Introduction
Quality of life is an increasingly important health goal1 and has begun to be used as a health outcome measure.2 This is especially the case in situations where health interventions cannot be aimed mainly at prolonging life, as occurs for very old persons and persons with a terminal illness.
Several authors have defined the concept of quality of life, and subjectivity is a constant feature of these definitions. For example, according to the WHO,3 quality of life is «an individual´s perception of their position in life in the context of the cultural and value systems in which they live, and in relation to their goals, expectations, standards and concerns.» This is a broad concept which comprises elements such as physical health, psychological state, level of independence, social relationships and relationship with the environment.
Health-related quality of life (HRQOL), the concept used most widely in health sciences, evaluates the repercussions of illness or health status, treatment, and other health care measures, on the patient´s social and personal life.4,5
A number of questionnaires and other instruments have been developed to evaluate HRQOL. Some instruments are generic and can be used either for the general population or specific groups of patients; others are specific for patients with a particular disease, individuals within a specific population group, etc.6
An extensive review of published studies7 detected few studies in Spain or other countries that used questionnaires or similar instruments to look at the quality of life of older persons in the general public. The aim of our study was to describe self-perceived health status and quality of life in noninstitutionalized persons older than 65 years residing in health care districts 2 and 4 in the Community of Madrid, an autonomous region in central Spain.
Material and methods
Participants
The target population was noninstitutionalized persons older than 65 years residing in health care districts 2 and 4 in Madrid, and able to respond to all items on the questionnaires on their own.
To calculate sample size we used as a reference the proportion of persons older than 65 years (6.7%) who considered their health to be poor or very poor according to a health survey carried out in Barcelona in 1986.8 The size necessary to detect a proportion of 6.5% of a finite population at the 95% confidence level and a precision of 1.5% was calculated as approximately 1000 persons. A sample of 3000 persons was selected (two substitutes per person) from the municipal census by systematic sampling with a random start algorithm. The field work was done in June 1999 to April 2000. A pilot study with the first 50 questionnaires showed that persons older than 65 years understood the instrument and answered the questions appropriately.
Deceased persons and those for whom the census was in error were excluded, as were those who had moved to a different address outside the two health care districts in the study, those living in residential facilities or who were hospitalized at the time of the study, those with moderate or severe cognitive impairment (score higher than 4 on the Pfeiffer test) and those who were unable to communicate with the interviewer for whatever reason (severe hearing impairment, language, dysarthria, dysphasia, etc.).
Data collection
This article describes and analyzes some of the variables from a more extensive questionnaire.7 The structured instrument consists of a number of items often used in other questionnaires to evaluate clinical status and social support in older persons. Tests that have been validated by others were also used.
Self-perceived health status was evaluated with the appropriate items from the National Health Survey.9 One item was «Which words best describe your current health status? Very good, good, fair, poor or very poor?» Quality of life in persons older than 65 years was investigated with the EuroQol6 instrument, and with the NHP6. The EuroQol instrument describes health status in five dimensions (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression), each of which is scored at three levels of function. The instrument also contains a millimeter visual analogue scale (VAS) on which the participant scores his or her health state on the day of the interview: a score of 0 indicates worst imaginable health, and a score of 100 indicates best imaginable health. The NHP consists of a series of yes-or-no items grouped in six domains (energy, pain, physical mobility, emotional reactions, sleep and social isolation). Each domain is scored on a scale from 0 to 100, with higher scores indicating greater problems.
Sociodemographic variables (age, sex, marital status, level of education and social class), economic resources (economic dependence and sufficiency of family income to cover basic daily needs), social and familial support (persons in the household, time spent alone during the day, presence of someone who acts as a caregiver and confidant), physical exercise (usual sports or frequent long walks, walks only near the home, walks only inside the home, bed-to-chair) were studied. Functional capacity was examined with the Older Americans Resources and Services Multidimensional Functional Assessment Questionnaire(OARS-MFAQ) dimensions of basic and instrumental activities of daily life. Mental health was evaluated with the Pfeiffer test for cognitive impairment and the Hospital Anxiety and Depression scale (HAD) for mood.
