INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In spite of the extensive literature that currently exists on obstructive sleep apnea syndrome (OSAS), most studies have been cross-sectional in design and our knowledge of the natural evolution of this disease is still limited. In a retrospective study based on the history of 118 OSAS patients, Lugaresi and coworkers found that a typical OSAS patient was a man who had snored every night for years or decades before the start of OSAS-associated illness. On the basis of these results, the investigators suggested that OSAS is a progressive disease which starts with simple snoring and develops into OSAS of increasing severity (1). In the few studies performed to evaluate the outcome of untreated OSAS patients, the theory that OSAS is a progressive disease has been supported by two (2, 3), whereas the result of one indicated that OSAS is stable over time (4).

The aim of this population-based study was to investigate the long-term outcome in men with snoring and excessive daytime sleepiness (EDS) over a 10-yr period. The subjects were included, regardless of whether the polysomnography at baseline showed apneas.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Population

The study design is presented in Figure 1. In short, 3,201 from a total of 4,021 males, randomly selected from the general population (response rate 79.6%), answered a postal questionnaire in 1984, including questions about somatic diseases, snoring, sleep disturbances, and associated daytime symptoms (5). On the basis of the answers, primarily those to questions relating to snoring and EDS, a group of 166 men who were strongly suspected of having OSAS were identified. A sample of 61 of these men was investigated using all-night polysomnography in 1985 (6). They also underwent a physical examination and their height and weight were measured.

View larger version (25K): [in this window] [in a new window]   Figure 1.   Study design.

All the subjects were informed of the result of the polysomnography in 1985. When indicated, they were advised to lose weight, stop smoking, and reduce their alcohol intake. They were also offered treatment with continuous positive airway pressure (CPAP) or a referral to an otorhinolaryngologist for surgical treatment.

Procedure at the Follow-up

In 1994, all the survivors from the 3,201 subjects who answered the first postal questionnaire were followed using a new questionnaire with questions identical to the ones in 1984. The follow-up questionnaire also included questions about smoking, alcohol consumption, and physical activity (7).

In 1995, seven of the 61 men studied with full polysomnography had died. The remaining 54 were invited to take part in this follow-up-study. Before the visit to the sleep laboratory, a structured interview was performed by telephone. The subjects were asked about their general health, medication, and whether they had ever attended medical check-ups or had been admitted to hospital because of hypertension, cardiac disease, or stroke. All subjects were asked to state whether their symptoms of snoring and EDS had decreased, increased, or been stable during the 10-yr period. They were also asked whether they had ever sought medical advice because of any of these symptoms and, if so, what kind of investigation and/or treatment they had been offered. Medical records were obtained from the relevant departments.

When the subjects arrived at the sleep center, a physical examination of their lungs, heart, blood pressure, and pulse was performed. Height and weight were measured and the body mass index (BMI) was calculated. The data on alcohol and physical activity were based on answers to the questionnaire in 1994; the data on smoking and hypertension were based on the interviews preceding the sleep recordings. Subjects were classified as hypertensive if they carried a previous diagnosis of hypertension or if they had a measured blood pressure on examination of greater than 180 systolic and/or 100 diastolic.

All the studied subjects were unpaid volunteers. The informed consent of all the participants was obtained, and the study was approved by the ethics committee at the medical faculty at Uppsala University.

Polysomnography

Polysomnographic recordings were performed in a manner similar to that used at baseline (6) and identical criteria were used in the scoring in 1985 and in 1995: An apnea was defined as a complete cessation of oronasal air flow for at least 10 s and hypopnea as a marked decrease in air flow for at least 10 s, followed by a reduction in oxygen saturation of at least 4% from the baseline level and/or an arousal. The apnea-hypopnea index (AHI) was calculated as the total number of such events divided by hours of sleep. The AHI was calculated as AHI 1995  AHI 1985.

