Vol. 96 (2002) 337^343
Long-term follow-up of untreated patients with
sleep apnoea syndrome
D. FISHER,G. PILLAR, A. MALHOTRA, N. PELED AND P. LAVIE
Sleep Laboratory, Rambam Medical Center, Bruce Rappaport Faculty of Medicine, Technion - Israel Institute of
Technology, Haifa, Israel
Abstract Obstructive sleep apnoea (OSA) is a common disorder with numerous potential sequelae. Although the
majority ofthese consequences can be reduced with appropriate treatment, onlylimited data exist regarding the natural
progression of this disorder in untreated individuals.
We herebyreport along-termfollow-up of alluntreated patients (n¼40) followed-up intheTechnion Sleep Clinic, using
both subjective and objective measurements.In addition, we report a long-term follow-up of11patients who attempted
dietary weight loss.
The average timeintervalbetweenthe first and second polysomnographies for the untreated group was 5?072?8 yrs,
and 2?572?3 yrs for the weight reduction group.There was no significant change in Body Mass Index (BMI) or Respiratory Disturbance Index (RDI) betweenthe two Polysomnographic (PSG) evaluationsinthe untreated patients.However,
eight patients developed hypertension (n¼5) or ischaemic heart disease (IHD) (n¼3) between the two evaluations.RDI,
age and BMI atthe time ofthe initial evaluation were not predictive of changesin RDI, snoring intensityor minimal oxygen
saturation. However, the patients who developed hypertension/IHD had significantly higher RDI than the patients who
did not (46727 vs. 23717 h1, Po0?005). In the weight-loss group, BMI decreased by a mean of 3?1kg m2, and RDI
decreased by 20 events h1, Po0?05 for both.There was a significant correlation between the weight loss and improvement in RDI (R¼0?75, P¼0?005).
We conclude that in untreated obstructive sleep apnoea patients RDI does not necessarily increase over time,
but associated hypertension or ischaemic heart disease may develop.When weight loss is successfully achieved, sleep
apnoea significantly improves with a high correlation between the extent of weight loss and the improvement in apnoea
status.r 2002 Elsevier Science Ltd
doi:10.1053/rmed.2001.1277, available online at http://www.idealibrary.com on
Keywords obstructive sleep apnoea syndrome; natural history; long-term follow-up; weight loss; hypertension; ischaemic
heart disease
INTRODUCTION
Obstructive sleep apnoea (OSA) is a common disorder
characterized by the repetitive collapse of the pharyngeal airway during sleep. Its prevalence is 2^ 4%, and the
potentially associated sequelae include reduced quality
of life, increased risk of motor vehicle accidents, hypertension, and possibly an increased frequency of myocardial infarction and stroke. While treatment is associated
with a reduction in these consequences, the literature
regarding the natural progression of the disorder as well
Received19 April 2001and accepted in revised form 26 November 2001.
Correspondence should be addressed to: Peretz Lavie, Sleep
Laboratory, Bruce Rappaport Faculty of Medicine, Technion - Israel
Institute of Technology, Gutwirth Building, Haifa 32000, Israel.
Fax: (972)- 4 - 8323045; E-mail: plavie@tx.technion.ac.il
as long-term follow-up of therapy is relatively sparse. As
treatment of OSA can be poorly tolerated, the risk and
bene¢t of no therapy is of clinical importance. The studies that have addressed this issue, were usually either
subjective (questionnaire based) or with relatively
short-term follow-up. The results of these revealed con£icting conclusions.
In a retrospective study of 118 OSA patients, Lugaresi
et al. reported that typical patients had snored for many
years prior to development of OSA.They concluded that
this is a progressive disease which starts with simple
snoring and evolves to full-blown OSA syndrome with
increased severity (1). Likewise, Martikainen et al. found
in a population-based questionnaire study that the prevalence of self-reported snoring and pauses between
snoring increased in a 5 yr follow-up study (2). In the
338
few studies performed to evaluate the outcome of
untreated OSA patients objectively, two found that
mild^ moderate OSA may indeed progress over a relatively short period of time (3,4), while two others
reported that the severity of OSA is stable over
time (5,6).
