Self-perceived health in functionally independent older people: associated factors

Population ageing is a worldwide trend in recent decades. In Spain, there were an estimated 7.9 million people (17 % of the total population) over the age of 65 in 2014 and this figure is expected to reach 15.4 million (38 % of the total population) by 2064 [1]. Active and healthy ageing (AHA) is a multi-factorial process that has become a health policy priority for local, national and international health authorities [2] in order to reduce the growing incidence of dependence. Self-perceived health (SPH) is included in one of the six group of determinants of the AHA model proposed by the World Health Organization (WHO) [3].

SPH is a complex measure that “represents a summary statement about the way in which numerous aspects of health, both subjective and objective, are combined within the perceptual framework of the individual respondent” [4]. It has been widely used, and is considered a valid and reliable indicator [5‐7] of overall health status, a predictor of mortality and of health services use [8]. It is usually evaluated with a single item, asking individuals to rate their overall health on a scale from poor to excellent. Several alternative phrasings of the SPH question exist, but all have similar response patterns [9].

SPH in older adults has been found to be associated to sociodemographic characteristics (namely, sex, age, education, and income), chronic diseases and functional status [10‐12]. Additional factors linked to SPH are social relationships [11‐13] and neighborhood environment [14, 15]. However, the effect magnitude of these factors varies depending on the study design, population and cultural context [16], and not all studies include all the above mentioned variables.

Functional status, in particular, is widely recognised to be a powerful determinant of SPH in older adults [10]. The assessment of functional status is a key element in healthcare. The detection of functionality loss can guide the implementation of health interventions aimed to avoid or delay the onset of dependence. Given the relevance of functional status as a factor of SPH, when it is studied at the same time with other related factors, the latter may become secondary with their role remaining undervalued when functional status is considered. Possibly, this is why less is known about how older people with good functional status perceive their health and the actual role of other factors related to SPH in this population. The assessment of factors beyond functional status may provide relevant information for the design and implementation of health promotion strategies.

This work is part of a larger study focused on understanding aspects of health and living conditions of community-dwelling elderly people, from a comprehensive perspective. The objectives of this study are to describe SPH in a functionally independent community-dwelling older population and to identify the main factors associated with SPH in this population group.

Of the 800 individuals selected, 125 were excluded from the analysis for cognitive impairment and 41 for BADL dependence. Those excluded were mostly women (67 %), were older (p < 0.0001) and had poorer health, with a higher percentage having ≥3 chronic conditions (p < 0.0001) and taking ≥3 prescription drugs (p < 0.0001) compared to the included individuals. Thus, the final sample included in this study consisted of 634 independent community-dwelling older adults. The mean age of the individuals included was 74.8 (SD 6.7) years and more than half (55 %) of them were women. A total of 126 subjects were accompanied during the interview and 12 (2 %) of the participants answered the battery of questions with some help by the accompanying person. The five SPH categories were replied as follows: excellent (n = 38 participants), very good (n = 147), good (n = 333), fair (n = 105) and poor (n = 11). Thus, about 18 % of the respondents rated their SPH as poor.

From an individual perspective, those with a higher level of education were more likely to report a good SPH (p = 0.035), as were those with higher monthly incomes (p = 0.053) (Table 2). In addition, those presenting ≥3 chronic diseases and taking ≥3 drugs daily (p < 0.0001), having depressive symptoms (p < 0.0001), sensory impairment (p = 0.0001), bad nutrition pattern (p = 0.0001) and failing to get a good night’s sleep (p = 0.002) were more likely to report a poor SPH.

From a social point of view, the percentage of those reporting good health was significantly higher among those who engaged in cognitively stimulating activities (p < 0.0001), social leisure activities (p = 0.001) or group social activities (p < 0.0001), with high social trust (p = 0.017) and satisfactory self-perceived social life (p < 0.0001) than it was for those who did not engage in the social activities considered, had low social trust or unsatisfactory social life.

