245.
Gæde P, Vedel P, Parving HH, Pedersen O: Intensified multifactorial intervention in patients with type
2 diabetes and microalbuminuria: The Steno Type 2 randomised study. Lancet 1999;353:617–622.
Hansson L, Lindholm LH, Niskanen L, et al:Effect of angiotensin-converting enzyme inhibition compared
with conventional therapy on cardiovascular morbidity and mortality in hypertension: The Captopril
Prevention Project (CAPPP) randomised trial. Lancet 1999;353:611–616.
Hansson L, Zanchetti A, Carruthers SG, et al: Effects of intensive blood pressure lowering and low-dose
aspirin in patients with hypertension: Principal results of the Hypertension Optimal Treatment
(HOT) randomised trial. Lancet 1998;351:1755–1762.
Multiple Risk Factor Intervention Trial Research Group: Multiple Risk Factor Intervention Trial: Risk

factor changes and mortality results. JAMA 1982;248:1465–1477.
Pyo
¨
ra
¨
la
¨
K, Pedersen TR, Kjekshus J, Færgeman O, Olsson AG, Thorgeirsson G: Cholesterol lowering
with simvastatin improves prognosis of diabetic patients with coronary heart disease. Diabetes Care
1997;20:614–620.
Ravid M, Savin H, Jutrin I, Bental T, Katz B, Lishner M: Long-term stabilizing effect of angiotensin-
converting enzyme inhibition on plasma creatinine and on proteinuria in normotensive type 2
diabetic patients. Ann Intern Med 1993;118:577–581.
Sacks FM, Pfeffer MA, Moye LA, et al: The effect of pravastatin on coronary events after myocardial
infarction in patients with average cholesterol levels. N Engl J Med 1996;335:1001–1009.
Tuomilehto J, Rastenyte D, Birkenha
¨
ger WH, et al: Effects of calcium-channel blockade in older patients
with diabetes and systolic hypertension. N Engl J Med 1999;340:677–684.
UKPDS Group: Intensive blood-glucose control with sulphonylureas or insulin compared with conven-
tional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet
1998a;352:837–853.
UKPDS Group: Tight blood pressure control and risk of macrovascular and microvascular complications
in type 2 diabetes: UKPDS 38. BMJ 1998b;317:703–713.
Yki-Ja
¨
rvinen H, Nikkila
¨
K, Ryysy L, Tulokas T, Vanamo R, Heikkila
¨

M: Comparison of bedtime
insulin regimens in NIDDM: Metformin prevents insulin-induced weight gain. Diabetologia 1996;
39(suppl 1):A33.
Dr Oluf Pedersen, Steno Diabetes Centre, Niels Steensens Vej 2,
DK–2820 Copenhagen (Denmark)
Tel. +45 44 43 90 50, Fax +45 44 43 82 32
240Gæde/Pedersen