In the non-apnoea elderly population the ¢ndings are
also inconclusive. Phoda et al. (7) and Bliwise et al. (8) reported mild increases in apnoea indices over time (3 and
5 yrs, respectively) in these populations. In contrast, Ancoli-Israel et al. (9) found no change with high variability
of RDI in a triple evaluation study over an 8?5 yrs
period.
In a recent study, Lindberg et al. (10) reported a longterm follow-up of 29 men with untreated snoring and
daytime somnolence. A repeat PSG 10 years after initial
evaluation revealed that 18 had similar RDI between the
two studies, while in 11 the RDI increased by more than
5 h1. While in the ¢rst evaluation only one man had an
RDI of above 10 h1, after 10 yrs eight men fell into this
category. This increase in RDI of snorers could not be
predicted by any of the baseline measurements (e.g.
RDI, BMI, age, or smoking). The group in whom RDI
deteriorated did not di¡er from the‘stable’ group in worsening of EDS or development of HTN. This study, however, was limited by the population selection of only
snorers and not patients with OSA.
As there are only few data in the literature regarding
the natural history of apnoea indices in OSA patients,
based mainly on relatively short-term follow-up, we
planned this study to evaluate changes in apnoea severity
and potential sequelae in all untreated patients followedup in our clinic. We hypothesized that patients who
choose not to be treated will have relatively mild OSA,
and that they would demonstrate deterioration over
long-term follow-up both subjectively and objectively.
For comparison, we also present data of follow-up of 11
patients who underwent a dietary weight loss program,
with somewhat shorter follow-up time.
METHODS
Subjects
All untreated patients who were followed-up in our clinic
and underwent re-evaluation polysomnography (PSG),
were included in the study. In addition, followed-up patients who were in our clinic after participation in a dietary program of weight loss were included as well. These
patients who participated in weight reduction programs
were instructed to return for a follow-up study once
they reached a stable target weight. Patients who underwent a surgical procedure (nasal: septoplasty/turbinectomy, or pharyngeal: LAUP/UPPP) or were on continuous
treatment with either continuous positive airway pres-
RESPIRATORY MEDICINE
sure (CPAP) or oral appliance were excluded. All untreated patients in the study were recommended to be
treated but chose not to due to either non-compliance
with CPAP/dental device or refusal of surgery or both.
None of the patients had major otolaryngologic pathology on ENT examination. All the participants were required to evaluate subjectively the spontaneous changes
which occurred between the two sleep studies regarding
snoring intensity and daytime somnolence. Each patient
underwent a medical interview including completing
sleep questionnaires, Epworth Sleepiness Scale, brief
physical examination including weight and height measurements by the same sleep physician in our lab, and a
full night PSG sleep study.Concomitant diseases (speci¢cally, HTN and IHD) were de¢ned based on diagnosis
made by other specialists (PCP, Cardiologist), regardless
of this study, and were retrieved from the patients’
primary care physician records.
Polysomnography
All patients underwent two full-night PSG studies, with
the identical equipment (Nihon Kohden polygraph, EEG4214, Tokyo, Japan). Recordings consisted of electroencephalography (EEG) (C3-A2), electrooculography (EOG),
submental electromyography (EMG), electrocardiography (ECG), £ow (thermistor), respiratory e¡ort (thoracic belt), and leg movements. Data were recorded both
on paper (with a paper speed of1cm sec1), and on computer. Bedtime was between 22:00 ^23:00 and 06:00 (all
patients were awakened at 06:00). In addition, quantitative snoring intensity and arterial oxygen saturation
levels were continuously recorded. An oxygen saturation sensor was placed on the ¢fth ¢nger of each subject,
and a microphone was placed one meter above the
patient’s bed.
Scoring and de¢nitions
All scoring was performed based on standard criteria
(11). Apnoea was de¢ned as the complete cessation of air£ow for at least 10 sec. Hypopnoea was de¢ned as any
decrement in £ow, if it was associated with either arousal (de¢ned as the appearance of 3 sec of alpha rhythm in
the EEG or an increment in submental EMG signal) or
oxygen desaturation (of 3% or more). Respiratory Disturbance Index was calculated by dividing the total
number of apnoeas + hypopnoeas by total sleep time.
