Self-rated Changes of Health Status during Stay-at-home Orders among Older Adults Using the Long-term Care Insurance System of Japan: A Cross-sectional Study

Abstract

Objectives: To reveal self-rated changes of health status during stay-at-home orders among older adults and to verify whether decrease in frequency of going outdoors during these orders was related to self-rated changes in health status.

Method: A self-completed questionnaire for older adults was provided in 2 dayservice facilities and a nursing station. We operationally defined health status with 4 domains (motor function, oral and swallowing function, depression, and social networks) and designed the questionnaire to determine self-rated changes in health status using factor analysis. After factor analysis, regression analyses were conducted.

Dependent variable was each factor score (self-rated changes of health status), and independent variable was decrease in frequency of going outdoors.

Results: Approximately 80% of participants answered that their health status had “worsened” in motor function (75.0%-87.2%). Moreover, more than 70% of participants answered “worsened” in “Feeling energy” and “Getting together and speaking with friends” (72.3% and 75.7%, respectively). Regression analyses demonstrated that, after adjusting for covariates, the decrease in frequency of going outdoors was related to self-rated changes of motor function and friend network.

Conclusion: During stay-at-home orders, older adults felt deterioration in their motor function, in feeling energy, and in their friend network, especially people who had decreased their frequency of going outdoors felt more deterioration in their motor function and in their friend network.

A new type of coronavirus disease called COVID-19 have been widespread around the world. The World Health Organization declared a state of emergency on January 31 2020. Since then, in order to prevent the spread of COVID-19, many countries around the world legislated or recommended stay-at-home orders, or “lockdowns,” which have resulted in individual and collective restrictions on participating in outdoor activities.

At the end of February 2020, the Japanese government requested that people reduce their attendance at sports and cultural events. On March 26, in 4 prefectures, including Tokyo, governors requested that residents stay at home and avoid non-essential trips outdoors. On April 7, the Japanese government declared a state of emergency and requested residents in 7 prefectures, including Tokyo, avoid unnecessary or non-urgent visits outdoors until the state of emergency was lifted (this was called stay-at-home orders, or “gaisyutsu-jisyuku-yosei” in Japanese.) On May 4, the stay-at-home declaration was extended until May 25, meaning that residents must remain at home for approximately 2 months.

Research has shown that going outdoors is related to various indicators of health status among older adults. Cross-sectional studies of older adults revealed that those who went outdoors infrequently was more likely to have mobility impairments^1,2), limitation in activities of daily living (ADLs)^1-3), limitation in instrumental activities of daily living(IADLs)^1-3), cognitive decline^1,2,4), depressive mood^1,2,5), less social participation¹⁾, low oral function¹⁾, low meal intake⁶⁾, low health rerated quality of life⁷⁾, and so on. Moreover, several prospective cohort studies of older adults showed that frequency of going outdoors was related to differences in mortality⁵⁾, mobility⁸⁾, ADLs⁹⁾, IADLs^5,8,10), and self-rated health⁹⁾. Thus, going outdoors is an important factor for health status.

Stay-at-home orders, namely social or collective restrictions on outdoor activities have been unprecedented in recent years. A Scoping Review including 41 articles showed physical activities (PA) significantly decreased regardless of the population studied¹¹⁾, and in Japan, a cross-sectional online survey among older adults also clarified PA during stay-at-home orders significantly decreased compared to PA before stay-at-home orders¹²⁾. Moreover, several studies demonstrated decrease PA during COVID-19 pandemic was related to higher level of depression^13,14).

In the same way as PA, stay-at-home orders might decrease frequency of going outdoors among older adults, which might lead to a decline in their health status. Therefore, it was necessary to study how health status might have changed during stay-at-home orders and to verify whether decrease in frequency of going outdoors due to the stay-at-home orders was related to health status, as previous studies^1-10) conducted in non-pandemic conditions showed that relationship.

Health status is measured both subjectively (self-rated health) and objectively; self-rated health has a predictive value for mortality equivalent to objectively measured health¹⁵⁾. Although, in general, objectively measured health is more reliable, to better understand the needs of older adults, it is also important to focus on their own subjective perspectives about changes in their health status.

