INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Inhaled corticosteroids constitute a widely used and effective anti-inflammatory therapy for asthma. Airway inflammation is present even in patients with mild asthma, and therefore inhaled corticosteroids are now recommended to most asthmatics (1). This therapy is often long-term since its discontinuation often results in exacerbation of the disease (3). In addition to inhaled corticosteroids, asthmatics may also need oral steroid courses or even continuous oral steroid therapy. Long-term use of oral steroids is a well-known risk factor of osteoporosis (4, 5). Adverse effects of long-term inhaled steroids on bone metabolism have recently been investigated (6).

There are also studies on the effects of inhaled steroids on bone mineral density (BMD) (10, 19). The results of these studies are inconsistent, however. The reason may be that there have been different sexes (5, 10, 11, 20, 21) and age ranges in these studies. Menopause is considered to be the most potent risk factor for osteoporosis. Its effect on bone density may not have been properly taken into account in previous studies. Similarly, the use of oral steroids confounds the results (10, 19). To our knowledge, no population-based study on bone density in patients with asthma has been conducted. And whether asthma itself constitutes a risk factor for osteoporosis is unknown.

In the present work a random stratified population sample of perimenopausal women with asthma was studied using dual-energy X-ray absorptiometry (DXA) of the spine and femoral neck. The control group consisted of nonasthmatic perimenopausal women from the same population. The aim of the present study was to evaluate whether perimenopausal patients with asthma in a general population are at increased risk of osteoporosis. In addition, the effects of oral and inhaled corticosteroids on BMD were studied.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population

The target population consisted of all 14, 220 women 47 to 56 yr of age residing in Kuopio Province, Eastern Finland, in 1989 (22). The baseline postal enquiry was sent to 14,121 of them in 1989, and 13,100 (92.8%) responded. The enquiry included questions about gynecological history, use of female hormones, parity, height, weight, physical activity, consumption of milk products, smoking, morbidity, and current use of prescribed drugs.

A total of 11,055 respondents (84.4%) were willing to undergo bone densitometry. BMD was measured in a random stratified sample of 3,222 women (29.1%) by using DXA (Lunar DPX, Madison, WI) at the spine (L2-L4) and the left femoral neck during 1990 and 1991 (23, 24). After exclusion of nonvalid BMD measurements, the study population consisted of 2,941 women (108 with asthma) for spinal and 3,202 women (119 with asthma) for femoral BMD analyses.

Another postal enquiry was sent during 1990 and 1991 to those women who underwent bone densitometry. It included questions about use of coffee, alcohol, and amount of physical activity. Weekly hours of regular physical exercise, kilometers of daily walking or running, and physical demands of work (1/sitting, 2/light, 3/medium, 4/ heavy) were combined into a three-category general physical activity score. Both questionnaires were checked through at the time of densitometry. The results from these questionnaires have been presented elsewhere (22, 23, 25).

Menopausal status at the time of the second enquiry was considered as postmenopausal if 6 mo had elapsed since the last menstruation. The grip strength of the dominant arm was assessed with a hand dynamometer (Martin Vigorimeter, Germany), as a mean of three measurements.

Patients with Asthma

Seven hundred seventy-five (5.9%) of the women who returned the baseline postal enquiry reported having asthma or other chronic pulmonary disease. An additional (third) questionnaire was sent to those 214 who had undergone densitometry. Special interest was paid to questions concerning the use of inhaled corticosteroids, their daily doses, durations, regularity of use, and the use of spacers. With regard to oral corticosteroids, the number of courses, their durations, daily doses, and possible regular use (more than 6 mo) was investigated. To test the validity of the asthma diagnosis and to check the data concerning drugs, the medical records were reviewed. Furthermore, 36 women were called by phone.

Eighty-two percent of the 214 patients returned the asthma questionnaire. According to the questionnaire and after checking the medical records, 119 women were considered to have definite asthma in 1990 (55.6%). The asthma diagnoses were based mainly on peak expiratory flow measurements. The rest of the asthma diagnoses were based on the results of spirometry, ergometry, and the inhalation challenge test, or were clinical diagnoses fulfilling the ATS criteria (26). The mean age of the asthmatic women at the time of BMD measurement was 53.7 ± 2.9 (SD) yr (range, 48 to 59 yr). Characteristics of the asthmatics and nonasthmatics are presented in Table 1.