The minimal oxygen saturation was determined as the
lowest saturation value that was associated with a respiratory event (i.e. excluding artifactually low values
due to movements). Likewise, maximal snoring intensity
was determined as the maximal decibels recorded by the
dB-meter, associated with snoring (i.e. excluding cough
or other noises).
NATURAL HISTORYOF SLEEPAPNOEA
339
Subgroups of patients who ‘improved’ or ‘worsened’
were de¢ned as a reduction or increment of 25% or
more of the initial RDI, if the change was at least 10
events per h (i.e. to avoid calling a change from an RDI
of 10 to 13 ‘worsening’. In this example ‘worsening’ would
require at least 20 events per h). These arbitrary de¢nitions were prespeci¢ed.
Data analysis
All statistical analyses were performed with commercially available software (Excel 97, Microsoft; and SigmaStat, SPSS, Chicago). All data are presented as mean71
SD unless otherwise stated. Two-tailed t-tests for independent samples were used to compare variables between groups or subgroups, and paired t-test to
statistically assess the impact of time within patient
groups. In each case, Po0?05 was taken to indicate
signi¢cance.
RESULTS
Summary data for the untreated and weight loss patients
is shown inTable1.The average time interval between the
¢rst and second polysomnographies for the untreated
group was 572?8 yrs, and 2?572?3 yrs for the weight
reduction group. In the weight-loss group, BMI decreased
by a mean of 3?1kg m2, and RDI decreased by 20
events h1, Po0?05 for both (Table1).There was a signi¢cant correlation between the weight loss (BMI) and improvement in RDI (R¼0?75, P¼0?005). In the untreated
group, there was no signi¢cant change in BMI or RDI
(Table 1). There was no signi¢cant correlation between
changes in weight (BMI) and in RDI in the untreated
group. Minimal oxygen saturation in the untreated
patients worsened (88710 vs. 85711%, Po0?05), and
maximal snoring intensity increased (6078 vs. 6578 dB,
Po0?05). Eight untreated patients developed hypertension (n¼5) or ischaemic heart disease (n¼3) over the
time course of 5 yrs between the two evaluations
(Table 1). None of the patients in the weight
reduction group developed HTN or IHD during the
study period.
The Epworth Sleepiness Scale (ESS) score did not
di¡er between the two evaluations (9?274?5 and
9?474?9, respectively). Although 19 patients reported
worsening of their complaints, there was no signi¢cant objective change in their ESS score (8?774?2
and 10?675?0 on the ¢rst and second evaluations,
respectively).
The severity of the disorder, patients’ BMI or age at
the time of the initial evaluation could not predict
changes in RDI, snoring intensity or minimal oxygen
TABLE 1. Comparison of measurements between untreated and weight loss patients, and between primary evaluation and
long-term follow-up evaluation
Variable
Untreated
Weight loss
Number of patients (M/F)
Age at ¢rst study (yrs)
Time between studies (yrs)
BMI (kg m2)
40 (36/4)
47710
5?072?8
28?974?8
29?474?6
27721
28721
88710
85711*
6078
6578*
6 (15)
11 (28)
2 (5)
5 (13)
0
22
18
11 (11/0)
46713
2?572?3
33?374?5
30?273?6*
46717}
26716*
82710
8576
67710
6575
4 (36)
4 (36)
0
0
3
7
1
RDI (events h1)
Minimal Oxygen Sat (%)
Maximal Snoring intensity (dB)
Hypertension Number (%)
IHD Number (%)
Subjective change (No. of patients)
First
Second
First
Second
First
Second
First
Second
First
Second
First
Second
Improved
Unchanged
Worsened
Values are mean7SD.
BMI: body mass index;RDI:Respiratory Disturbance Index;IHD: Ischaemic Heart Disease; dB: decibeles.Values for hypertension,
IHD and subjective changes are presented as number of patients (percentage in brackets).
