Association between self-rated health and asthma: a population-based study

Authors


  • Authorship and contributorship Jörgen Syk participated in study design, data analysis and writing the manuscript. Kjell Alving participated in data analysis and in writing the manuscript. Anna-Lena Undén participated in study design and performed the study. She also took part in data analysis and in writing the manuscript.

  • Ethics The study protocol was approved by the local ethical committee at Karolinska Hospital, Stockholm, Sweden (Dnr: 94-289).

  • Conflicts of interest The authors have stated explicitly that there are no conflicts of interest in connection with this article.

Jörgen Syk, MD, Runby Vårdcentral, P.O. Box 474, SE-194 04 Upplands Väsby, Sweden. Tel: +46 8 590 990 60/+46 70 769 37 01 Fax: +46 8 590 317 65 email: jorgen.syk@ptj.se

Abstract

Introduction:  Self-rated health (SRH) is a relevant measure of health as it can predict morbidity, mortality and health-care use. Studies have shown an association between poor SRH and elevated levels of circulating inflammatory cytokines. It is therefore interesting to learn more about the association between asthma, a chronic inflammatory disease with a recognised systemic component and SRH.

Objectives:  To compare SRH ratings in respondents with and without current asthma. A second aim was to compare SRH with quality-of-life ratings in the same groups.

Methods:  In 1995, we randomly selected 8200 persons ≥18 years from the population of Stockholm County, Sweden and mailed them a questionnaire. A total of 5355 persons (67.5%) responded. Respondents were divided in two groups, those with and those without current asthma. The groups were further divided by sex and age (18–44 and ≥45 years). SRH was measured with the question ‘How do you rate your general health status?’ and quality of life with the Gothenburg Quality of Life Instrument and the Ladder of Life.

Results:  Respondents with asthma rated their health significantly worse than did those without asthma, except women aged 18–44 years. SRH was associated at least as strong as quality of life to asthma with the advantage of being easier to apply (only one item).

Conclusion:  Information on SRH is easy to obtain and represents an important dimension of health status that potentially can be used as a complement to identify patients who need extra attention to manage their asthma and its consequences.

Please cite this paper as: Syk J, Alving K and Undén A-L. Association between self-rated health and asthma: a population-based study. Clin Respir J 2012; 6: 150–158.

Introduction

Self-rated health (SRH), the answer to the question ‘How do you rate your general health status?’, is a relevant measure of health, as it can predict morbidity and mortality (1). Mossey and Shapiro, writing in 1982 (2), were the first to demonstrate that poor SRH can predict mortality independent of objective health status. Since then, several studies in various populations have confirmed that SRH is a powerful predictor of morbidity and mortality, even after controlling for confounders such as medical diagnosis, functional ability and psychosocial factors (1, 3–5). SRH can also predict hospitalisation and frequent examinations in outpatient clinics (6–8). Because SRH is a useful predictor of morbidity, mortality and other health outcomes, it seems logical to hypothesise that it would be useful to register SRH in clinical practice and use it as a complement to the ordinary examination to find patients with unsatisfactory health status.

Little is known about the biological mechanisms that connect poor SRH with morbidity and mortality. A number of recent studies have shown an association between poor SRH and elevated levels of circulating inflammatory cytokines (9–11). Inflammatory cytokines can affect both the immune system and the brain and give rise to a sickness response that includes symptoms such as depression, tiredness, loss of appetite and loss of interest in one's environment (12, 13). This link makes it interesting to look at the association between inflammatory diseases and SRH. In this study, we have focused on investigating the relationship between asthma, a chronic inflammatory disease with a recognised systemic component, and SRH. Asthma-related inflammation is not restricted to local inflammation in the central airways. Changes also occur in the nose, sinuses and small airways. There is also communication between the airways and the bone marrow through inflammatory cytokines in circulating blood (14).

In recent years, quality of life, measured with various questionnaires, has become an important factor in the evaluation of asthma and asthma treatment in studies (15–18). Although SRH is strongly correlated with health-rated quality of life (19), knowledge about the association between asthma and SRH among adults remains limited. Some studies report an association between asthma and SRH among elderly persons, but little is known about the relationship between asthma and SRH in younger adults (20–22).

The aim of this study was to compare SRH in respondents with and without current asthma. A second aim was to compare SRH with quality-of-life ratings in the same groups.