To relate the profession or occupation of the head of the household during most of his or her life as determined in the questionnaire9 to social class according to the occupations proposed by Domingo and Marcos10 and widely used in other studies, we recorded occupational categories as follows: classes I and II (self-employed or responsible for a business with 5 employees or fewer, responsible for a business with 6 employees or more, self-employed professional or professional employed by another business, holding an executive position, responsible for 5 subordinates or fewer, responsible for 6 subordinates or more); class III (intermediate-level executive, other employees who work in an office, other employees who work outside an office); class IV and V (skilled worker, unskilled manual worker, retired) or farmer (small-scale farming, large-scale farming).
The activities of daily life (ADL) dimension of the OARS-MFAQ questionnaire6 consists of 7 items on basic activities or activities related with self-care, and 7 items on instrumental activities or activities aimed at maintaining the person´s environment. Three summary variables were calculated: basic ADL, instrumental ADL and global functional capacity. For the basic and instrumental ADL variables we classified persons as «independent» if they were able to perform all activities in the variable without help, «needing help» if they required help for at least one activity, and «dependent» if they could not perform at least one activity. Within the global functional capacity variable we included completely independent persons in the «good capacity» category, persons who were dependent only for instrumental or basic ADL in the «slight impairment» category, and persons who were dependent in both types of activity in the «severe impairment» category.
The HAD6 questionnaire consists of 14 items divided into two subscales (anxiety and depression) with 7 items each, which are evaluated on a 4-point Likert scale from 0 to 3, in which 0 is the most favorable response and 3 the least favorable. The score for each subscale is calculated, and scores of 11 or higher are considered to indicate depressive disorder or anxiety. Scores of 7 or lower indicate that the individual should not be identified as a case, and scores between 8 and 10 are considered borderline findings.
The questionnaire was completed in the participants´ homes by an appropriately trained researcher. To reduce the number of persons who declined to participate, a personalized letter on health authority (INSALUD) letterhead was sent to each resident chosen as a participant, to request their help with the study and explain how the study was to be done. Participants were advised that the interviewers would contact them by telephone within the following 15 days to arrange a visit. The letter also provided a phone number the participants could call for additional information about the study. The
health centers in the areas where the study was to be conducted were also notified.
Individuals who could not be contacted by telephone or in person at their home address after three tries at different times of day were excluded and replaced with other participants.
Data analysis
The first part of the analysis consisted of a description of the study population (means and proportions with 95% confidence intervals). Data were stratified by age (younger than 80 years and 80 years or older) and by sex. For the qualitative analysis we calculated Pearson´s χ2 and the χ2 value for linear trends, or Fisher´s exact test when appropriate. Quantitative analyses were based on Student´s t test and nonparametric tests (Mann-Whitney U and Kruskal- Wallis K) as appropriate.
Results
Of the 2002 recipients of the letter requesting cooperation, 270 fulfilled the exclusion criteria, 453 declined to participate, 49 were away from their habitual residence and 319 could not be contacted after three tries. In all 911 questionnaires were completed, for a response rate of 66.8% (95% CI, 64.2%-69.3%) calculated as the number of questionnaires completed divided by the number of persons who fulfilled the inclusion criteria and who could be contacted.
We found no significant differences in sex or age distribution between the group of participants who completed the questionnaire and the group who did not complete the survey for reasons unrelated with any of the exclusion criteria.
Mean age of the 911 persons interviewed was 74.7 years (95% CI, 74.3%-75.1%). Women made up 59.7% (95% CI, 56.4%-62.9%) of the sample. About one-fourth of the women (26.3%, 95% CI, 22.7%-30.2%) and about one-fifth of the men (22.9%, 95% CI, 18.8%-27.6%) were 80 years old or more; the difference between sexes was not significant.
Table 1 shows the social and economic characteristics of the sample. Women had significantly lower levels of education, socioeconomic level and social and family support than men. Persons 80 years of age or more spent significantly more time alone than did those younger than 80 years.
Table 2 shows the physical and mental health characteristics of the sample. When perceived health status was classified in only two categories (very good or good vs fair, poor or very poor), men´s perceived health was better than women´s perceived health: 57.5% of all men perceived their health to be very good or good, vs 48.4% of the women (P=.007). We found no significant difference between persons older (53.4%) and younger than 80 years (51.7%). Women exercised significantly less and had significantly worse functional capacity and worse scores for anxiety and depression than men. Persons older than 80 years exercised less and, as expected, had a worse functional capacity than those younger than 80 years, although the differences between those older and younger than 80 years on the anxiety and depression subscales were not significant.