*Po0?05 Between ¢rst and follow-up studies.
wPo0?05 between groups.
340
RESPIRATORY MEDICINE
TABLE 2. Change in RDI in di¡erent subgroups of the untreated patients based on severity,BMI and age
Category
Subgroup classi¢cation
based on ¢rst study
RDI
Mild (0^20)
Moderate (21^40)
Severe (440)
Lean (o25)
Overweight (25^30)
Obese (430)
o40
40^50
450
BMI
Age
Number (Patients
in category)
Time between
studies (mo)
RDI At ¢rst
evaluation
RDI at second
evaluation
23
10
7
7
20
13
8
19
13
54737
62733
81722
27711?6
66733?7
71735?8
45736?2
64737?7
66729?4
13?475?0
33?078?3
64?7718?6
13?473?3
23?3716?2
40?8726?9
12?877?5
26?1720?0
37?9724?1
16?575?9
28?7710?9
61?6724?6
12?773?4
20?1712?9
39?7728?3
16?074?1
25?4717?3
39?2+26?9
All the patients presented in this table were untreated.Time between the two evaluations is presented in months.In none of the
subgroups (in all categories) was the change in RDI between the two PSG evaluations signi¢cant.
TABLE 3. Comparison between untreated patients in whom RDI worsened, and those in whom it improved
Group (All untreated)
Improved
Worsened
BMI kg m2
Age (yrs)
n
5
6
RDI events hour1
First
Second
First
Second
First
Second
56714
4578
63714
5278
3074
3074
3073
3073
4579
21722
26711
38725
TABLE 4. Comparison between untreated patients with newly diagnosed HTN/IHD and the others
Group (All untreated)
HTN/IHD
Others
n
8
32
Age (yrs)
BMI kg m2
RDI events hour1
Oxi min (%)
First
Second
First
Second
First
Second
First
Second
5779*
4578
6379*
5079
3375*
2874
3275
2975
46727*
23717
46731*
23715
78712*
9078
81711
85711
The patients who developed HTN/IHD were older, heavier and had more severe OSA at the initial evaluation than those in
whom these diseases were not newly diagnosed. However, the diseases were developed despite tendency to reduce BMI in this
group as opposed to increased BMI in the other group, such that in the second evaluation the BMIs did not di¡er signi¢cantly between the groups.
*Po0?05 between groups.
saturation. Breakdown of the untreated group to three
subgroups according to RDI, BMI and age is presented in
Table 2. In none of the subgroups was the change in RDI
between the two PSG evaluations signi¢cant. In ¢ve patients the RDI in the second study decreased (improved),
while in other six it increased (worsened). Characteristics of these two subgroups are shown inTable 3.The ‘improved’ group consisted of older patients with worse
RDI than the ‘worsened’ group. Data of the patients
who developed hypertension/ischaemic heart disease
during the study period are presented in Table 4. These
patients were signi¢cantly older (57 vs. 45 yrs, Po0?05)
and had signi¢cantly higher RDI than the other untreated
patients (46727 vs. 23717 h1, Po0?05). Interestingly,
they were also heavier at the time of diagnosis, but they
tended to reduce weight compared to the others who
tended to gain weight, such that at the time of second
evaluation the BMIs of the two groups were similar.
There were no di¡erences in other risk factors of HTN/
IHD such as smoking habits (1/8 and 4/32, respectively),
hyperlipidaemia (1/8 and 2/32, respectively), diabetes
mellitus (1/8 and 2/32, respectively) and alcohol consumption (0/8 and 3/32, respectively) among the patients
who developed cardiovascular consequence and those
who did not.The subjective evaluation of change in sleepiness/snoring between the two groups was also similar.
NATURAL HISTORYOF SLEEPAPNOEA
DISCUSSION
The results of this study suggest that untreated patients
with obstructive sleep apnoea syndrome do not
necessarily worsen over time, as manifested by similar
respiratory disturbance indices. Although relatively
long-term follow-up (5 yrs on the average) PSG evaluation of untreated patients revealed in this study somewhat deeper oxygen desaturations and increased
intensity of snoring, the clinical relevance of this is unclear. The respiratory disturbance index remained unchanged over 5 yrs of follow-up. None of the tested
variables could predict which patients would worsen
and which would not.