Regarding contextual variables, individuals without obstacles inside their home (p = 0.031) and with a supermarket within walking distance (p = 0.003) were more likely to perceive their health as good.

No statistically significant differences were observed by age or sex, living arrangements, history of falls, social support, physical activity, smoking status, housing condition, active leisure activities or adequacy of community services (Table 2).

At the multivariate analysis stage, polypharmacy, sensory impairment, bad sleep quality, bad nutrition pattern and not engaging in certain types of leisure activities were identified as the set of factors that best explained SPH. Specifically, according to this model (Table 3), respondents with polypharmacy were more likely to report poor health (OR: 5.76, 95 % CI: 3.60–9.18). Further, respondents with at least one sensory impairment, those with bad sleep quality and those with bad nutrition pattern were around twice likely to rate their SPH as poor (OR: 1.87, 95 % CI: 1.15–3.04; OR: 1.82, 95 % CI: 1.02–3.28; and OR: 2.37, 95 % CI: 1.08–5.21, respectively). Finally, individuals who did not engage in cognitively stimulating activities or group social activities were also more likely to rate their SPH as poor (OR: 4.08, 95 % CI: 1.64–10.20 and OR: 2.62, 95 % CI: 1.63–4.23, respectively). Collinearity diagnostics did not reveal any problems with the included variables. The AUC of this model was 0.810, suggesting good discriminatory capacity [37].

This study aimed to assess the role of various individual, social and contextual factors in the assessment of SPH in independent older people without cognitive impairment. For this purpose, individuals were considered independent if they had good cognitive status and BADL function. In past research [11], people with cognitive impairment have also been excluded, given their level of deterioration could interfere with their ability to provide valid answers. Inability to perform daily life activities was also an exclusion criteria applied in a previous older people study [11]. Therefore, the approach followed in the present study can provide further research information on which other factors besides autonomy are associated with SPH.

The multivariate model that best explained poor SPH in our data, after adjusting for age and sex, included the following variables: polypharmacy, having sensory impairment, bad sleep quality, bad nutrition pattern and not engaging in cognitively stimulating activities or group social activities. These results are consistent with the complex definition of AHA [3, 38, 39]. The presented model includes factors related to individual health status, like polypharmacy, clearly related to the presence of chronic diseases. Also three elements rarely relevant in younger populations, namely, sensory problems, sleep quality and nutrition patterns [38‐40]. Additionally, our model includes two variables related to enjoyable activities: cognitively stimulating activities and group social activities. As far as cognitively stimulating activities are concerned, reading may be associated with educational attainment, but not listening to the radio. This finding may indicate the importance of maintaining mental vitality into late adulthood as a component of healthy ageing. A WHO report of active ageing [39] encourages older people to keep learning to maintain cognitive capacities. Other relevant aspects associated with a satisfactory cognitive ageing are the absence of stress in older people, found to be related to a lower deleterious impact in the brain [41], and brain reserve, where higher mental activities have been found to be associated with less cognitive decline [42]. Group social activities are a multicomponent variable that reflects a good level of interest in socialising and maintaining a social network. Both cognitively stimulating activities and group social activities are factors that may contribute to improving quality of life of older people, being the latter also important to enhancing social connectedness [43]. Previous works have also examined the relation between social activities, social participation or social integration and SPH [11, 13]. Although different social activities have been studied, such as, participation in recreational activities at a public hall or community association for the aged [11], in a hobby association [11], in educational and cultural activities [13], in sports or physical activities with other people, in family/friendship activities outside the household [13], or going to a social centre for the elderly [12]; in all cases, participants had better SPH than non-participants.

Age, sex, education, income, chronic diseases, functional status and mental health have also been found to be linked to SPH in previous studies [10‐12]. In our data of functionally independent subjects, all other factors but sex were associated with SPH in the univariate analysis stage. The lack of association with sex is completely unexpected and is probably due to the selection criteria applied, in particular, the exclusion of dependent individuals, with dependence being more common among women. However, sex was not a predictor of SPH in a previous study of older Thai people [10].

