International Journal of Epidemiology © International Epidemiological Association 1997

Vol. 26, No. 2 Printed in Great Britain

Prevalence of Sleep Apnoea Syndrome in the Spanish Adult Population J M MARIN,* J M GASCON,* S CARRIZO* AND J GISPERT† Marin J M (Servicio de Neumología, Hospital Miguel Servet, Avda Isabel la Católica 1–3, 50006-Zaragoza, Spain), Gascon J M, Carrizo S and Gispert J. Prevalence of sleep apnoea syndrome in the Spanish adult population. International Journal of Epidemiology 1997; 26: 381–386. Background. Some data indicate that obstructive sleep apnoea syndrome (OSAS), a disorder characterized by recurrent episodes of cessation of respiratory airflow during sleep, is highly prevalent in the general population but no such data exist in southern Europe. Methods. In the Zaragoza metropolitan area (northeast of Spain) a representative sample of 1360 subjects aged .18 years and selected by quota methods according to age, sex and geographical distribution agreed to participate. Trained interviewers visited selected residents to administer a sleep questionnaire in the presence of a bedmate or another closely-related person who lived in the subject’s home; anthropometric data and arterial blood pressure were also recorded. All participants were invited to record nocturnal home oximetry (NHO). The NHO results were classified as ‘abnormal’ (or consistent with OSAS) in the presence of repetitive, short-duration arterial oxyhaemoglobin saturation (SaO2) fluctuations. Results. The diagnosis of OSAS was established in subjects with loud (severe) snoring + excessive daytime sleepiness + abnormal oximetry. In the group of 1222 subjects (597 males, 625 females) who agreed to have NHO, 63.7% of men and 36.3% of women snored ‘usually’ or ‘always’ (severe snorers); daytime sleepiness in active situations occurred in 12.1% and 14.4% respectively. The association of severe snoring plus daytime sleepiness plus abnormal NHO was found in 18 subjects (13 males, 5 females). Conclusions. We estimated that among Spanish adults, 0.8% of women and 2.2% of men meet the minimal criteria to diagnose sleep apnoea syndrome. Keywords: apnoea, sleep, snoring, prevalence, obstructive sleep apnoea, sleep-disordered breathing, nocturnal home oximetry

prevalent in the UK8, Northern Europe,6,9,10 Australia13 and US.11,12 Due to the absence of data about the prevalence of sleep apnoea and sleep-related medical problems in Southern Europe, we decided to investigate that topic in a general population of adults studied at home.

Obstructive sleep apnoea syndrome (OSAS) is characterized by repetitive episodes of upper airway obstruction that occur during sleep, usually associated with a reduction in blood oxygen saturation.1,2 Almost all such patients have a long history of loud snoring. The chief clinical consequence of OSAS is excessive daytime sleepiness; most patients also report cognitive dysfunction and memory loss. Sleep apnoea is known to contribute to the development of both systemic and pulmonary hypertension and to biventricular failure.3,4 Previous epidemiological investigation of sleep apnoea has been hampered by difficulties in obtaining an adequate population-based sample, the lack of spouse interviews or they have been limited to specific groups of subjects: middle-aged men or industrial workers.5–13 In spite of the limitations, these studies concluded that OSAS is

METHODS Sample The study was performed in 1993–1994 in the municipality of Zaragoza, in the northeast of Spain. At that time there were 580 000 inhabitants in the population register; the population was representative for the Spanish population except for the absence of fishermen and miners. In all 360 000 people were .18 years and from that population 1640 people were randomly selected with respect to sex, age and neighbourhood distribution. Of these, 50.2% were females and 49.8% males. A written invitation to participate followed by telephone

* Servicio de Neumología, Hospital Miguel Servet, Avda Isabel la Católica 1–3, 50006-Zaragoza, Spain. † Departamento de Bioestadistica, Universidad de Zaragoza, Spain.