Materials and methods

Study group

A total of 8200 persons aged ≥18 years was randomly selected from the population register of two health-care regions in Stockholm County: the north-western health-care region and Södertälje health-care region. All data were collected from the answers to a public health questionnaire posted to the potential respondents in the spring of 1995. The original purpose of the questionnaire was to investigate health, quality of life and health-care use in the population. After two reminders, 5355 persons (67.5%) responded. Persons reported dead, no longer living at the address or temporarily absent from the address (by relatives or the post office) were excluded (271 persons). Respondents were divided into those with and those without asthma (please see below for more details). Both groups were further divided by sex and age (18–44 and ≥45 years). The cut-off point between the age groups was chosen to make the groups approximately equal in size and to minimise the risk of including persons with chronic obstructive pulmonary disease (COPD) in the younger age group.

Health variables

Asthma

Respondents were categorised as having asthma (i.e. current asthma) if they answered yes to both of two questions: ‘Do you have or have you had asthma?’ and ‘Do you use any asthma medicine?’ All others were included in the group of respondents who did not have asthma.

Self-rated health

Information on SRH was gathered using the question ‘How do you rate your general health status?’ There were five response alternatives: ‘very good’, ‘quite good’, ‘neither good nor poor’, ‘quite poor’ and ‘very poor’. In the text, the health of those who chose the alternative ‘neither good nor poor’ is referred to as ‘fair’(23).

Quality of life

The Gothenburg Quality of Life Instrument (GQLI) was used to assess social, physical and mental well-being. Social well-being was rated on the basis of answers to questions about home and family situation, housing, work situation, personal finances, health and leisure time. Physical well-being was rated on the basis of answers to questions about hearing, vision, memory, fitness and appetite. Mental well-being was rated on the basis of answers to questions about mood, energy, patience, self-confidence and sleep (24). All items were rated on a 7-point Likert scale from ‘very poor’ (1 point) to ‘excellent, could not be better’ (7 points).

Quality of life was also measured with Cantril's Ladder of Life, a 10-point scale that respondents used to rate their present, past (1 year ago) and future (1 year from now) quality of life (25, 26). The highest step on the ladder represents the ‘best possible life’ and the lowest step the ‘worst possible life’. The higher the score, the better.

Physical health

We used the question ‘Are you suffering from any of the following chronic diseases? (yes or no)’ as a measure of physical health. The chronic diseases in question included cardiovascular disease, diabetes, hypertension, joint/musculoskeletal disease and chronic pain. The total number of positive responses (0–5) constituted the variable physical health.

Sociodemographic factors

Educational status was divided into three levels: Compulsory schooling only (7 years of school before 1961 and 9 years thereafter), secondary schooling (10–12 years) and university education. The questionnaire included six response alternatives for occupational status: ‘working’, ‘studying’, ‘unemployed’, ‘on sick leave’, ‘retired’ and ‘other’. Few respondents in the 18–44-year age group chose ‘unemployed’, ‘on sick leave’, ‘retired’ or ‘other’, so for the purposes of analysis, we collapsed these categories in a new category, ‘other younger’. Few respondents ≥45 years chose ‘studying’, ‘unemployed’, ‘on sick leave’ or ‘other’, so for the same reason, these categories were also collapsed into a new category, ‘other older’. Respondents were divided into two categories on the basis of whether they were married or cohabiting (the first category) or not married or cohabiting (the second category).

Physical activity during leisure time in the past year was also reported. There were four response alternatives: ‘inactive leisure time’, ‘moderate physical activity during leisure time’, ‘moderate regular physical activity during leisure time’ and ‘regular physical activity and exercise’. The two groups with most frequent exercise were merged into one group for the purpose of logistic regression analysis. Respondents were categorised into smokers and non-smokers on the basis of their response to the question ‘Do you smoke?’ (Tables 1a and 1b)

Table 1a. Sociodemographic characteristics of men by age and self-report of asthma or no asthma*
CharacteristicsMen
18–44 years≥45 years
Asthma n = 58No asthma n = 1058PAsthma n = 65No asthma n = 1303P
  • *

    Differences between the groups with and without asthma were analysed with χ2 tests and t-tests. There were no significant differences between the groups with and without asthma except by occupational status and age for men ≥ 45 years.