Statistical Methods

The computations were performed using the Statistica 4.0 software package (StatSoft Inc., Tulsa, OK). The results are presented as mean ± SD. Significance test for changes between baseline and the follow-up were conducted using the paired t test for continuous variables and the chi-square test for proportions. Comparisons between subgroups were performed using the unpaired t test for continuous variables, the Mann-Whitney U test for variables on an ordinal scale, and the chi-square test for proportions. A simple linear regression was used to calculate correlations between continuous variables. Correlations between continuous variables and variables on an ordinal scale were performed with the Spearman rank correlation test. To achieve normal distributions for continuous variables, the variables were log transformed. The null hypothesis was rejected at a level of p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Response Rate, Nonparticipants

Of the 54 subjects who were still alive in 1995, 38 (70.4%) participated in this follow-up. A total of 13 subjects replied saying that they did not wish to participate. Three men, one of whom lived abroad, did not answer the invitation or reminders. When the seven who had died are also included, a total of 23 subjects were only analyzed in 1985. All the participants who took part in the follow-up study also answered the postal questionnaire in 1994.

No significant difference was found between the 38 participants and the 23 who did not participate in the follow-up in terms of age, BMI, or AHI at baseline or in terms of any of the sleep parameters. The nonparticipants were more often smokers and had a higher prevalence of hypertension at baseline (Table 1).

Treated Participants

During the 10-yr period a total of nine of the men had been treated for OSAS or snoring by surgery and/or by CPAP. Compared with the untreated subjects, the treated group as a whole had a significantly higher mean AHI at baseline (11.8 ± 7.7 versus 2.1 ± 4.2, p < 0.001), whereas the groups did not differ significantly in terms of age or BMI at baseline. All nine men who had been treated were excluded from the following analyses.

Sleep Results

The sleep results in 1985 have been reported previously (6). In 1995, the mean total sleep time (TST) for the 29 untreated participants was 337 ± 68 min and the mean sleep efficiency (TST/time in bed) was 76 ± 14%. All the men slept for longer than 240 min, apart from two men who slept for 222 and 118 min respectively. The man who slept 118 min (11% rapid eye movement [REM]) at the follow-up had a sleep efficiency of only 29%. As he experienced major difficulty sleeping away from home, he did not want to come for another investigation. This man had a total of four hypopneas but no apneas during the study night.

For all the 29 men at the follow-up, sleep stages 1 and 2 constituted 63 ± 11% of the TST, stages 3 and 4 19 ± 8%, and REM sleep 19 ± 6%. All the subjects had REM sleep for at least 7% of their TST.

Respiratory Events

Among the 29 men, the number of respiratory events (apneas and hypopneas) during the night ranged from 0 to 149 (mean 38 ± 40) and the AHI ranged from 0 to 26 (mean 6.8 ± 7.2). In 16 of the men, the AHI was less than 5; five had an AHI of 5 to 10 and eight had an AHI of > 10. Subjects with an AHI of > 10 spent significantly less of their TST in sleep stages 3 and 4 compared with the others (13 ± 7 versus 21 ± 8%, p < 0.05).

Natural Evolution of Sleep-disordered Breathing

The characteristics of the untreated men at baseline and at the follow-up are presented in Table 2. In 1985, four of the 29 men had OSAS defined as an AHI of  5/h and 10 yr later all of them still fulfilled this criterion. In addition, nine of the 25 (36%) who had an AHI of < 5 at baseline had developed OSAS at the follow-up, two of whom now had an AHI of > 25 (Figure 2). During the 10-yr period, the AHI for the group as a whole increased from 2.1 ± 4.2 to 6.8 ± 7.2 (p > 0.01). When a 50% change in AHI was used to divide patients into worse, stable, and improved, 21 (72%) were worse, only two (7%) had improved, and six (21%) remained stable. No significant associations were found between AHI (i.e., AHI 1995  AHI 1985) and age or BMI at baseline, weight gain, smoking, alcohol dependence, or physical activity in 1994. 

View larger version (27K): [in this window] [in a new window]   Figure 2.   Changes in AHI between baseline and at the follow-up among the 29 untreated men.

Predictors of an Increase in AHI

To further analyze the predictors of a deterioration in sleep-disordered breathing, the subjects with a AHI of  5 were compared with the remainder. As shown in Table 3, subjects who increased their AHI by at least five units during the 10-yr period did not differ significantly from the remainder in terms of any of the variables investigated.

In the interview preceding the polysomnography in 1995, four of the 29 subjects reported that they suffered from an increase in snoring at the follow-up, while seven reported less snoring and for 18 this symptom had been stable over the 10-yr period. The four men with increase in snoring had a somewhat higher mean AHI compared with the remainder (8.8 ± 9.9 versus 4.1 ± 7.3, not significant [NS]). An increase in EDS during the follow-up period was reported by nine, only one of whom reported increase in both snoring and EDS. All the subjects who had an increase in EDS also had an increase in AHI and their mean AHI was 10.3 ± 9.7, which was significantly higher than the AHI of 2.2 ± 5.1 found among the remainder (p = 0.01). No significant differences were found between the men with increasing EDS and the remainder in terms of age, BMI, smoking, weight gain, alcohol dependence, or level of physical activity.