On the other hand, our finding that good night sleep is related to a better SPH is consistent with a cohort study of non-institutionalized Spanish adults aged 60 years or more where extreme sleep durations (≤5 or ≥ 10 h) were associated with a poorer health-related quality of life [44]. Sleep is a vital physiological process and the mechanisms implicated in its duration and quality are complex and multidimensional [45]. Several changes are seen in older people, as aged individuals, are more prone to wake up at earlier times or experienced an involuntary reduction of their total sleep duration [46]. Finally, disturbances in the circadian rhythm have been associated with a worse health [47]. All the above reveals the importance of maintaining adequate sleep patterns in this population group.

Sensory-related abilities are also found to be related to SPH. This is in concordance with a study of non-disabled older people living alone [11], where problems in visual, hearing and chewing abilities were associated with poor SPH. One of the challenges of an ageing population according to WHO are precisely vision and hearing losses as well as oral health [39]. The WHO urges countries to develop policies and programmes to prevent visual and hearing impairment and promote good oral health habits [39]. Similarly, the development of guidelines promoting an adequate nutrition is a need to solve eating problems, understanding as such both under nutrition and excess energy intakes [39].

Another difference with other published studies [14, 15] is the lack of associations with neighbourhood characteristics (adequate community services and proximity to a number of facilities), either in the univariate (except for proximity to supermarket) or in the multivariate analysis. One possible explanation is that people with poor SPH may have a different perception of their neighbourhood environment than those in good health [48] and may not even perceive these problems as they go out less frequently. It is also likely that, if individuals live in the same neighbourhood for a long time, they get used to the environment. Another plausible justification may be that the measure created to analyse this factor was not suitable for use in older people. Finally, it is also important to consider that the questions used to assess community services were based on WHO age friendly city criteria [35] and were not part of an specific or validated questionnaire. Therefore, further research will be conducted exploring this issue and, perhaps, new tools should be developed to explore elderly peoples’ perception of their communities.

On the other hand, there are limitations that should be recognised. The study design was cross-sectional, and hence, it is not possible to establish cause-effect relationships. Nonetheless, as stated by Giron et al. [19], this type of research is necessary to help us understand factors that determine or alter health. Moreover, the battery of questions developed is extensive (145 items that were responded in approximately 60 min) and may explain why individuals who agreed to participate tended to have a relatively good functional status. Another limitation of this study is that the data were collected via face-to-face interviews by 20 interviewers. So, although interviewers were trained, there is always a possibility that the data collection technique varied, as was explained in the study of Tajvar et al. [49]. Nevertheless, face-to-face interviews are the recommended way to collect information in the case of long questionnaires, and generally obtain higher response rates than postal or telephone interviews [50]. Even though information of chronic conditions was self-reported by the individuals and thus may affect the accuracy of information, it is an approach used in similar previous studies [10, 11]. It is also important to discuss about the representativeness of the sample. The original sample (800 individuals) was representative of the Gipuzkoa elderly population according to sex and age. The exclusion of non-independent and cognitively impaired individuals did not cause major alterations to the percentage of the considered variables, as in the finally included sample (634 individuals) all categories were between 1 and 5 % of the original sample. Finally, missing data was not a particular problem in the current study but for the income level variable, which 22 % of the sample did not reply. It is likely that full information on this very variable would have allowed for further explorations.

In conclusion, several different factors related to health and social variables account for the SPH of older people besides autonomy. Among them, the study found that drugs taken per day, sensory-related abilities, sleep quality, nutrition and leisure activities were the most strongly associated with SPH in a functionally independent elder population. In this sense, our results confirm the multidimensional nature of SPH and are consistent with those of other studies. Our work highlights the need for a thorough assessment of factors associated with SPH in older adults, this being essential for developing health-related programmes that promote active and healthy ageing.