381

382

INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

contact was accepted by 1360 and they were then visited at home by two trained nurses. Subjects who lived alone (n = 101) were substituted by others who did not live alone, choosen at random, but with the same characteristics (age, sex and area of residence). Non-responders (n = 320) had a slightly higher age but showed no difference in sex distribution, education level or marital status. Study in the Home Subjects were questioned always in the presence of the most closely-related person, preferably the husband/ wife. First of all, subjects were asked to complete an 85-item questionnaire divided in two parts. Part 1 included questions regarding lifestyle factors, occupation, general heath status and medications. Part 2 included items concerning sleep hygiene, sleep disorders and diurnal hypersomnia. A typical question was: do you feel paralyzed (unable to move) when falling asleep? Replies were coded as: 1: never, 2: rarely, 3: sometimes, 4: usually and 5: always. Specific questions (n = 6) about snoring and sleep behaviour were asked of subjects and their closest family member. Unless otherwise specified, responses of ‘usually’ and ‘always’ constitute a positive response and ‘never’, ‘rarely’ or ‘sometimes’ constitute a negative response. Loud or severe snoring were identified when the bedmate (or housemate) reported this symptom as having occurred during sleep every night or most nights in a week. Physical measurements made in the home included height and weight to determine body mass index (BMI = kg/m2), neck circumference and blood pressure according to standard guidelines.14 Nocturnal Home Oximetry All participants (n = 1320) were offered a single nocturnal examination in their homes (nocturnal home oximetry [NHO]); the test was performed in 1222 subjects (597 males, 625 females). No differences were found in age, sex, anthropometric data, general health status and sleep questionnaire responses between those who accepted or not the overnight oximetry. The NHO was done with a BIOX 3700 oximeter (Ohmeda, Louisville, CO, USA); arterial oxygen saturation (SaO2); signal was obtained through a finger probe. A technician from the sleep laboratory visited the subjects to fit the device at their usual bedtime. Subjects were instructed to interrupt the recording if they awoke during the night and to stop it when they awoke in the morning. The technician removed the device in the morning at the subject’s home and transferred data to the computer in the sleep laboratory. The NHO was done twice in 24 subjects who reported not sleeping well during the first recording.

The oximetry results were manually inspected and counted from traces printed from the computer storage. The test was considered as abnormal (consistent with OSAS) if there were repetitive episodes (mean for the whole night .10/hour) of transient SaO2 desaturation followed by a rapid return to the baseline SaO2 level;15 the total number of desaturation episodes during the recording time was divided by the estimated sleep time to obtain the desaturation index (DI). The minimum SaO2 value and the percentage of time recording with a SaO2 below 90% were also registered. Statistical Analysis All data were analysed by the SPSS statistical package program.16 The data are represented as percentage and mean (standard deviation). The 95% confidence intervals (CI) of the prevalence figures were estimated on the assumption of a normal distribution.8 Relationships among the data were examined by analysis of variance or χ2 test as appropriate. To assess the influence of different variables, stepwise multiple regression analysis was used. The protocol was approved by the Miguel Servet Hospital Clinical Research Committee.

RESULTS Sample Characteristics Of the 1640 adults selected, 160 could not be contacted after many attempts either by telephone or home visits and 120 refused to participate due to multiple reasons; therefore 1360 subjects were studied, of whom 1222 agreed to nocturnal home oximetry. Subjects who accepted or refused (n = 138) NHO were compared with regard to their responses to all questionnaire items on sleep complaints, additional medical problems, sex and age; no differences were found. Table 1 shows selected characteristics of the population. The sample included most occupations and ranged from unskilled manual workers to professionals; a total of 46% were employed full-time and 26% were retired. The prevalence rates for chronic diseases (hypertension, ischaemic heart disease, chronic bronchitis...) obtained from the questionnaires are similar to those reported annually by the Ministry of Health for the Spanish general population. There was no evidence of selection bias that would limit the representativeness of the sample. Prevalence of Snoring The prevalence of snoring ‘usually’ or ‘always’ (severe snoring) as reported by the spouse was much higher among males than females (Table 2) and increased with age with a peak at 50–59 years (Figure 1). There was a significant relation of severe snoring to smoking in