  • Few respondents in the 18–44-year age group chose response alternatives ‘unemployed’, ‘on sick leave’, ‘retired’ or ‘other’, so for the purposes of analysis, we collapsed these categories into a new category, ‘Other younger’.

  • Few respondents in the ≥45-year age group chose the response alternatives ‘studying’, ‘unemployed’, ‘on sick leave’ or ‘other’, so for the purposes of analysis, we collapsed these categories into a new category, ‘Other older’.

  • ns, not significant; SD, standard deviation.

Age, mean (SD)29.5 (7.58)31.0 (7.49)ns62.2 (11.30)59.2 (10.38)<0.05
Educational level in %  ns  ns
 Compulsory22.619.5 68.963.9 
 Secondary62.360.6 14.816.4 
 University15.119.8 16.419.7 
Married/cohabitating in %71.765.5ns80.679.6ns
Occupational status in %  ns  <0.01
 Working65.573.2 36.956.6 
 Studying20.013.9  
 Retired 56.934.5 
 Other younger14.512.9  
 Other older 6.28.8 
Physical activity, mean rank2.32.6ns2.02.2ns
Smoking in %15.520.1ns23.122.1ns
Table 1b. Sociodemographic characteristics of women by age and self-report of asthma or no asthma*
CharacteristicsWomen
18–44 years≥45 years
Asthma n = 113No asthma n = 1202PAsthma n = 105No asthma n = 1451P
  • *

    Differences between asthma groups were analysed with χ2 tests and t-tests. There were no significant differences between the groups of women with asthma and the corresponding groups of women without asthma.

  • Few respondents in the 18–44-year age group chose response alternatives ‘unemployed’, ‘on sick leave’, ‘retired’ or ‘other’, so for the purposes of analysis, we collapsed these categories into a new category, ‘Other younger’.

  • Few respondents in the ≥45-year age group chose the response alternatives ‘studying’, ‘unemployed’, ‘on sick leave’ or ‘other’, so for the purposes of analysis, we collapsed these categories into a new category, ‘Other older’.

  • ns, not significant; SD, standard deviation.

Age, mean (SD)31.3 (7.80)31.5 (7.52)ns60.7 (12.02)60.8 (11.70)ns
Educational level in %  ns  ns
 Compulsory19.319.9 63.470.4 
 Secondary53.257.4 15.113.7 
 University27.522.8 21.515.9 
Married/cohabitating in %65.171.6ns61.463.8ns
Occupational status in %  ns  ns
 Working62.767.6 45.151.4 
 Studying19.115.1  
 Retired 43.140.5 
 Other younger18.217.3  
 Other older 11.88.2 
Physical activity, mean rank2.32.6ns2.12.1ns
Smoking in %32.727.1ns24.821.6ns

Statistical methods

Differences between the groups with and without asthma were analysed using t-test and χ2 analyses. To exclude the possibility that the association between SRH and asthma was because of confounding factors, we performed a multivariate analysis with fair/poor SRH as a dependent variable and education, physical activity, smoking and marital status as independent variables (Tables 4a and 4b).

Table 4a. Results of logistic regression analysis, showing associations between fair/poor self-rated health in men by sociodemographic characteristics in a full model and an age-adjusted model
 Men
18–44 years ≥45 years
Age-adjusted model OR (95% CI)Full model OR (95% CI) Age-adjusted model OR (95% CI)Full model OR (95% CI)
  1. CI, confidence interval; OR, odds ratio.

18–29 years1145–64 years11
30–44 years1.16 (0.85–1.59)0.77 (0.53–1.14)≥65 years1.41 (1.09–1.83)1.36 (1.01–1.83)
No asthma11 11
Asthma2.85 (1.62–5.00)2.77 (1.46–5.26) 2.83 (1.71–4.69)3.05 (1.74–5.34)
Education     
Compulsory 1  1
 Secondary 0.73 (0.48–1.11)  0.98 (0.67–1.45)
 University 0.39 (0.21–0.72)  0.70 (0.47–1.02)
Physical activity     
 High 1  1
 Medium 2.60 (1.70–4.00)  2.17 (1.49–3.16)
 Low 4.54 (2.81–7.38)  4.92 (3.20–7.59)
Smoking     
 No 1  1
 Yes 1.50 (1.00–2.25)  1.45 (1.06–1.99)
Married/cohabitating     
 Yes 1  1
 No 1.54 (1.08–2.22)  1.90 (1.38–2.60)
Table 4b. Results of logistic regression analysis, showing associations between fair/poor self-rated health in women by sociodemographic characteristics in a full model and an age-adjusted model
 Women
18–44 years ≥45 years
Age-adjusted model OR (95% CI)Full model OR (95% CI) Age-adjusted model OR (95% CI)Full model OR (95% CI)
  1. CI, confidence interval; OR, odds ratio.

