Determinants of ADL and IADL disability in older adults in southeastern Poland

This cross-sectional study was carried out by researchers at the University of Rzeszow among a randomly selected representative population of people aged 60 and older living in southeastern Poland (region of the Podkarpackie Voivodeship). There are 350,000 people aged 60 and over in Podkarpackie Voivodeship [18]. The inclusion criteria for participation in the study group were an age of 60 or older, normal cognitive state (Abbreviated Mental Test Score (AMTS) > 6 points), and provision of informed consent for participation in the study.

The calculation of the sample size was based on the following assumptions: a 95% (0.95) confidence level and a fraction size of 0.5 with a maximum estimation error of 3%. A flow chart shows the participant selection and drop-out process (Fig. 1).

The study was carried out by appropriately prepared and trained interviewers who conducted direct, pen-and-paper interviews at the participants’ places of residence. Households were replaced after 3 unsuccessful interviewer visits of the interviewer due to the resident not being at home or an older person’s refusal to participate in the survey of an older person, death or inability to participate. Eventually, 2207 complete interviews were included in the analysis.

To assess the participants’ cognitive states, an abbreviated version of the AMTS questionnaire [19] was used. Since the interview was conducted directly with older persons, this questionnaire served as a screening criterion (AMTS > 6 points) to ensure the collection of reliable answers.

Information on socioeconomic variables such as age, sex, place of residence, marital status, education, and income was collected by means of an Socioeconomic Status Index.

The Katz Index of Independence in Activities of Daily Living [20] was used to assess basic activities in everyday life, whereas the Lawton Instrumental Activities of Daily Living Scale was used to assess complex daily activities [21].

To examine various factors affecting the functioning of older people in Poland, codes from the WHOICF checklist were selected [10]. A review of the literature and a selection (mapping) of relevant items/questions for the selected categories (category) in accordance with WHO recommendations were carried out [22]. The selected codes were assigned to questions from the standardized questionnaires (e.g., WHODAS 2.0, WHOQOL). Questions regarding the physical activity of older adults were prepared according to the recommendations of the European Network for Action on Aging and Physical Activity [23]. Based on pilot study (that preceded the main study), the prepared tools were checked. The repeatability and correlations among the questionnaires included in the mapping were assessed. To reduce the number of questions that would be asked in the main study, the questions that significantly differentiated the population depending on the functional level in the pilot study were selected.

In the main study, the following information concerning physical health was collected: height and body mass (including BMI calculations and ICF b530 Weight maintenance functions), pain occurrence (ICF b280 Sensation of pain) and the number of chronic diseases (a quantitative variable related to health conditions was selected as a general reference to the ICD-10 classification complementary to the ICF and a popular indicator of multimorbidity and the maintenance of good health in older people) [24]. There were questions about physical activity, including the amount of time devoted to any physical activity during the week (at least moderate physical activity, defined as activity causing at least a slight shortness of breath, sweating, and fatigue), as well as about planned physical exercises performed to improve strength and endurance (ICF d570 Looking after one’s health, including the maintenance of appropriate physical activity) [25]. Furthermore, the researchers also collected information on social activity and participation, including participation in groups or social organizations (ICF d910 Community life), and maintaining good relations with relatives (ICF d760 Family relationships) and acquaintances (ICF d720 Complex interpersonal interactions). The participants were also asked about their living environments, including the existence of barriers and obstacles (including architectural, communication, social and other barriers) (a comprehensive selection of the environmental factors included in the Short List of Environment on the WHO ICF checklist) and their housing conditions (ICF e155 Design, construction and building products and technology of buildings for private use). For the physical health analysis, responses were given in the form of quantitative variables (BMI, the VAS pain scale, and the number of diseases).

To use the collected information in a logistic regression model, the obtained information was coded as dichotomous data (yes/no), except for housing conditions, which had three options for further analysis. The cut-off point for the assessment of physical activity was established based on the WHO recommendations [25].