Causes of Death

Seven men died during the 10-yr period. Their mean age at death was 61 ± 13 yr (range 34 to 72). Ishaemic heart disease was the cause of death in three of them, while malignancy was the underlying cause of death in three. For one man who died in Spain, it was not possible to obtain a death certificate. Compared with the remainder, those who died were older (57 ± 12 versus 50 ± 10 yr in 1985, p = 0.08) and had a somewhat higher mean AHI in 1985 (7.5 ± 11 versus 3.7 ± 6.1, p = 0.2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our main finding is that among this small group of men who were selected because of symptoms related to sleep apnea, sleep-disordered breathing became significantly worse over the 10-yr period.

The follow-up period is longer than previously described in the literature. Furthermore, all the subjects investigated here were enrolled from a general male population because of a reported symptom profile in a questionnaire. This is in contrast to previous follow-up studies in this field in which only patients who themselves have sought medical advice and have had a diagnosis of OSAS have been included (2). Because of the population-based design, our results should be representative of changes seen in middle-aged men who are symptomatic with sleepiness and snoring. However, it seems reasonable to assume that there could be a selection bias at follow-up, as individuals with increasing symptoms are more eager to be investigated again.

A progression over time has also been reported by the two investigations performed to study the evolution of mild to moderate OSAS (2, 3). In contrast, in the study by Sforza and coworkers, no significant changes in apnea frequency were found in 32 OSAS patients at a follow-up at least 5 yr later (4). However, the subjects enrolled in their study had more severe OSAS (mean AHI 52.2). In their subgroup of seven patients who had deteriorated, the AHI at baseline was significantly lower than in the group that had improved (mean 14.1 versus 56.3, p < 0.001), thereby indicating that their result did not differ significantly from those of other studies with respect to the evolution from mild to moderate OSAS.

In this study, no predictor of a deterioration in sleep-disordered breathing could be identified. This could possibly be explained by the relatively low number of examined subjects and thereby the low statistical power of our study. Furthermore, as no asymptomatic control group was investigated here, it is not possible to establish the role of age with regard to changes in the number of respiratory events over time. The results of longitudinal studies in populations enrolled without reference to sleep-wake complaints have conflicted, as a slight decrease in AHI over time has been found by some (8), whereas others have reported a slight increase (9, 10). However, in studies in which patients who already have OSAS have been followed, no association has been found between higher age and deterioration (2, 3). In the study by Sforza and coworkers, the group which deteriorated actually tended to be younger than those who improved or remained stable (4). Moreover, with respect to BMI, the lack of correlation between overweight, weight gain, and deterioration in sleep-disordered breathing has been observed in two of the previous follow-up studies (3, 4). Svanborg and Larsson found an association between increase in weight and an increase in oxygen desaturation index but also noted considerable increases in respiratory disturbance in some patients despite weight loss (2). A progressive neurogenic lesion in the palatopharyngeal muscle in patients with different degrees of upper airway obstruction, including nonapneic heavy snorers, was recently reported by Friberg and coworkers (11). As suggested by these investigators, this progression, which is due to local mechanical trauma as a result of snoring, is a possible contributory factor to upper airway collapsibility. This would be a plausible explanation for the progressive nature of snorers disease seen in some subjects, regardless of other predictors of deterioration.

Reporting an increase in daytime sleepiness at the follow-up was highly associated with the progress of sleep-disordered breathing. However, the definition of sleepiness becoming worse over time was based on retrospective recall of symptoms made at the interview in 1995 which prevents us from drawing any definite conclusions. Another disadvantage is that alcohol consumption and the level of physical activity were only included in the questionnaire in 1994. The possibility cannot be ruled out that changes over time in these parameters might influence the changes in AHI to some extent.

We conclude that in this small sample of middle-aged males with snoring and daytime sleepiness, sleep-disordered breathing became significantly worse over a 10-yr period. It appears that deterioration was independent of age, BMI, weight gain, and smoking.