Trying to conclude regarding the natural history of
OSA from a combination of our results with results of
previous studies, we believe that OSA is a very gradually
progressive disorder. We speculate that it starts with a
long period of indolent snoring (1), then progresses to a
relatively rapid period of increased frequency of respiratory events (3,12), again followed by relatively stable
status (5) with potential mild worsening of snoring and
oxygen desaturations, as we observed in the present
study. In the absence of weight gain, it is impossible at this
time to predict which patient will deteriorate and which
will not. Both in the present study, and in the previous
studies that reported deterioration of apnoea with
time, neither age nor BMI correlated with the change in
RDI (10).
The fact that in the two studies reporting deterioration, the RDI was mild-moderate, and in the study that
reported stable RDI, it was severe, raises the possibility
of a ceiling e¡ect. However, the ¢ndings of this study plus
those of Rosenthal et al., that even with low RDIs there
is not necessarily progression, somewhat oppose this
theory. Likewise, it could be argued that a threshold
e¡ect caused increase in RDI in the Svanborg study (12),
however, again, we and others (6,9) found that even with
low RDI there is not necessarily deterioration with time
(Table 2). Furthermore, the very mild change of
minimal oxygen saturation recorded in the second
evaluation of the untreated group is within the
con¢dence limits for this measure, and its clinical
relevance is unclear.
Night-to-night variability in RDI can undoubtedly play
a role in the ¢ndings of this as well as all the above mentioned papers. It has been previously shown that nightto-night variability in RDI can be as high as 730% (13,14).
Thus, our, and others’ (5), attempt to distinguish patients
who ‘improved’ from those who ‘worsened’ on an individual basis is potentially problematic. For a large group,
however, it can be assumed that this random variability
will be self-canceling. In addition, it is unlikely that ‘¢rstnight’ e¡ect plays a role given the long time duration
between the two studies. Therefore, the group ¢nding
of increased ODI from 10?1 to 20?9 over 15?6 months
341
(12), and of RDI from 21?8 to 33?4 over 17 months (3),
probably re£ects a real worsening rather than variability.
Likewise, the RDI stability we found as well as that reported by others (5) is also probably real and not a result
of nightly variability. One possible explanation for the
con£icting results is that the patients in the Svandborg
and Pendlebury studies (3,12) were still in the ‘rapid
deterioration phase’ of the disorders, whereas
those in Sforza’s study and in the current one were beyond that period.We do not know whether the patients
in the Sforza study tended to develop hypertension, as
this information was not given in that study. In the
present study, however, ¢ve out of 34 developed this
over a 5-yr period (15%), which is substantially higher
than the expected 3^ 4% incidence of newly diagnosed
hypertension in the general population over a
similar period (15). This brings us to the conclusion
that even in the absence of RDI deterioration, OSA may
continue to progress gradually, with worsening oxygen
desaturations and posing a risk for associated diseases
such as hypertension (16). However, the patients who developed new hypertension or ischaemic heart disease
were older and heavier than those who did not, which
could also contribute to the development of these
diseases.
The ¢nding that in the weight-loss group, signi¢cant
weight loss was associated with signi¢cant reduction in
RDI, is consistent with many previous works (17^22).
While these studies have followed up these patients for
0?5^2 yrs, in the present study the follow-up period is
2?572?3 yrs. As previous studies failed to demonstrate
correlation between the amount of weight loss and the
improvement in apnoea indices (17,18), it was unclear
whether a critical threshold level for weight loss exists,
and which patients were most likely to bene¢t from
treatment. The ¢nding in the present study, of a signi¢cant correlation between the weight loss and improvement in RDI (R¼0?75, P¼0?005), supports the concept
that any patient may bene¢t from weight loss, without
such a critical threshold.