Among the reasons given by persons who did not rate their health status as good or very good, disease was the most frequent cause (73%; 95% CI, 69.7%-76%), followed by age (65%; 95% CI, 61.5%-68.3%), and much less frequently, toxic habits (28.6%; 95% CI, 25.4%-31.9%), economic causes (10.7%; 95% CI, 8.6%-13.1%) and work (10%; 95% CI, 8.1%-12.4%).
Mean scores on each of the dimensions of the NHP are shown in table 3 along with mean scores for men and women older and younger than 80 years. In general, sleep was the highest-scoring dimension and social isolation the lowest scoring dimension. Moreover, women had worse scores than men in all dimensions, and persons older than 80 years scored worse than those younger than 80 years on all dimension except emotional reactions.
Table 4 shows the percentages of persons who had some or considerable difficulty with the dimensions of the EuroQol instrument. The largest percentage of persons with problems appeared in the pain/discomfort dimension, and the self-care dimension showed the lowest percentage of persons with problems. In all dimensions except self-care, more women than men had problems. Persons older than 80 years had more problems than younger participants in the mobility, self-care and daily activities dimensions.
Mean score on the VAS of the EuroQol instrument was 66.6 (95% CI, 65.3%-68%); men scored lightly higher overall (68.8; 95% CI, 66.8%-70.9%) than women (65.1; 95% CI, 63.3%-66.9%; P=.004). Nearly half of the sample (45.7%; 95% CI, 42.4%-49%) had no problems with any of the dimensions of the EuroQol instrument. More men than women had no problems on any of the dimensions (53.7%; 95% CI, 48.4%-58.8%, vs 40.1%, 95% CI, 35.9%-44.3%; P<.001), and more persons younger than 80 years had no problems on any of the dimensions (47.5%; 95% CI, 43.7%-51.3%) as compared to persons older than 80 years (39.6%; 95% CI, 33.3%-46.4%; P=.039).
Discussion
Many studies have used self-perceived health status to investigate subjectivity in the concept of health in older persons; however, fewer studies have used quality of life questionnaires with this population. The present study reports the results obtained with three instruments to measure health status, and the results were comparable across instruments. However, the aim of the present report is to describe the results to lay the foundation for subsequent reports that will analyze the data in greater depth.
We describe a large (n=911), representative urban sample of persons older than 65 years residing in the Community of Madrid and reflecting the entire spectrum of sociodemographic characteristics, although the middle classes predominated. In comparison to other studies,8,11-14 the population we worked with had a higher level of education and a higher socioeconomic level, as shown by a comparison of the percentages of the samples that belonged to social classes I and II: 29.2% in the present study vs a maximum of 25% in earlier reports, in which social classes IV and V made up more than 50% of the sample.
Our response rate (66.8%) was somewhat lower than in similar reports such as the Envejecer en Leganés (Growing old in Leganés)15 (80%) and the ANCO studies16 (88%). The response rates in these surveys were enhanced with a media campaign involving the press, television and radio.
We noted considerable differences between men and women: the latter had a worse perceived health status and quality of life than men. Women are a disadvantaged collective, especially in the age group we studied, and their socioeconomic level is considerably lower than that of men. Moreover, women express more loneliness, greater functional impairment and worse affective disorders. Although fewer men than women reach great old age, those men that do reach the age of 80 years or older apparently do so under more favorable conditions than women.
There were also differences in perceived quality of life between persons younger and older than 80 years; however, we found no differences in perceived health status. The differences between the two age groups were smaller than those between the two sexes: no significant differences were noted in socioeconomic level or affective disorders, although persons older than 80 years felt lonelier and had a greater degree of functional impairment than those younger than 80 years. Older persons may assume these changes to be natural, and therefore may not consider them to diminish their health status.
Self-perceived health status, the VAS scale of the EuroQol instrument, and the profile provided by the five dimensions of the EuroQol instrument can be viewed as three instruments that aim to measure approximately the same thing: an individual´s subjective perception of his or her general health status.