383

SLEEP APNOEA IN SPAIN

TABLE 1 Characteristics of participants according to gender Males n = 625

Females n = 597

41.5 ± 9.1 (18–86)

43.6 ± 8.3 (18–88)

Body mass index, kg/m2 a (range)

24.8 ± 2.8 (16.3–46.8)

23.4 ± 2.7 (15.1–44.4)

Neck circumference, cma (range)

39.9 ± 2.7 (33–51)

34.1 ± 2.3 (28–48)

Current smokers, no. (%) (.10 cig/day)

229 (38.3)

190 (30.4)

28 (4.2)

10 (1.6)

Age, yearsa (range)

Alcohol consumption, no. (%) (.20 g/day) Nightly sleeping pills, no. (%) Hypertension, no. (%) (>140/90 mmHg) a

33 (5.5)

56 (8.9)

154 (25.8)

153 (24.5)

FIGURE 1 Combination of major symptoms of obstructive sleep apnoea with abnormal nocturnal home oximetry result in male and female adults. Figures represent percentage of subjects

TABLE 3 Several characteristics of subjects in relation to nocturnal home oximetry (NHO) results mean ± SD.

Normal NHO n = 1088 (89%)

Abnormal NHO n = 134 (11%)

507/581 39.8 ± 11.8 487 (44.8) 9 (0.8) 134 (12.3) 260 (23.9) 371 (34.1)

90/44** 46.2 ± 10.1 120 (89.6)* 30 (22.4)** 36 (21.1)* 47 (35.1)* 48 (35.8)

TABLE 2 Prevalence of symptoms of sleep-disordered breathing by gender

Severe snoring Mild/moderate snoring Stops breathing Snoring disturbs others Struggle for breath Fall asleep: reading at work as driver Severe snoring + hypersomnolence a

Males n = 597

Females n = 625

63.6 (59.8–76.5)a 32.3 (28.6–36.1) 4.4 (2.9–6.3) 15.4 (13.1–19.0) 4.4 (2.9–6.3)

36.3 (32.5–40.1) 34.1 (30.4–37.8) 2.1 (1.1–3.53) 8 (6–10.4) 3.2 (1.9–4.9)

12.1 (9.4–14.7) 3.2 (1.9–4.9) 1.8 (0.9–3.3)

14.4 (11.6–17.2) 1.9 (0.9–3.3) 0.8 (0.2–1.8)

14 (11.3–16.9)

10.4 (8.01–12.8)

Percentage (95% confidence interval).

both males and females. In comparison with non-severe snorers, those who reported severe snoring were twice as likely to be current smokers and had higher daily alcohol consumption (9 ± 15 versus 13 ± 12 g/day, P , 0.01). Logistic regression analysis was performed to examine the effects of continuous variables on snoring. Male sex (P , 0.0001), BMI (P , 0.01), neck circumference (P = 0.007), smoking (P = 0.03) and daily alcohol consumption .10 g were related to severe snoring. In this model, in our population, a smoker who drinks .10 g of alcohol daily, with a BMI .30 and a

Male/female Age, years, mean ± SD Severe snoring, no. (%) Stops breathing, no. (%) Hypersomnolence, no. (%) Hypertension, no. (%) Current smokers, no. (%) * P , 0.05; ** P , 0.01.

neck size .40 cm, is five times more likely to be a severe snorer than is a non-smoking, non-drinker with a BMI ,30 and a neck circumference ,30 cm. Nocturnal Home Oximetry Almost all subjects reported normal quality of sleep in the overnight oximetry. Many subjects had some episodes of nocturnal desaturations; in fact, 60% of severe snorers and 28% of those who reported sometimes or rarely snoring had more than one episode of desaturation per hour of recording. Overall, 11% of NHO were considered abnormal, that is, there was more than 10 desaturation events/h of recording (Table 3). Abnormal nocturnal oximetry was more common among males than females (P , 0.01). In the group with abnormal NHO there were also more subjects with nocturnal respiratory complaints, diurnal hypersomnolence and arterial hypertension; there was no difference in the percentage of subjects with concomitant respiratory,