For the analysis, the participants were divided into persons without any difficulties and persons reporting at least one ADL limitation and at least one IADL limitation. Dichotomous variables were created, with a value of 1 if the participant showed a limitation in one or more ADLs (1+ ADLs) and IADLs (1+ IADLs) and a value of 0 if the participant did not show any limitations. These cut-offs are provided by the SHARE project [7]. The following sections present only the results of people with at least one difficulty in the ADL and IADL scales. The data were analyzed using Statistica version 13.1. Demographic data are presented as descriptive statistics. Two logistic regression models were used to identify factors related to ADLs and IADLs. The level of significance was set at p <  0.05. The chi-square test (in the case of qualitative variables) and the Mann-Whitney test (in the case of quantitative variables) were used for the initial analysis of the relationship between the individual demographic variables and ADLs and IADLs. The normal distribution of the quantitative variables was verified using the Shapiro-Wilk test. Logistic regression models were used to identify sets of factors that had statistically significant effects on the occurrence of at least one ADL limitation and at least one IADL limitation in the entire study group. The model parameters were estimated by means of stepwise regression (forward selection). The quality of the model estimation was tested using the Hosmer-Lemeshow test and pseudo R² values. For the variables that were included in the above models, their clustered influence on the increase in the odds ratio for the occurrence of an ADL or IADL limitation was also examined when the factors occurred in pairs for people who did not report these variables. The following dichotomous variables were used for this analysis: age, pain level and the number of diseases. In order to obtain the dichotomous variable, the WHO age division was used, i.e., the 60–74 years (early old age) and 75 and more (late old age and very old age) group [26], while the number of diseases was divided according to the definition of multimorbidity, i.e., 0–1 diseases and 2 and more diseases [27]. The pain level variable as measured by the VAS scale was divided into 0–2 points and 3 and more points. Three points was adopted as the cut-off point because this level significantly differentiated the respondents with at least one ADL limitation and one IADL limitation. The significance level was set at p <  0.05.

In total, the study included 2207 people aged 60 and older, including 1325 women and 882 men. The average age of the participants was 72.12 (SD = 7.77). The overwhelming majority of participants had primary or vocational education (63.43%) and were in a relationship (62.62%). Among the majority of participants who answered the income question, the average income per household member was PLN 2000 or less per month. On average, each participant had an average of 4.94 (SD = 3.55) chronic diseases, while their average pain level as measured by the VAS scale (0–10) was 3.58 points (SD = 2.91). Most of the participants did not perform at least a moderate level of physical activity for 150 min/week (71.05%) and did not perform physical exercises aimed at strengthening the muscles and improving physical fitness (77.07%). Moreover, most of the participants did not belong to a social group (77.75%). On the other hand, the majority of people maintained social contacts (62.62%) and claimed that they had good relations with their relatives (63.25%). Almost half of the surveyed group (48.71%) stated that their environments had some barriers and obstacles (communication, social, or architectural), and only 40.82% of participants were satisfied with their living conditions (Table 1).

The research results showed that 35.75% of participants reported at least one problem with IADLs. Most problems with IADLs were found to be related to walking farther than the normal walking distance/moving within the community (27.46%). At least one problem with ADLs was reported by 17.13% of people. Most often, the participants had problems getting out of bed and moving around (17.54%). The prevalence of at least one problem with ADLs and IADLs increased gradually in the older age groups (Table 1).

Problems with ADLs and IADLs occurred significantly more frequently for the following individuals: older people, women, lonely people, those with lower incomes, those with lower physical activity levels, those who did not engage in physical exercise, those who did not maintain social contacts, those with worse relations with their relatives, those living in an environment with barriers and obstacles, those with more chronic diseases and those with higher levels of pain (Table 1).

The most common limitations in ADL in the study group were bathing and showering (8.38%) and dressing (6.52%). In IADL, moving within the community posed the most problems (27.46%). The percentage of people experiencing at least one problem with ADL and IADL increased in older age groups (Table 2).

The logistic regression models included variables that significantly differentiated the studied population in terms of the occurrence of at least one problem with ADLs and IADLs. Due to the large number of missing answers (685 missing items), a variable for the income of the surveyed population was not included in the models.