Asthma was diagnosed at a mean age of 44.1 ± 8.4 (SD) yr (range, 13 to 57 yr). Seventy-two (60.5%) of the asthmatics used inhaled corticosteroids in 1990. The majority of them, 90.3% (n = 65), used inhaled corticosteroids regularly. Seventy-one women (59.7%) also reported having allergic rhinitis, but only 24.4% of them (n = 29) used intranasal topical steroids. Ten asthmatics had used oral corticosteroids regularly for more than 6 mo. Sixty-one (51.3%) asthmatic women reported at least one oral steroid course (Table 2).

Statistical Analysis

The statistical analyses were carried out by using the SPSS/PC program. The data are expressed as means with SDs. The mean BMD values of asthmatics are also expressed as z-score (i.e., SD units) deviations from the nonasthmatic population means. Student's two-tailed unpaired t tests were used to test differences in continuous variables between the asthmatics and the nonasthmatics.

In order to study the effect of corticosteroids on bone, patients with asthma were divided into three groups according to their corticosteroid use: 28 of them (23.5%) had never used corticosteroids (Group I), 26 (21.9%) had used only inhaled corticosteroids (Group II), and 65 (54.6%) had used oral corticosteroids (Group III). Group III included women with regular use of inhaled corticosteroids and 19 women with only oral steroids.

Analysis of variance was used to test differences in continuous variables between the asthma groups (Group I to III). The differences were located by the Newman-Keuls test. Analysis of covariance was used to adjust for weight, height, grip strength, age, and menopause. Category data were compared by the chi-square test. The relationship between variables within the asthma group was tested with Pearson's correlation coefficient. Multiple stepwise regression analysis was used to assess independent determinants of BMD and to derive equations for the prediction of bone density in asthmatics by using the following continuous variables: age, years since menopause, menopause age, duration of hormone use, parity, height, weight, BMI, alcohol and coffee comsumption, daily kilometers of walking or running, calcium intake, number of oral corticosteroid courses, duration and doses of regular oral corticosteroid treatment, duration and daily dose of inhaled corticosteroid treatment, duration of asthma, and age at the time of asthma diagnosis.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Preliminary examination of the data showed no statistically significant differences between the asthmatics and the non-asthmatics except in body mass index (BMI) (Table 1). Between the different asthmatic patient groups there were statistically significant differences only in age at the time of asthma diagnosis, which was lowest for those who had used oral corticosteroids (Table 1).

Only 24% of asthmatics, but 32% of nonasthmatics reported no other health disorders (p < 0.05). Hypertension, lactose intolerance, venous thrombosis in a leg, and pulmonary embolism were statistically significantly more prevalent in asthmatics. There were no significant differences in menopausal status, HRT use, or BMDs between the asthmatics and the nonasthmatics (Table 1). Lastly, there were no statistically significant BMD differences between the three asthma groups (Table 1), and adjustment for weight, height, grip strength, age, and menopause did not change the results.

In asthmatics, age and time since menopause correlated negatively with spinal BMD, whereas weight and BMI correlated positively with both spinal and femoral BMD. The durations of inhaled corticosteroid and of regular oral corticosteroid treatments correlated negatively with spinal BMD. The latter duration also correlated negatively with femoral BMD, whereas hormone repleacement therapy (HRT) correlated positively with femoral BMD (Table 3).

The asthmatics with no HRT use had significantly lower spinal and femoral BMDs than did the corresponding non-asthmatics (spinal BMD, 1.083 ± 0.150 [SD] versus 1.128 ± 0.160 g/cm², p < 0.05; femoral BMD, 0.894 ± 0.112 [SD] versus 0.929 ± 0.128 g/cm², p < 0.05) (Figure 1). The z-score values were 0.29 for spinal BMD and 0.27 for femoral BMD in these asthmatics without HRT (Figures 2 and 3). Spinal but not femoral BMD was also lower in the asthmatics receiving oral corticosteroids than in the corresponding nonasthmatics (no HRT) (Figure 1).

