Relationship between pain and Quality of Life—Findings from the Swedish National Study on Aging and Care—Blekinge study

2.1 Sample

The Swedish National Study on Aging and Care (SNAC) is a national, longitudinal, multicentre study based on a representative randomised sample of age-stratified older adults in ten age cohorts (60, 66, 72, 78, 81, 84, 87, 90, 93, 96-years). Individuals were randomly selected from the national Population Register. The data collection was conducted at four research centres and approved by the Ethics Committees of Lund University and the Karolinska Institutet, Stockholm, Sweden. The structure and design of the SNAC project have previously been presented by Lagergren et al. [22].

Participants in the present study were recruited from the SNAC baseline sample at one of the four research centres, SNAC-Blekinge (B) in Karlskrona. A t otal of 2312 individuals were invited to participate, of whom 1402 (61%) agreed. There sponse ratevaried between 55 and 75% in the different age groups, with a higher rate in the younger groups. Thus a total of 910 subjects declined participation. Reasons for non-participation were lack of willingness (83%), feeling too ill (10%) and impossible to contact (7%). No statistical difference was found in terms of gender among the participants and dropouts. Data collection was carried out during the period 2001–2003. This sample of 585 men and 817 women is similar to the other rural and urban samples of the overall SNAC study in terms of age, sex, functional ability and perceived QoL. In 2001, the municipality of Karlskrona, a medium sized town with rural surroundings, had 60,596 inhabitants, of whom 30,389 were women and 30,207 men. The population > 60 years was 14,627 (24.1%), of whom 8242 (56.3%) were women and 6385 (43.7%) men.

2.2 Procedure

Potential participants were invited by letter to take part in the study. Medical examination and testing by research personnel was conducted in two sessions, each lasting about 3h. Physicians and nurses were membersof the research team and collected data from the medical examination, structured interview and questionnaire. Access to medical records and informed consent were obtained from each patient and the questionnaire was completed in the period between the two sessions. Those who agreed to participate but were unable to travel to the research centre were investigated in their home or nursing home.

2.3 Measures

Information on health related Quality of Life (HRQOL) was obtained by using the health survey short form (SF-12) as a dependent variable. The SF-12 instrument has been validated and provides a score estimate of an individual’s health in eight dimensions: Physical functioning, Physical activity, Pain, General health, Vitality, Social functioning, Emotional capacity and Mental health [23,24]. The response alternatives to the twelve questions were ‘Yes’ or ‘No’ or in the form of 3, 5 or 6 statements. The scores were processed in accordance with the Swedish Manual and Interpretation Guide [23]. The physical component scores (PCS-12) and mental component scores (MCS-12) are obtained by means of the standard scoring algorithm providedby the designers of the instrument [25]. As the concept of pain partly overlaps the PCS-12 in the SF-12, the analyses were conducted using the MCS-12. The holistic approach and health perspective meant that high QoL was defined as the upper 25th percentile, corresponding to a cut-off point of 60.4. Participants in the lower 75th percentile were thus defined as having low QoL.

The independent variable pain was self-reported and based on the question: Have you had ache/pain during the last 4 weeks? The response alternatives were ‘Yes’ and ‘No’.

Data on age, gender and living alone were collected from the questionnaire.

Sleep disturbances were addressed by the question: Do you suffer from insomnia? The response alternatives were ‘Yes’ and ‘No’.

Co-morbidity was measured using the Johns Hopkins Adjusted Clinical Groups (ACG) Case Mix System 6.0. Special algorithms based on various diagnoses were used and validated [26,27]. The five criteria employed were: (1) likely persistence of the morbidity conditions in question, (2) their severity, (3) their aetiology, (4) their diagnostic certainty and (5) the person’s need for special care. The individuals were then assigned to some of the 82 ACG groups with the same degree of co-morbidity but without measuring the individual’s functional capacity. The diagnoses were obtained from electronic patient records for a period of up to two years prior to the SNAC study (n=1379). The 82 ACG groups were then merged into six (0–5) groups, 0 indicating no need and 5 a great need of health care. The mean co-morbidity was 1.73 (s.d. 1.40).