384

INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

FIGURE 2 Proportion of subjects related to age groups who reported severe snoring, daytime hypersomnolence and struggle for breath during sleep

cardiac (except for hypertension) or other systemic diseases. Logistics regression analysis demonstrated that age (P , 0.005), gender (P , 0.001), severe snoring (P , 0.0001) and cessation of breathing (P , 0.05) were associated with abnormal NHO. Prevalence of Sleep Apnoea Syndrome Figure 2 shows the prevalence of the combinations of the three main essential features in OSAS in males and females. Diurnal hypersomnolence plus severe snoring was higher among men (14%) than women (10.4%) (P , 0.01) and increased with age in both groups (Figure 1). There was also a higher prevalence of hypersomnolence plus abnormal NHO in men (3%) than women (1.7%) (P , 0.05) with a peak at age 60–69 years. The association of severe snoring, daytime hypersomnolence and abnormal nocturnal home oximetry was detected in 13 males and five females. This group had a higher BMI (P , 0.05) and a larger neck size (P , 0.01) than a selected subgroup of 525 subjects who did not complain of snoring, did not have daytime sleepiness and whose NHO was normal. Alcohol consumption and current smoking were also more common in the former group. From this data, we estimated that a minimum of 2.2% of men and 0.8% of women of the Spanish general adult population had sleep-disordered breathing consistent with severe obstructed sleep apnoea.

DISCUSSION The major finding from our data is that undiagnosed obstructive sleep apnoea syndrome is quite prevalent among Spanish adults. A minimum of 2.2% of men and 0.8% of women meet minimal diagnostic criteria for OSAS: severe snoring + daytime hypersomnolence + abnormal nocturnal home oximetry. We used the

finding of abnormal NHO as an indicator of sleepdisordered breathing because this technique has shown that nocturnal SaO2 recording accurately selects most patients clinically suspected of having OSAS at clinical rounds.8,15 Home oximetry testing has a sensitivity of 98% for case-finding in patients with suspected sleep apnoea based on clinical symptoms; in addition, a negative NHO test result reduces the probability of having OSAS to 3.1% (negative predictive value of 96.9%).15 When using a clinical model of suspicion of sleep apnoea which included the presence of loud snoring and hypersomnolence, Rauscher et al.17 found that NHO had a sensitivity of 100% to detect a subject with .10 apnoeas per h of sleep. The NHO test has the advantage of great simplicity and minimum discomfort for the subject. The accuracy of the method and the nearly complete uptake (93%) of testing by the sample population make our results very representative of the whole adult population. Subjects who refused to participate were not different from those who did so in terms of degree of obesity or neck size, two of the major predictive variables for OSAS that could have biased our data. In epidemiological studies, figures for prevalence of sleep apnoea syndrome depend on the cutoff used for the so-called respiratory distrurbance index (RDI, number of apnoeas or hypopnoeas per h of sleep) or what criteria are used to define symptoms (described by subjects with clinically undiagnosed moderate/severe OSAS) for which most sleep specialists would recommend an immediate therapeutic intervention.18 The prevalence estimates of sleep apnoea syndrome in the study population are similar to those reported by other community studies which used portable monitors for testing subjects at home. For example, Stradling et al.8 studied 1001 middle-aged men using nocturnal oximetry. They found 5% of subjects with abnormal NHO and 0.3% with clinically severe symptomatic OSAS. Bearpark et al.13 reported subjective daytime sleepiness plus RDI >5 in a minimum of 3% of 294 men aged 40–65 years from the volunteer register of a coastal town in Australia. Finally, Redline12 et al. in an ongoing genetic-epidemiological study in the US in 389 subjects >16 years of age, found 4.6% of cases with daytime sleepiness plus RDI >15. In contrast with those studies, Young et al.11 using nocturnal polysomnography, the gold-standard method to diagnosis of OSAS, observed a prevalence (RDI >15 events per h) of 4% of women and 9.1% of men among middle-aged working adults. The differences in prevalence reported by different groups may be explained both by sampling biases and different diagnostic approaches. Our figures are conservative because an OSAS case had to complain of daytime sleepiness and severe snoring in addition to an