Our study has several limitations:
(1) Although we chose an unselected group, namely
took all patients who were followed-up in our clinic,
the group may be biased by the fact that only patients
who came to the clinic were analyzed. We cannot
predict what would have been the status of
untreated patients who had not returned to clinic.
One could argue that the most symptomatic would
likely return for visits. Alternatively, the most
severely a¡ected patients may have been forced
into treatment or succumbed to their disease
(23,24). In order to minimize sampling bias from
this population, we chose to study all eligible
individuals rather than a speci¢c severity or subgroup.
342
(2) The number of subjects who participated in this
study is too small to draw conclusions regarding the
natural history of OSA. However, given that they
were followed up both subjectively and objectively
for a relatively long period of time, we believe these
¢ndings are important and add to the existing
literature. In addition, although only 40 patients
participated, these were all eligible individuals, with
no patients dropping out due to missing data or
technical problems.
(3) The follow-up period of the weight loss group is
shorter than that of the untreated group. As a
result, their OSA severity could have changed over
an additional 2?5 yrs of follow up. Previously, we
assessed OSA patients, both at short and long
periods after weight reduction surgery, and
found that they improved in the short-term
following surgery, but again worsened in
long-term follow-up (25). Anatomical abnormalities
of the upper airway may also be important in
predicting the e¡ect of weight loss (26). However,
again, we did not ¢nd gross ENT abnormalities in
our patients.
(4) Using a patients’/family physician’s report as the
source of diagnosis of hypertension or ischaemic
heart disease may result in underdiagnosis of the
problems, since only symptomatic patients were
probably diagnosed. Possibly this represents the
lower limit of cardiovascular complications in
patients with OSA (27). Including cardiologic
evaluation in the protocol may have improved the
accuracy of these cardiovascular complications
evaluations. However, we believe the new onset
diagnosis of these complications could equally
be made in all patients and therefore believe the
results have not been skewed in any particular
direction.
In conclusion, the most striking observations of this
study are the relatively stable RDIs in untreated OSA patients over a 5-yr period. RDI, age and BMI at the time of
the initial evaluation were not predictive of changes in
RDI, snoring intensity or minimal oxygen saturation, but
the patients who developed hypertension/IHD had significantly higher RDI than the patients who did not.Dietary
weight reduction is an e¡ective treatment in a ‘dose’related fashion.
REFERENCES
1. Lugaresi E, Cirignotta f, Gerardi R, Montagna P. Snoring and
sleep apnoea: natural history of heavy snorers disease. In:
Guilleminault C, Partinen M (eds.) Obstructive sleep apnoea syndrome: clinical research and treatment. New York: Raven Press,
1990:25^36.
RESPIRATORY MEDICINE
2. Martikainen K, Partinen M, Urponen H,Vuori I, Laippala P, Hasan J.
Natural evolution of snoring: a 5-year follow-up study. Acta Neurol
Scand 1994; 90: 437^ 442.
3. Pendlebury ST, Pepin JL,Veale D, Levy P. Natural evolution of moderate sleep apnoea syndrome: signi¢cant progression over a mean
of17 months.Thorax 1997; 52: 872^ 878.
4. Svanborg E, Larsson H. Natural evolution of obstructive sleep
apnoea syndrome. Sleep1993; 16: S124 ^S125.
5. Sforza E, Addati G, Cirignotta F, Lugaresi E. Natural evolution of
sleep apnoea syndrome: a ¢ve year longitudinal study [see comments]. Eur Respir J1994; 7: 1765^1770.
6. Rosenthal L, Rohers T, Roth T. Natural course of sleep apnoea:
a two year follow-up. In: Kuna S, Suratt P, Remmers J (eds.)
Sleep and respiration in aging adults. New York: Elsevier,
1991:348.
7. Phoha RL, Dickel MJ, Mosko SS. Preliminary longitudinal assessment
of sleep in the elderly. Sleep 1990; 13: 425^ 429.
8. Bliwise D,Carskadon M,Carey E, Dement W. Longitudinal development of sleep-related respiratory disturbance in adult humans.
J Gerontol 1984; 39: 290 ^293.