View larger version (16K): [in this window] [in a new window]   Figure 1.   Bone mineral density (BMD) with 95% CIs of the spine and femoral neck in asthmatics, in their subgroups, and in nonasthmatics with no HRT. Spinal and femoral BMDs were significantly lower in the asthmatics than in the nonasthmatics, and spinal BMD was significantly lower in the asthmatics receiving oral corticosteroids than in the nonasthmastics (p < 0.05).

View larger version (15K): [in this window] [in a new window]   Figure 2.   The z-score values with 95% CIs are shown for spinal BMD of the asthmatics and their different subgroups. The z-score is the standard deviation from the mean of the nonasthmatic population.

View larger version (17K): [in this window] [in a new window]   Figure 3.   The z-score values with 95% CIs are shown for femoral neck BMD of the asthmatics and their different subgroups. The z-score is the standard deviation from the mean of the nonasthmatic population.

Multiple regression analysis showed that the duration of inhaled corticosteroid treatment, years since menopause, age at the time of asthma diagnosis, dose (mg) of regular oral corticosteroid treatment, and duration of hormone use were significantly associated with spinal BMD (R² = 0.50, p < 0.001). Weight and HRT were significantly associated with femoral BMD (R² = 0.31, p < 0.001). The regression equations were: Spinal BMD = 1.565  0.018 ·Duration of inhaled corticosteroid treatment (years)  0.012 ·Years since menopause  0.008 · Age at the time of asthma diagnosis (years)  0.048 · milligrams of regular oral corticosteroid treatment + 0.022 · HRT (years).

Femoral Neck BMD = 0.534 + 0.005 · Weight (kg) + 0.015 · HRT (years).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study showed that asthmatic women from a general perimenopausal population had slightly decreased spinal and femoral neck BMDs when compared with nonasthmatic women. These differences were more prominent and statistically significant in women without HRT. It is well known that HRT users have higher lumbar BMDs (20, 25, 27). In this respect the exclusion of HRT use in some analyses can be justified to show the pure effect of asthma and its treatment on bone density.

To improve the validity of asthma diagnosis the medical records were reviewed. Cross-sectional design of this study allows only very cautious causal considerations. Selection and differential recall of hormone or corticosteroid treatments might also have biased the results. However, there is no reason to assume that the asthmatics would have recalled their HRT in a way different from that of the nonasthmatics. It is also unlikely that the asthmatics with low BMD had recalled their corticosteroid use in a way different from that of the asthmatics with normal BMD.

The duration of inhaled corticosteroid treatment and the duration of regular oral corticosteroid treatment correlated negatively with spinal BMD. Furthermore, in multiple regression analysis the duration of inhaled corticosteroid treatment and the daily dose of regular oral corticosteroid treatment were significantly associated with spinal BMD. However, with regard to femoral neck BMD the deleterious effect of corticosteroids was not so evident. Femoral neck BMD was negatively associated only with the duration of regular oral corticosteroid treatment, and in multiple regression analysis corticosteroid treatments were not associated with femoral BMD. The explanation for this might be to do with the amount of trabecular bone, which is more susceptible to corticosteroid treatment (5, 17, 28,). However, decreases of femoral BMD resulting from corticosteroid treatment have also been reported (11, 16).

To our knowledge, no population-based study of BMD in asthmatics has been conducted before, and the results of previous studies concerning BMD in cases of asthma, as well as the effect of inhaled corticosteroids on bone loss, have been inconsistent. Reasons for this inconsistency may be that there have been mixed patient groups of men and women in same studies (5, 10, 20, 21), age ranges have varied, the effect of menopause on bone density has not been adequately controlled, and the effect of oral steroids has not always been controlled (10, 19). It is also important to take into account the age when the subjects started their inhaled or oral corticosteroid treatment. It can be assumed that at least long-term oral corticosteroid treatment has an adverse effect on the skeleton before the peak bone mass has been reached. In our study, the mean age when inhaled corticosteroid treatment was started was 49.0 ± 4.9 (SD) yr. Hence, most asthmatics had reached their peak bone mass before the corticosteroid treatment. Ruegsegger and colleagues (17) showed that young patients (< 40 yr) receiving corticosteroids lose bone more rapidly than do older patients (> 50 yr), not including postmenopausal women.