Our findings for self-perceived health status are within the range of findings reported by others.8,12,17-24 As in our study, some earlier reports noted differences between men and women.8,22,25
Older persons have seldom been asked what they believe to be the main causes of their poor health. A search of the literature identified only one study, by Martín-Almendros et al.,26 in which a sample of the adult population was asked which two factors from a list of nine (smoking, diet, drinking, stress, physical activity, environment, weight, family support and heredity) they considered to have the greatest influence on their health. For persons older than 65 years the most important factors were diet (51.6%), smoking (39.1%) and drinking (31.1%). In the present study, smoking and drinking were the third most important cause of poor health, after illness and old age.
Mean score on the VAS (66.6) was between the values obtained in other national-level studies, and was higher than the mean score for persons older than 65 years in the region of Catalonia (60.6)27 but slightly lower than the mean score for the population of the province of Navarra (69.92).23,28 However, this figure was lower than the score for persons older than 60 years in the United Kingdom (76.9),29 where the VAS score decreased as age increased (80 for persons aged 60 to 69 years; 75 for persons aged 70 to 79 years; and 72 for persons 80 years old or more). In the present study, persons older than 80 years scored slightly lower (65.3) than persons younger than 80 years (67.1), but this difference was not significant.
Slightly less than half (45.7%) of the persons interviewed for this study reported a profile that reflected optimum health (11111); the percentage of women in this category (40.1%) was significantly lower than the percentage of men (53.7%). Badia et al.27 reported the 11111 profile in 67% of the respondents in the general population of Catalonia, and in the 2000 Navarra health survey23 70% of the general population reported the 11111 profile. A few years previously, Gaminde et al.28 found this profile in only 56.2% of the general population of Navarra, and the percentage was lower (38%) for persons older than 60 years, with no significant differences between men and women.
Our findings differ from those of other Spanish studies with regard to the percentage of persons who had problems with one or more dimensions of the EuroQol instrument (table 5). Badia et al.,27 in their study of the population of persons older than 65 years in Catalonia, and the 2000 Navarra health survey23, both found larger percentages of participants who had problems with mobility and smaller proportions with
anxiety/depression problems than in persons older than 65 years in the Community of Madrid. The differences between the percentage of men and women in Madrid with problems on one or more dimensions of the EuroQol instrument were greater than in the Navarra study.23 Kind et al.29 found that women aged 60 to 69 years in the United Kingdom had fewer problems than men of the same age group in all dimensions of the EuroQol instruments except anxiety/depression; however, for persons aged 70 years or older, women had more problems than men in all dimensions. These authors also found that the percentage of respondents with problems in all dimensions increased with age, except for the anxiety/depression dimension. In our study population the percentage of participants with problems increased with age for all dimensions except for anxiety/depression and pain/discomfort.
When we compared our results with those of the only Spanish study that centered on a similar population, we found that the scores on the NHP were similar to those reported by Alonso et al.30 in persons older than 70 years in Barcelona (energy, 23.3; pain, 21.8; physical mobility, 27.3; sleep, 32.8; emotional reaction, 25; social isolation, 11.5), whereas the differences between our results and those of a British31 and Italian study31 were larger. This may have been due to sociocultural differences between countries. In general, persons older than 65 years in the population we studied scored higher than their British counterparts except on the social isolation dimension, where scores in the two countries were similar, and the energy dimension, where scores were higher for British men and lower than Italian men. In the sleep dimension, men in Madrid scored higher, and in the pain and mobility dimensions, women in Madrid scored higher.
It would be interesting to determine, in a separate study, whether the results for these three measures of health status differ in any significant way between the general population and persons older than 65 years, and which potential differences are most useful to detect needs and deficiencies in health status.
Correspondence: Mercedes Azpiazu Garrido. C/ Guzmán el Bueno, 63, 4.º E. 28015 Madrid, Spain. E-mail: mazpiazug@nexo.es
This research was supported by Fondo de Investigaciones Sanitarias project 99/0248.
This study forms part of the research project titled «Quality of life and associated factors in persons older than 65 years» carried out by Mercedes Azpiazu Garrido for the degree of Ph.D., and was presented as a plenary lecture at the XX Reunión de la Sociedad Española de Geriatría y Gerontología in San Sebastian, Spain, in June 2001.
Manuscript received 11 March 2002.
Manuscript accepted for publication 17 July 2002.