9. Ancoli-Israel S, Kripke DF, Klauber MR, et al. Natural history of
sleep disordered breathing in community dwelling elderly. Sleep
1993; 16: S25^S29.
10. Lindberg E, Elmasry A, Gislason T, et al. Evolution of sleep
apnoea syndrome in sleepy snorers: a population-based
prospective study. Am J Respir Crit Care Med 1999; 159:
2024 ^2027.
11. Rechtscha¡en A, Kales A. A manual of standardized terminology,
techniques and scoring system for sleep stages of human subjects. Los
Angeles: Brain Information Service/Brain Research Institute,
UCLA,1968.
12. Svanborg E, Larsson H. Development of nocturnal respiratory
disturbance in untreated patients with obstructive sleep apnoea
syndrome.Chest 1993; 104: 340 ^343.
13. Mosko SS, Dickel MJ, Ashurst J. Night-to-night variability in sleep
apnoea and sleep-related periodic leg movements in the elderly.
Sleep 1988; 11: 340 ^348.
14. MeyerTJ, Evelo¡ SE, Kline LR, Millman RP.One negative polysomnogram does not exclude obstructive sleep apnoea. Chest 1993; 103:
756 ^760.
15. Bakx JC, van den Hoogen HJ, van den Bosch WJ, et al. Development
of blood pressure and the incidence of hypertension in men and
women over an 18 -year period: results of the Nijmegen Cohort
Study.J Clin Epidemiol 1999; 52: 531^538.
16. Peppard P, Young, T, Palta, M, Skatrud, J. Prospective study of the
association between sleep disordered breathing and hypertension.
N Engl J Med 2000; 342: 1378 ^1384.
17. Smith PL, Gold AR, Meyers DA, Haponik EF, Bleecker ER. Weight
loss in mildly to moderately obese patients with obstructive sleep
apnoea. Ann Intern Med 1985; 103: 850 ^ 855.
18. Peiser J, Lavie P, Ovnat A, Charuzi I. Sleep apnoea syndrome in the
morbidly obese as an indication for weight reduction surgery. Ann
Surg1984; 199: 112^115.
19. Rubinstein I, Colapinto N, Rotstein LE, Brown IG, Ho¡stein V.
Improvement in upper airway function after weight loss in patients
with obstructive sleep apnoea. Am Rev Respir Dis 1988; 138:
1192^1195.
20. Rajala R, Partinen M, Sane T, Pelkonen R, Huikuri K, Seppalainen
AM. Obstructive sleep apnoea syndrome in morbidly obese
patients.J Intern Med 1991; 230: 125^129.
21. Suratt PM, McTier RF, Findley LJ, Pohl SL,Wilhoit SC. E¡ect of verylow-calorie diets with weight loss on obstructive sleep apnoea. Am
J Clin Nutr1992; 56: 182S ^184S.
22. Schwartz AR, Gold AR, Schubert N, et al. E¡ect of weight loss on
upper airway collapsibility in obstructive sleep apnoea. Am Rev
Respir Dis1991; 144: 494 ^ 498.
NATURAL HISTORYOF SLEEPAPNOEA
23. He J, Kryger MH, Zorick FJ, Conway W, Roth T. Mortality and
apnoea index in obstructive sleep apnoea. Experience in 385 male
patients.Chest 1988; 94: 9^14.
24. Lavie P, Herer P, Peled R, et al. Mortality in sleep apnoea patients: a
multivariate analysis of risk factors [see comments]. Sleep1995; 18:
149^157.
25. Pillar G, Peled R, Lavie P. Recurrence of sleep apnoea without concomitant weight increase 7?5 years after weight reduction surgery.
Chest 1994; 106: 1702^1704.
343
26. Pasquali R, Colella P, Cirignotta F, et al.Treatment of obese patients
with obstructive sleep apnoea syndrome (OSAS): E¡ect of weight
loss and interference of otolaryngoiatric pathology. Int J Obes 1990;
14: 207^217.
27. Trilling JS, Froom J. The urgent need to improve hypertension care.
Arch Fam Med 2000; 9: 794 ^ 801.