18–29 years1145–64 years11
30–44 years1.23 (0.93–1.64)1.13 (0.82–1.56)≥65 years1.51 (1.21–1.91)1.01 (0.76–1.33)
No asthma11 11
Asthma1.19 (0.75–1.90)1.18 (0.72–1.95) 2.71 (1.80–4.08)2.78 (1.78–4.36)
Education     
Compulsory 1  1
 Secondary 0.75 (0.52–1.08)  0.99 (0.68–1.44)
 University 0.58 (0.37–0.91)  0.76 (0.52–1.11)
Physical activity     
 High 1  1
 Medium 1.52 (1.07–2.16)  1.89 (1.33–2.69)
 Low 2.62 (1.71–4.01)  5.07 (3.39–7.59)
Smoking     
 No 1  1
 Yes 1.40 (1.02–1.94)  1.07 (0.80–1.44)
Married/cohabitating     
 Yes 1  1
 No 1.52 (1.11–2.08)  1.34 (1.03–1.74)

SRH was dichotomised into ‘good health’ (response alternatives ‘very good’ and ‘quite good’) and ‘fair/poor health’ (response alternatives ‘neither good nor poor’, ‘quite poor’ and ‘very poor’). We also tried the alternative to dichotomise SRH into ‘good/fair health’ and ‘poor health’ but the group of respondents who reported ‘poor health’ (response alternatives ‘quite poor’ and ‘very poor’) was too small to obtain a good statistical model for multiple regression analyses. We estimated two logistical models, one age-adjusted and one full model by sex and age group. Results are shown with odds ratios (ORs) and 95% confidence intervals. The fit of the models was assessed by the Hosmer–Lemeshow goodness-of-fit test. The models were considered acceptable if P > 0.05, and all models met this demand. Analysis was also performed with occupational status and physical health as independent variables; this did not affect the association between fair/poor SRH and asthma (data not shown). Statistical analyses were performed with SPSS version15 (IBM, New York, USA) and Stata/IC10.0 (StataCorp LP, Texas, USA).

Results

Characteristics of the study group

The study group consisted of 5355 people (54% women), 18–100 years of age.

A total of 341 of the participants (6.4%) reported that they had asthma, including 123 men (2.3%) and 218 women (4.1%). There was a significant difference in age and occupational status between men ≥45 years with and without asthma. The men with asthma were older, fewer of them were employed and more of them were retired (Table 1a).

Association between asthma and SRH and asthma and quality of life

There was a significant difference in SRH between those with and those without asthma. Respondents with asthma rated their health worse. This association was more marked in respondents ≥45 years (Table 2). The most common response to the SRH question in each group was ‘quite good’. However, respondents with asthma rated their general health as less than good approximately twice as often as respondents without asthma except among women aged 18–44 years. Approximately half the respondents with asthma who were ≥45 years rated their health as less than good. Over 20% in this group rated their health as poor – twice as many as those without asthma who were also ≥45 years.

Table 2. Comparison of self-rated health in men and women by age and self-report of asthma or no asthma* (data given in percent)
Variable18–44 years≥45 years
AsthmaNo asthmaχ2PAsthmaNo asthmaχ2P
  • *

    Differences between the groups with and without asthma were analysed with χ2 tests.