SLEEP APNOEA IN SPAIN

abnormal NHO. This is in contrast with most of the above reports in which RDI is the sole criteria for estimating the OSAS prevalence. In fact if we only consider NHO results, 15% of men and 6% of women had a desaturation index .10, a figure which agrees with most studies on sleep-disordered breathing prevalence. In our population, 64% of men and 36% of women snored loudly, as reported by a spouse or closelyrelated person. Using subjective reports, the prevalence of habitual snoring may previously have been underestimated, as it was said to range from 30 to 45% in men. When comparing self and spouse reports of snoring poor agreement has been found19 and so in epidemiological studies on sleep disturbances the bedpartner questionnaire or a direct measurement during sleep is mandatory. Using these approaches habitual snoring was detected in 64–81% of subjects in community studies.11–13 Detecting severe snoring may have implications for public health because ischaemic heart disease has been linked to snoring.20 As previously described,8 we also found a direct relationship between snoring, neck circumference and obesity. In patients with OSAS, obesity is responsible for an increase in neck size; this so-called ‘neck obesity’ has been implicated in the pathophysiology of upper airway obstruction during sleep due to fat compression of the pharynx.21,22 Other major independent risk factors for snoring found in this study were male gender and alcohol and cigarette consumption. Smoking may produce inflammation and oedema of the pharynx and so quitting may reduce snoring. Alcohol tends to induce obstructive sleep apnoea in snorers probably due to its depressant effects on the central nervous system.23 The ratio of males: females with OSAS in our population (3.6:1) was in agreement with that reported by Young et al.11 (2.25:1) and most recently that found by Redline et al. They specifically addressed this issue and found a male: female ratio for nocturnal apnoeas of 2:1.12 Those studies were also community-based and not clinicallybased24 where under-recognition of sleep symptoms and apnoeic events in women has been demonstrated.25 However, the importance of male sex was the most powerful predictor of snoring and obstructive sleep apnoea. In the elderly we found a prevalence of snoring, hypersomnolence and OSAS similar to other authors.26 Among seniors, the subgroup with abnormal NHO did not have a higher percentage of concomitant cardiovascular or neuromuscular diseases than the subgroup with normal NHO, therefore, central apnoeas or Cheyne Stokes breathing were unlikely causes of nocturnal desaturations on oximetry in senior patients with OSAS. We concluded that snoring and sleep-related symptoms are common among Spanish adults. The prevalence of

385

undiagnosed moderate-severe OSA in our population is prevalent both in women and men. Subjects with sleep apnoea in our community were distinguished from nonOSA by being older, smokers, more likely to be male with obesity and larger necks and have higher alcohol consumption.

ACKNOWLEDGEMENT Supported by Grant FIS 92/0447 from the Fondo de Investigaciones Sanitarias, Spain. REFERENCES 1