The model of the effect of ADL factors was well adjusted to the data, as indicated by the results of the Hosmer-Lemeshow test (χ²_HL = 4.311, p = 0.828), and the pseudo R² value was equal to 0.8559, indicating that the model correctly classified 85.59% of the cases. An important factor related to the occurrence of ADL limitations was the presence of barriers and obstacles in the respondent’s environment, including architectural, communication, social and other barriers. The presence of barriers and obstacles increased the risk of occurrence of at least one ADL limitation by almost four times (exactly 3.74 times) compared to that of people who did not report such barriers in their environment. Another important factor was engaging in daily physical activity that caused shortness of breath, sweating, and slight fatigue (e.g., doing housework, gardening, brisk walking, or participating in sports) for at least 30 min a day for a total of at least 150 min a week. People who did not engage in such activity had almost two and a half times greater odds (exactly 2.33 times) of having at least one ADL limitation. Moreover, maintaining social contacts was also a significant factor. People who did not maintain social contacts were twice as likely (exactly 2.04 times) to have at least one ADL limitation. Another important factor influencing ADL disability was maintaining good relations with relatives. People who did not maintain good relations with their family members were 1.5 times more likely to experience ADL disability than people with good relations with their family. Other important factors affecting ADL limitations were pain, age and the number of chronic diseases. With each successive level on the VAS scale, the odds of disability increased by 27%. In addition, with each subsequent year of participants’ age, the odds of disability increased by 8%, and with each subsequent chronic disorder, the odds increased by 7%.

The model of the effect of IADL factors was well suited to the data, as indicated by the results of the Hosmer-Lemeshow test (χ²_HL = 11,473, p value = 0.176) as well as a pseudo R² value of 0.7748, indicating that the model correctly classified 77.48% of the cases. As in the ADL model, an important factor associated with the presence of IADL limitations was the existence of barriers and obstacles in the participant’s environment. The presence of barriers and obstacles increased the risk of at least one IADL limitation by three times (exactly 2.98 times) compared to that in people who did not have such barriers in their environment. Another important factor was participating in daily physical activity for a total of at least 150 min a week. Persons who did not perform such activity were almost one and a half times (exactly 1.36 times) more likely to live with at least one IADL limitation. Similarly, people who did not maintain social contacts had almost one and a half times (exactly 1.35) greater odds of having at least one IADL limitation. Other dominant factors affecting the occurrence of IADL limitations were pain, age and the number of chronic diseases. With each subsequent level of the VAS scale, the odds of disability increased by 27%. In addition, with each subsequent year of participants’ age, the odds of disability increased by 10%, and with each subsequent chronic illness, the odds increased by 4% (Table 3).

An assessment of the impact of pairs of factors that significantly differentiated the examined population was conducted in terms of the occurrence of at least one problem with ADLs or IADLs and they entered in to two logistic regression models presented above.

The incidence of disability varied widely depending on the specific pairs of factors, but the incidence significantly increased in each pair. The highest odds ratio of at least one ADL limitation was observed for the combination of barriers in the participant’s environment with multimorbidity (odds ratio (OR) 74.07), pain on the VAS scale ≥3 points (OR 50.93), a lack of at least 150 min of physical activity in a week (OR 44.51) or older age (OR 42.40). A very high odds ratio was also found for the combination of multimorbidity and a lack of social contacts (OR 47.50) (Table 4).

With regard to IADL disability, the highest odds ratio of at least one limitation was observed for the combination of pain on the VAS scale ≥3 points with older age (OR 19.47), a lack of social contacts (OR 16.33), or a lack of good relations with relatives (16.28). A high odds ratio of disability also was observed for the combination of barriers in the participant’s environment with older age (OR 18.57), pain (OR 18.31) or multimorbidity (OR 17.68) (Table 4).

This study has some limitations. First, the cross-sectional nature of this study does not allow the researchers to make strict cause effect interpretations of the associations between ADL and IADL disability and its determinants. A longitudinal study is recommended to establish such associations. Second, the population of older people under institutional care was excluded from the study, and therefore, the prevalence of disability may have been completely underestimated.