There are also other factors that should be borne in mind when interpreting the results. The number of asthmatics is relatively small and they are likely to have mild disease because of the population-based study design. This may rather underestimate the BMD differences between the asthmatics and the nonasthmatics. Also, if the mean BMI in women with asthma is higher, as was the case in our study, this may result in underestimation of the differences in BMD because weight correlates positively with BMD.

Physical activity during leisure was similar in the asthmatics and the nonasthmatics. Hence, it can be assumed that most of the asthmatics did not have severe asthma, and exercise differences do not explain our results. The asthmatics reported lactose intolerance more often than did the nonasthmatics. However, the mean dietary daily calcium intake was similar in the asthmatics and the nonasthmatics, and lactose intolerance was not a common disorder (7%). Lactase intolerance has only a slight lowering effect on BMD (29).

Taking full account of oral corticosteroid courses and regular oral corticosteroid use was one strength of our study. The asthmatics who had used oral corticosteroids but had had no HRT had lower spinal and femoral BMDs than did corresponding nonasthmatics. Regular oral corticosteroid treatment was also associated with low spinal BMD in multiple regression analysis. There is no consensus about skeletal effects of cumulative doses, daily doses, or duration of oral corticosteroid treatment, but bone loss has been reported to be most rapid during early phases of corticosteroid treatment (18, 28).

Previous studies have questioned the effect of inhaled corticosteroids on bone mass. However, oral corticosteroid courses may have confounded the results. In our study the mean spinal and femoral BMDs were not significantly different in the group with inhaled corticosteroids compared with the group without steroids. However, the duration of inhaled corticosteroid treatment was negatively associated with spinal BMD and it was also associated with spinal BMD in multivariate analysis. The mean duration of inhaled corticosteroids was 5.2 yr and the mean daily dose was 1.0 mg. In previous studies, the duration of inhaled corticosteroid treatment varied, and it has not always been specified (15). It is not known whether the cumulative or daily dose or the duration of inhaled corticosteroid treatment is the most important risk factor for osteoporosis.

There were also 28 asthmatics with no history of corticosteroid use. This allowed us to evaluate the independent effect of asthma on bone mass. The duration of menopause was shorter in these subjects than in corticosteroid users, and this could have a positive effect on BMD. Their spinal and femoral BMDs were higher than those in the other groups of asthmatic patients and did not differ from those in the nonasthmatic population. Hence we can assume that asthma itself, without corticosteroids and with normal ambulation, is not a major risk factor with regard to osteoporosis.

Hormone replacement therapy is a well-known protective factor with regard to bone loss. In agreement with the results of previous studies (4, 20, 27), we also noticed a positive association between HRT and BMD. In fact, significant BMD decreases were shown only in women without HRT. This indicates that HRT can be effectively used in the prevention of osteoporosis in postmenopausal asthmatic women.

An association between HRT and asthma has been reported (30). In our series, asthma was diagnosed at the mean age of 44 yr, whereas the women had used HRT 2.4 yr on average by densitometry (with a mean age of 54 yr). Thus, we can assume that asthma, as a rule, was diagnosed premenopausally before HRT was started. Therefore, we suggest that there was no association between HRT and asthma in our study.

Osteoporosis may well be an acceptable hazard for a minority of patients with severe asthma, but the majority with mild asthma should not be exposed to the risk. It is important to identify patients at high risk of osteoporosis. Known risk factors account for less than half the variability of bone mass in various studies, and they do not accurately predict bone mass in individual patients (23, 25). Thus, measurement of BMD is important. However, it is also important to map the risk factor status as well since some risks can be avoided. It is easier to prevent osteoporosis when there is sufficient bone mass left than to treat established disease. In clinical situations the possibility of HRT use should be considered in each postmenopausal asthmatic woman using at least oral corticosteroids. It is important to decrease the doses of inhaled corticosteroids if possible and keep oral courses to a minimum. By treating asthmatics well with inhaled corticosteroids their osteoporosis risk can perhaps be decreased if they do not need so many oral courses and also have a better level of physical activity.

In conclusion, our results suggest that women with asthma in a general population represent a risk group with regard to osteoporosis. Oral corticosteroids represent the greatest risk. The use of inhaled corticosteroids may negatively affect spinal BMD. On the other hand, the results suggest that HRT is beneficial in asthmatic women. Follow-up of this study population is expected to provide more exact information about causes of bone loss in these women.