The objectives of this study were to reveal self-rated change in health status among older adults during stay-at-home orders and to verify whether decrease in frequency of going outdoors during stay-at-home orders were related to self-rated change in health status.

Method

Procedure and participants

This cross-sectional study was done with questionnaires completed by older adults using the long-term care insurance system of Japan. The survey was conducted in 2 day-service facilities and a nursing station in Setagaya Ward, Tokyo, from May 11 to May 31, 2020. We defined stay-at-home as staying at home because of social or collective restrictions on outdoor activities that is not attributed to personal factors. Moreover, during stay-at-home and before stay-at-home were defined as periods in April and May 2020 and in January and February 2020, respectively.

During the survey period, 262 people were registered in the above-listed places. Exclusion criteria was people with diagnosis of dementia or cognitive deficits who had no family to represent them in answering questionnaire. We searched those with diagnosis of dementia or cognitive deficits in medial or care record and ask medical or care staffs if they have family to represent them in answering. Consequently, 11 people were excluded, and we sent questionnaire to 251 people. Informed consent was obtained from all participants or their family members. We confirmed provisions of the Declaration of Helsinki, and this study was approved by the ethical committee in the Arakawa campus of Tokyo Metropolitan University (approval number: 20016).

Measurements

Participant characteristics

We collected participant characteristics of age, sex, and living alone in the questionnaire. The level of care in the long-term care insurance system of Japan¹⁶⁾, and disease information were collected from medical or care records. We investigated if each participant had the following diseases or conditions: orthopedic disease, stroke, neurodegenerative diseases (e.g. Parkinson's diseases, not including dementia), spinal cord injury, peripheral nerve disease, heart disease, pulmonary disease, kidney disease, digestive disease, psychiatric diseases (except for dementia), cancer, dementia, diabetes mellitus, and other diseases. We defined sum of disease domain as degree of multi-morbidity (e.g., if a participant had knee arthritis and stroke, the degree of multi-morbidity was defined as 2).

Going outdoors

We defined going outdoors as follows, with minor modification to Fujita's definition¹⁾: going outdoors was going out from the house with or without a caregiver and includes shopping, taking a walk, going to a doctor or day service facility, and going for leisure activities; it excludes outdoor gardening or taking out the garbage.

In the questionnaire, we investigated frequency of going outdoors and destinations or purposes for going outdoors “before stay-at-home” and “during stay-at-home.” Our survey was conducted in May 2020. Therefore, “during stay-at-home” data were collected at that time, whereas the “before stay-at-home” data were recalled from participant memory. For frequency, we set 7 options (1: never go outdoors, 2: go outdoors once a month, 3: once in 2 weeks, 4: once a week, 5: 2-3 times a week, 6: 4-6 times a week, 7: everyday). If the frequency of going outdoors during the stay-at-home orders was lower than the frequency before the stay-at-home orders, that was defined as a “decrease.” For destination or purposes, we set 11 options (walking, going to the hospital, going to day-services, shopping, visiting friends, visiting family, doing leisure activities, doing intellectual activities, participating in club activities, volunteering, or other).

Self-rated changes in health status

We operationally defined health status in 4 domains: (1) motor function, (2) oral and swallowing function, (3) depressive mood, and (4) social networks. We designed the questionnaire with 24 questions from these domains. For motor function, we defined 6 construction concepts (muscle strength of lower limb, muscle strength of upper limb, body flexibility, balance ability, gait ability, and endurance) and set content for each concept. For swallowing and oral function, depressive mood, and social networks, we defined construction concepts using constructs and question contents from existing scales: questionnaire to screen dysphagia¹⁷⁾, Geriatric Depression Scale 15¹⁸⁾, and Lubben Social Network Scale¹⁹⁾, respectively. All content was designed to detect worsening health status, with 4 ordinal options (0: never worsened, 1: not worsened, 2: somewhat worsened, and 3: much worsened). See Appendix 1 for question contents.