 Men
   19.8<0.01  34.8<0.001
Very good13.834.4  7.720.5  
Quite good50.048.7  43.154.5  
Neither good nor poor27.613.2  26.217.9  
Quite poor8.63.3  23.16.2  
Very poor00.4  00.9  
 100100  100100  
 Women
   18.7<0.01  30.6<0.001
Very good18.628.7  5.921.1  
Quite good59.351.9  41.650.0  
Neither good nor poor8.813.6  30.718.6  
Quite poor10.65.3  18.89.0  
Very poor2.70.5  3.01.3  
 100100  100100  

The analysis of the answers to the questions on the GQLI showed no difference between men with and without asthma in the 18–44-year age group. However, quality of life was significantly lower for all three categories of well-being (social, physical, and mental) in men ≥45 years with asthma than in men in the same age group without asthma. In both age groups, women with asthma had significantly lower rates of social well-being than women without asthma.

The analysis of the answers to the Ladder of Life showed no difference between respondents with and without asthma in the 18–44-year age group. Respondents with asthma who were ≥45 years had significantly lower quality of life scores than respondents in the same age group without asthma, except for women predicting their future quality of life (Table 3).

Table 3. Comparison of quality of life, measured with the Gothenburg Quality of Life Instrument and the Ladder of Life, in men and women by age and self-report of asthma or no asthma*
Characteristics18–44 years≥45 years
AsthmaNo asthmaPAsthmaNo asthmaP
  • *

    Differences between the groups with and without asthma were analysed with t-tests.

  • GQLI, Gothenburg Quality of Life Instrument; ns, not significant.

Mean rankMen
GQLI      
 Social well-being5.005.24ns5.265.56<0.05
 Mental well-being5.095.24ns4.775.33<0.001
 Physical well-being5.405.59ns4.745.05<0.05
Ladder of Life      
 Present quality of life6.786.92ns6.577.35<0.05
 Past quality of life6.486.50ns6.697.26<0.05
 Future quality of life8.088.00ns6.257.48<0.01
Mean rankWomen
GQLI      
 Social well-being5.035.27<0.055.075.54<0.001
 Mental well-being5.015.11ns5.055.24ns
 Physical well-being5.495.61ns5.075.22ns
Ladder of Life      
 Present quality of life6.847.13ns6.617.13<0.05
 Past quality of life6.466.47ns6.326.90<0.05
 Future quality of life8.428.24ns7.057.35ns

Multiple regression analyses

Further analysis with multiple regression to exclude confounding factors in the relation between SRH and asthma showed a significant association between asthma and fair/poor SRH in all groups except women 18–44 years. Women ≥45 years who had asthma and all men who had asthma had approximately three times higher odds of fair/poor SRH than those in the corresponding sex and age groups who did not have asthma (Tables 4a and 4b).

Separate analyses comparing the prevalence of fair/poor SRH in respondents with and without asthma in different age groups (18–29, 30–44, 45–64 and ≥65) showed that differences in prevalence were greatest among men ≥65 years (data not shown).

Discussion

In this study, univariate analysis showed that in both sexes and both age groups (18–44 and ≥45 years), respondents with asthma rated their health significantly worse than did respondents without asthma. Further investigation with multiple regression analyses, controlled for various sociodemographic, lifestyle and health factors, confirmed the association between SRH and asthma, except among women 18–44 years.

There were few differences between respondents with and without asthma with respect to sociodemographic and lifestyle factors (Table 1a) and no major difference that could have affected the results of the analysis. The sole significant differences were found in male respondents who were ≥45 years. Within this group, the mean age of the men with asthma was 3 years higher than the mean age of the men without asthma. As a consequence, the percentage of those who were working was higher among the men without asthma and the percentage of those who were retired was higher among the men with asthma.

Three previous studies have had results similar to the results of the current study.

A Swedish study on chronic diseases and symptoms among middle-aged to old men and women found associations between asthma and poor SRH (OR 2.6) (21). A study from west Texas on SRH and quality of life in participants ≥65 years (20) and a population-based study on general self-reported health and quality of life in US adults (22) showed associations between fair/poor SRH and asthma (OR 2.71 and 2.26).

It may seem self-evident that people with asthma would have poorer SRH than people without asthma. However, most people with asthma have a mild to moderate form of the disease and should be able to live a normal life with the right treatment. The availability of good asthma treatments means that it is not a foregone conclusion that asthma is associated with poor SRH. For example, in a newly published Swedish follow-up study of a cohort of young adults with asthma, the asthma cohort and controls scored similarly regarding quality of life (27).