Guilleminault C. Clinical features and evaluation of obstructive sleep apnoea. In: Kryger M H, Roth T, Demment W C (eds). Principles and Practice of Sleep Medicine, 2nd Edn. Philadelphia: W B Saunders, 1994, pp. 667–77. 2 International classification of sleep disorders. In: Thorpy M J (ed.). Diagnostic and Coding Manual. Lawrence, KA: Allen Press, 1990. 3 Hung J, Whitford E G, Parsons R W, Hillman D R. Association of sleep apnoea with myocardial infarction in men. Lancet 1990; 336: 261–64. 4 He J, Kryeger M H, Zorick M J, Conway W, Roth T. Mortality and apnea index in obstructive sleep apnea: experience in 385 male patients. Chest 1988; 94: 9–14. 5 Lavie P. Incidence of sleep apnea in a presumably healthy working population. Sleep 1983; 6: 312–18. 6 Gislason T, Almqvist M, Eriksson G, Taube A, Boman G. Prevalence of sleep apnea syndrome among Swedish men– an epidemiological study. J Clin Epidemiol 1988; 41: 571–76. 7 Cirignotta F, D’Alessandro R, Partinen M et al. Prevalence of every night snoring and obstructive sleep apnoeas among 30–69 year-old men in Bologna, Italy. Acta Psychiatr Scand 1989; 79: 366–72. 8 Stradling J R, Crosby J H. Predictors and prevalence of obstructive sleep apnoea and snoring in 1001 middle aged men. Thorax 1991; 46: 85–90. 9 Haraldsson P O, Carenfelt C, Tingvall C J. Sleep apnea syndrome symptoms and automobile driving in a general population. Clin Epidemiol 1992; 45: 821–25. 10 Jennum P, Sjøl A. Epidemiology of snoring and obstructive sleep apnoea in a Danish population, age 30–60. J Sleep Res 1992; 1: 240–44. 11 Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing among middleaged adults. N Engl J Med 1993; 328: 1230–35. 12 Redline S, Kump K, Tshler P V, Browner I, Ferrette V. Gender differences in sleep disordered breathing in a communitybased sample. Am J Respir Crit Care Med 1994; 149: 722–26. 13 Bearpark H, Elliot L, Grinstein R et al. Snoring and sleep apnea. A population study in Australian men. Am J Respir Crit Care Med 1995; 151: 1459–65. 14 Joint National Committee on Detection, Evaluation and Treatment of High Blood Pressure. The 1988 Report of the Joint National Committee on Detection, Evaluation and Treatment of High Blood Pressure. Arch Intern Med 1988; 148: 1023–38.

386 15

INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

Sériès F, Marc I, Cormier Y, La Forge J. Utility of nocturnal home oximetry for case finding in patients with suspected sleep apnea hypopnea syndrome. Ann Intern Med 1993; 119: 449–53. 16 SPSS Reference Guide. Chicago: SPSS Ins, 1993. 17 Rauscher H, Popp W, Zwick H. Model for investigating snorers with suspected sleep apnea. Thorax 1993; 48: 275–79. 18 Sanders M H. Medical therapy for sleep apnea. In: Kryger M H, Roth T, Demment W C, (eds). Principles and Practice of Sleep Medicine, 2nd Edn. Philadelphia: W B Saunders, 1994, pp. 678–93. 19 Wiggins C L, Schmidt-Nowara W W, Coultas D B, Samet J M. Comparison of self- and spouse reports of snoring and other symtoms associated with sleep apnea syndrome. Sleep 1990; 13: 245–52. 20 Koskenvuo M, Kaprio J, Telakivi T, Partinen M, Heikkila K, Sarna S. Snoring as a risk factor for ischaemic heart disease and stroke in men. Br Med J 1987; 294: 630–32. 21 Davies R J O, Stradling J R. The relationship between neck circumference, radiographic pharyngeal anatomy, and the

constructive sleep apniea syndrome. Eur Respir J 1990; 3: 509–14. Koenig J E, Thach B T. Effects of mass loading on the upper airway. J Appl Physiol 1988; 64: 2294–99. 23 Issa F G, Sullivan C E. Alcohol, snoring and sleep apnea. J Neurol Neurosurg Psychiatry 1982; 45: 353–59. 24 Guillemineault C, Quera-Salva M A, Partinen M, Jamieson A. Women and the obstructive sleep apnea syndrome. Chest 1988; 93: 104–09. 25 Bliwisw D L, Nekich M A, Dement W C. Relative validity of self-reported snoring as a symptom of sleep apnea in a sleep clinic population. Chest 1991; 99: 600–08. 26 Mant A, Kimg M, Saunders N A, Pond C D, Goode E, Hewett H. Four-years follow-up of mortality and sleep-related respiratory disturbance in non-dement seniors. Sleep 1995; 18: 433–38. 22

(Revised version received July 1996)