Statistical analysis

We analyzed data with SPSS var. 26 for Mac (IBM Japan), and the level of statistical significance was defined as p ＜ 0.05.

To verify whether construct of self-rated changes of health status corresponded with predetermined domains, we conducted factor analysis. Method of factor extraction was least squares solution, and method of rotation was varimax rotation. The contents for each factor were determined by criteria that factor loadings were greater than 0.4 or smaller than －0.4. If an item was related to 2 or more factors, we considered bigger factor loadings in the item. After factors were determined, factor scores for each case were calculated in each factor. Higher factor scores indicated worsening health status.

To verify decrease in frequency of going outdoors during stay-at-home orders was related to self-rated change in health status, regression analyses were conducted for each factor score. The dependent variable for each regression analysis was each factor score. First, we checked whether each factor score was normally distributed using the Shapiro-Wilk test, and 3 out of 5 factor score did not normally distributed. Therefore, we converted all factor score (a continuous variable) into a dichotomous variable (0: under 50 percentile, 1: more than 50 percentile) and conducted multiple logistic regression analysis. Independent variable was decrease in frequency of going outdoors and covariates were age, sex, living alone, level of care needs, degree of multi-morbidity, and frequency of going outdoors before the stay-at-home orders.

Results

A total of 193 people responded to the questionnaire. After excluding those who did not permit us to collect data from their own medical or care records (n = 6), those who had missing values (n = 26), and those who were younger than 65 years (n = 13), 148 participants were included in this study.

Participant characteristics were shown in Table 1. People who decreased frequency of going outdoors were 35.1%. Going outdoors for walking, shopping, attending day services, and going to the hospital were likely to continue during the stay-at-home orders（proportion of interruption: 14.6%-34.2%）, whereas going outdoors to visit friends or family, doing leisure activities, doing intellectual activities, attending club activities, and volunteering tended to be interrupted (proportion of interruption: 64.3%-85.7%).

Table 1.

Participants Demographics (N = 148)

SD: standard deviation
Age, mean (SD)	81.4 (7.6)
Sex, n (%)
Male	67 (45.3)
Female	81 (54.7)
Living alone, n (%)	36 (24.3)
Decrease in frequency of going outdoors, n (%)	52 (35.1)
Level of care needs, n (%)
Not-eligible	3 (2.0)
Need-support-1	30 (20.3)
Need-support-2	30 (20.3)
Level-1	33 (22.3)
Level-2	31 (20.9)
Level-3	13 (8.8)
Level-4	6 (4.1)
Level-5	2 (1.4)
Disease, n (%)
Orthopedics / musculoskeletal disease	78 (52.7)
Stroke	41 (27.7)
Neurodegenerative disease except for dementia	20 (13.5)
Spinal cord injury	2 (1.4)
Peripheral nerve disease	7 (4.7)
Heart disease	24 (16.2)
Pulmonary disease	11 (7.4)
Kidney disease	9 (6.1)
Digestive disease	5 (3.4)
Psychiatric disease	6 (4.1)
Dementia	7 (4.7)
Cancer	12 (8.1)
Diabetes mellitus	19 (12.8)
Other diseases	14 (9.5)
Degree of multimorbidity, median (min-max)	2 (1-5)

In self-rated changes of health status, in 6 contents related to motor function, approximately 80% of participants answered “worsened” (75.0%-87.2%), whereas fewer participants answered “worsened” for 6 contents related to oral and swallowing function (20.9%-50.7%). In contents related to depressive mood and social networks, 72.3% of participants answered “worsened” in “Feeling energy,” and 75.7% answered “worsened” in “Getting together and speaking with friends.” More detail results were shown in Appendix 2.

While we had defined health status in 4 domains before analysis, factor analysis extracted 5 factors and the cumulative proportion of variance explained was 57.6% (See Table 2). We named these factors as follows: 1st factor: “motor function,” 2nd factor: “depressive mood.” 3rd factor: “oral and swallowing function,” 4th factor: “family network,” and 5th factor: “friend network.”

Table 2.