There are several possible explanations for the differences in SRH between respondents with and without asthma, such as lower well-being because of unsatisfactory asthma control, awareness of suffering from a chronic disease, fear and anxiety about getting worse or losing control of their asthma, the influence of circulating inflammatory cytokines or a combination of these alternatives.

The multivariate analyses in this study showed that younger men with asthma had poorer SRH than younger men without asthma, but the same was not true for younger women. One possible explanation is that asthma affects physical fitness, which may be more important to men than women in this age group. This hypothesis is supported by the results of a Finnish study that investigated the variables that affected the way men rated SRH. The researchers compared three age groups: 31–35, 51–55 and 71–75 years. They found that in the youngest age group, SRH was best explained by the symptoms the men felt and by the men's physical fitness; in the middle-aged group, by symptoms and well-being; and in the older group, by chronic diseases (28).

Quality of life, measured with various questionnaires (15–18), is being used in a growing number of studies to evaluate asthma treatment and can also be used to evaluate asthma treatment in daily care. In this study, we measured general well-being in three ways: with the GQLI, the Ladder of Life and SRH. Quality of life measured with the GQLI and the Ladder of Life was not as strongly associated with asthma as SRH. Moreover, GQLI and the Ladder of Life showed no significant difference in quality of life between young men with and without asthma, whereas a significant difference in SRH was found between the two groups. In summary, asthma was more strongly and consistently associated to SRH than with quality of life as measured by the instruments in this study. Thus, although there is a strong correlation between SRH and quality of life (19), it seems that they may reflect and include different things. The SRH question is focused on health and is a more comprehensive and inclusive measure than the GQLI. It can cover dimensions of health that could not be covered by more detailed or guided questions. Another possible explanation is that a poorly controlled asthma may be characterised by elevated levels of circulating inflammatory cytokines, which in turn may create various interoceptive sensations that can be sensed and incorporated in the self-assessment process.

The present study has some limitations: the data are self-reported and do not include total tobacco exposure, type of asthma or asthma duration. It could also be considered a limitation that the study data were collected in the mid-1990s. Since then, there has been a change in prescription habits in Sweden towards more use of combination therapy (corticosteroids and long-acting β2 agonists) but no major change in levels of asthma control (29). Additionally, it is possible that patients with COPD were included in the study population. Until the early 1990s, the Swedish health-care system provided medication free of charge to patients with asthma but not to patients with COPD. Because of that, some patients with COPD may have been classified as having asthma. However, this is not likely to have influenced the association between SRH and asthma since adjusting for smoking in multivariate analysis did not affect the association between SRH and asthma in the older age group and the proportion of respondents with asthma in this study is very close to the prevalence of asthma in Sweden (30, 31). Strengths of the study include the large population-based sample and the wide age range of respondents (18–100 years).

The importance of SRH as an independent predictor of morbidity, mortality and health-care use by SRH has previously been reported mainly in behavioural science literature. Many researchers are familiar with this association, but it is not as well-known among clinicians. SRH can capture aspects of health that cannot easily be measured during a consultation and can provide a fuller picture of a patient's health status than otherwise might be available. SRH can easily be noted in a patient's medical record and followed over time. Improvement in SRH is associated with reduced risk of mortality (32). Human suffering could be prevented and health-care costs reduced by adding information on SRH to consultations.

A potentially useful step would be to conduct randomised prospective studies to test the usefulness of SRH as a screening tool for the identification of patients with unsatisfactory asthma control. It is important to better understand how asthma affects SRH, and conversely, how impaired SRH may affect asthma control. Further research into the association between SRH and systemic inflammatory markers in asthma, such as cytokines, would be of great interest (9, 10, 14).

Conclusions

In this population-based study, we found that respondents with asthma had lower SRH than respondents without asthma, except women 18–44 years. SRH was associated at least as strong as quality of life to asthma with the advantage of being easier to apply (only one item). Information on SRH is easy to obtain and represents an important dimension of health status that potentially can be used as a complement to the ordinary examination to identify patients who need extra attention to manage their asthma and its consequences.

Acknowledgements

This work was supported by PickUp funds from the Stockholm County Council and by a grant from Centre for Allergy Research, Karolinska Institutet. Scientific Editor Kimberly Kane of the Centre for Family and Community Medicine, Karolinska Institutet, revised the language and provided useful comments on the text.

Ancillary