SOC was measured by a translated short form of 13 questions from the original 29-question instrument, using the scoring algorithm provided by the designer [28]. There are 7 response alternatives and a score range of 1–100. As low, normal and high SOC is not defined, the mean score (72.23) was used. Thus scores below the mean were classified as low SOC, because it was considered to be normally distributed.

The economic status was based on whether or not the person had savings. A question from the Swedish statistics survey on income and living conditions was used: Could you, if necessary, raise 14,000 SEK (about 2000 USD) within a week for unexpected expenses? The response alternatives were ‘Yes’ and ‘No’ [29].

Educational level was measured by the question: Did you finish secondary school? The response alternatives were ‘Yes’ and ‘No.

2.4 Statistical analysis

Descriptive statistics were used to characterise the data as well as the association between the independent variables and the dependent variable, QoL. Spearman’s chi-square test was applied to control for differences between the 25% with the highest scores and those reporting scores in the MCS 75th. Multivariate logistic (forward) regression analyses were conducted separately for each sex to provide odds ratios (ORs) for the effect of the independent variables on the dependent variable and to study whether the relationship between pain and QoL changed when covariates were entered.

All analyses were carried out using the SPSS program, version 17.0. The probability values (p-value) revealed statistically significant results. The level of significance was set at 0.05. Missing data are indicated by (n) in all analyses. If a single question was missed the subject was excluded and reported as missing data.

4.1 Limitations and strength

Some limitations should be mentioned. This cross-sectional study included older people from an urban area (a medium-sized town) with rural surroundings, so perhaps the results cannot be generalised to rural areas and big cities. However, the study population resembles the other rural and urban samples of the overall SNAC study in terms of age, sex, functional ability and perceived QoL.

In this study, the subjective experience of pain was self-reported and pertained to the four preceding weeks. The results do not reveal whether the reported pain was persistent or acute nor its intensity. As pain is an unpleasant sensory and emotional experience associated with or described in terms of actual or potential tissue damage, pain assessment may reveal an imbalance in the body, psyche or both. This cross-sectional study does not deal with the consequences in relation to activity or suggest pain-treatment and whether the latter could have an influence on the differences identified. However, the reported prevalence of pain among older adults and the gender difference where more women reported pain are consistent with findings from a study in the UK with similar age cohorts [41].

In the Swedish general population, the Mental component scores (MCS) of the health survey short form, SF-12, for the upper 25th percentile in the age groups 60–80 years and over were distributed between 57.9 and 59.3 with a standard deviation from 10.4 to 11.7 [33]. In our study, the age groups were divided into shorter intervals and extended to 96 years. As younger cohorts score lower on the MCS, the mean MCS-score in this study is considered representative. The strength of the study is the number of older individuals in the oldest-old cohorts. The material is consistent with similar studies from other parts of Europe in which several factors apart from health were important for QoL[3].A British study of people aged 65 years and over living in their own homes revealed that social contacts are equally valuable as health [21]. Similar findings have been reported in a Canadian study of participants aged 60 years and over [39].

4.2 Conclusions

Pain has a strong relationship with low QoL among elderly women. Insomnia is associated with low QoL among men who suffer less from pain. Thus the main result is a striking gender difference: Elderly women suffer from pain, elderly men suffer from insomnia.

4.3 Implications

The aim of this study was to determine whether biological/psychological and social factors influence the relationship between pain and QoL among older adults. Extending life requires biological, psychological and social well-being in combination with an individual’s will to live. There is an obvious necessity to satisfy the needs of the older section of the population and improve their perceived QoL. Our findings suggest that it is important for all healthcare professionals to take account of the fact that aspects other than pain and insomnia, for example co-morbidities, need to be highlighted during consultations.

The influence of biological, social, psychological and existential factors on the relationship between pain and QoL was studied in women and men aged 60 years and over from a gender perspective. There are gender-based differences in the population that are stronger among older adults with pain [4,5]. Our data indicate that pain, age and female sex in addition to social and psychological factors have an independent, negative effect on QoL. This suggests that each factor should be assessed, although in view of the ORs, focus should be placed on pain and co-morbidity among women and insomnia in men.

The mortality rate of older people reporting dissatisfaction with their health has been found to be double that of people who do not report dissatisfaction [19]. The differences in mean age between the groups with and without pain indicate the need for future studies to investigate whether or not pain contributes to morbidity/mortality in the general population of older adults.