Question content and results of Factor Analysis

content	1st factor	2nd factor	3rd factor	4th factor	5th factor
Muscle strength of lower limb	.812	.157	.059	.072	.104
Muscle strength of upper limb	.740	.221	.178	.013	.028
Body flexibility	.727	.238	.168	.154	.064
Balance ability	.768	.159	.206	.223	.081
Gait ability	.730	-.022	.255	.114	.133
Endurance	.768	.133	.188	.119	.118
Difficulty to drink fluids	.227	.230	.674	.000	.010
Choking while eating	.142	.063	.842	.026	.037
Choking while drinking liquids	.161	.060	.736	.171	.063
Difficulty to eat hard foods	.354	.284	.358	-.119	.150
Food falls out of the mouth	.192	.035	.543	.173	.293
Food to be left in the mouth	.135	.167	.542	.077	.102
Satisfaction with life	.165	.766	.132	.171	.184
Feeling happy	.161	.713	.110	.194	.120
Feeling helpless	.257	.571	.159	.257	.012
Feeling worthless	.076	.561	.032	.188	.099
Being bored	.066	.564	.152	.137	.128
Feeling energy	.357	.549	.183	.143	.296
Getting together and speaking with friends	.186	.238	.167	.139	.629
Speaking with friends about private life	.133	.338	.154	.169	.795
Asking friends for help	.115	.133	.252	.382	.217
Getting together and speaking with family	.143	.302	.107	.653	.061
Speaking with family about private life	.108	.440	.039	.728	.204
Asking family for help	.135	.201	.058	.792	.050
variance explained (VE)	4.10	3.21	2.87	2.17	1.46
proportion of VE [%]	17.1	13.4	12.0	9.0	6.1
cumulative proportion of VE[%]	17.1	30.5	42.4	51.5	57.6
Cronbach α	0.91	0.85	0.83	0.85	0.79

Results of the multiple logistic regression analyses are shown in Tables 3. Decrease in the frequency of going outdoors was related to the 1st factor score (OR 2.16, 95%CI 1.04-4.47) and the 5th factor score (OR 2.22, 95%CI 1.08-4.59).

Table 3.

Results of multiple regression analysis for factor scores

Independent variable and covariates	1st factor score	2nd factor score	3rd factor score	4th factor score	5th factor score
Values in this Table mean Odds ratio (95% confident interval). 1st to 5th factor scores were indicators of motor function, depressive mood, oral and swallowing function, family network and friend network, respectively and they were converted into dichotomous values (0: under 50 percentile, 1: more than 50 percentile), which value “1” indicated greater deterioration of health status.†: Covariates Bold letters: p < 0.05
Decrease in frequency of going outdoors [decrease: 1]	2.16 (1.04-4.47)	1.18 (0.05-2.41)	1.01 (0.05-2.03)	0.87 (0.43-1.73)	2.22 (1.08-4.59)
Age†	1.05 (1.00-1.11)	0.99 (0.94-1.03)	0.99 (0.94-1.03)	0.98 (0.94-1.03)	1.02 (0.97-1.06)
Sex [female: 1] †	0.87 (0.42-1.84)	2.05 (0.99-4.24)	0.97 (0.47-1.97)	0.73 (0.36-1.49)	0.40 (0.19-0.85)
Care need level†	0.78 (0.61-1.00)	0.82 (0.64-1.05)	1.15 (0.91-1.46)	0.87 (0.69-1.09)	0.89 (0.70-1.13)
Living alone [alone: 1] †	1.20 (0.52-2.75)	1.28 (0.57-2.92)	0.48 (0.21-1.09)	1.33 (0.60-2.95)	0.86 (0.38-1.95)
Degree of multimorbidity†	1.58 (1.00-2.51)	0.78 (0.50-1.21)	0.92 (0.61-1.41)	1.09 (0.71-1.66)	0.73 (0.47-1.13)
Frequency of going outdoors before stay-at-home†	0.89 (0.66-1.20)	0.93 (0.69-1.24)	0.93 (0.70-1.24)	1.02 (0.77-1.73)	1.10 (0.82-1.48)

Discussion

Our study revealed self-rated changes of health status during stay-at-home orders among older adults using the long-term care insurance system of Japan. In the questionnaire, many participants answered “worsened” in motor function, feeling energy, and getting together and speaking with friends. During stay-at-home orders, in addition to restrictions for going outdoors, people were requested to keep “social distancing, or staying 2 meter (6 feet) apart”. Moreover, the fear of infection of COVID-19 expanded among people. These conditions forced people to alter their behavior and changed their environments, which might lead older adults to feel deteriorations in motor function, depressive mood, and friend network.

However, participants in this study were recruited from 2 day-service facilities and a nursing station, and these placed emphasis on rehabilitation services. Therefore, the participants in this study might be concerned with their motor function and thus sensitive to tiny changes therein, which means the participants might overestimate the changes in their motor function.

Multiple logistic regression analyses revealed that decrease in the frequency of going outdoors was related to self-rated changes in motor function and friend network. Previous studies showed that lower frequency of going outdoors led to impairments in mobility⁸⁾, ADLs⁹⁾, and IADLs^5,8,10). Moreover, lower frequency of going outdoors was also related to poor social networks¹⁾. Although outcomes in this study were not equal to those in previous studies, the results of present study corresponded with those of the previous studies. Decrease in frequency of going outdoors might causes lower amounts of physical activities and fewer opportunities to meet with friends, which can lead to feeling deterioration in motor function and friend network. Although this self-rated deterioration in motor function and friend network shown in this study did not necessarily indicate real changes of them, in pandemic, supports for older adults to help maintain their motor function and friend network might be needed.

Previous studies found that lower frequency of going outdoors was related to depression^1,2,5), whereas this current study demonstrated no relationship between decrease in frequency of going outdoors and change of depressive mood. This finding might have occurred because of differences in the contexts where each study was conducted. Our current study was conducted during an infectious disease pandemic and subsequent stay-at-home orders, which meant that other factors might be more related to depressive mood.

While previous studies found that lower frequency of going outdoors was related to oral function¹⁾ and reduction of meal intake⁶⁾, and we hypothesized that a decrease in frequency of going outdoors causes a reduction of meal intake and that leads to a decrease in oral and swallowing function, no relationship existed between decrease in frequency of going outdoors and self-rated change of oral and swallowing function. Decrease in frequency of going outdoors might indirectly lead to deterioration in oral and swallowing function because of a reduction in meal intake, which means stay-at-home for 2 months might be insufficient to lead to a decrease in oral and swallowing function.

Decrease in frequency of going outdoors was not related to deterioration in family network. Many participants in this study lived with family members. Moreover, their families might have visited them because Japanese government did not prohibit them from visiting in necessary or urgent situations. Therefore, even If the participants decreased their frequency of going outdoors, contact with the family network was likely to be maintained.

The present study had several limitations. First, among approximately 85% of participants in this study, care needs were not greater than Level 2. A statistical report of long-term care insurance system in May 2020 showed that the proportion of those who have a Need-Support-1 to Level-2 certification was 65.5%²⁰⁾, meaning that participants in this study were biased for those with mild or moderate impairments. Therefore, results of present study should not be adapted for people with more severe impairments (i.e., Level-3 to Level-5). Second, changes in health status in this study were self-rated and we asked participants whether their health status worsened. This situation might lead them answer “worsened”. Moreover, as mentioned above, participants in this study might overestimate the changes in their motor function. Therefore, it is needed to consider changes in health status in this study, especially, motor function might be biased. Further research based on objective data are needed, especially for change in motor function. Third, a previous study has shown a misclassification in the frequency of going outdoors⁸⁾. Moreover, variables of going outdoors before the stay-at-home orders were based on participants' recall. Thus, misclassification should be also considered in results of this study.

Conclusion

During stay-at-home orders, older adults felt deterioration in their motor function, in their feeling energy, and in their friend network, especially those adults who decreased the frequency of going outdoors felt more deterioration in their motor function and in their friend network.

Conflict of Interest

The authors declare no conflict of interest.

Appendix

Supplementary material（Appendix）: