Understanding the Impact of Care Literacy on Preventive Care: Evidence from Family Carers in Japan

Abstract

To address the importance of family carers’ understanding of care, encompassing their own care as well as the broader care and health social systems, this paper leverages the notion of ‘care literacy’. The aim of this study is to understand the variation in care literacy and the impact of care literacy on preventive care. The empirical focus is on working family carers for older relatives in Japan, through a cross-sectional online survey that includes a novel operationalization of care literacy, an established measure of health literacy, the assessment of information used to understand care, and measures of preventive care. The participants’ (n = 281) mean age was 53, with 44% women, and an average of 8.3 h per week caring for their parent(s). The measure of care literacy is shown to be correlated, as expected, but distinct to health literacy (correlation 0.60). Based on regression analysis of care literacy, significant explanatory variables are health literacy (p < 0.001), gender (p = 0.044), number of sources of information on care (p = 0.029), and care hours (p < 0.001). In contrast, proximity in living arrangements of the carer and care receiver and severity of care needs were not significant predictors. Turning to the impact of care literacy, care literacy is a significant explanatory variable for use of preventive measures (p = 0.002), in particular as related to nutrition (p < 0.001), frailty (p = 0.028), and general home renovations (p = 0.018). The pattern of results from this cross-sectional analysis indicates the importance of understanding the potential for improved care literacy as an enabler of better care.

1. Introduction

Given the continued importance of family carers, a significant issue is the extent of carers’ understanding of how to address their older relatives’ care needs. This entails their own care provision as well as care provided by other family members, the involvement of care and health social systems, and also adapting living conditions to reflect care needs. To address such family carers’ understanding of care, this paper leverages the notion of ‘care literacy’ proposed by Costantini et al. [1], which is defined as the knowledge and capabilities that enable people in need of care to live their daily lives in the community and facilitate potential health and care solutions. The notion of care literacy applies to the diverse stakeholders across the community: for these stakeholders to fulfill their distinct roles to support those needing care, the stakeholders need to have the relevant knowledge and capabilities.

Such care literacy is distinct from health literacy, a person’s broad ability to understand health-related issues [2], including the functional, communicative, and critical literacy elements [3,4]. While health literacy refers to understanding and deciding for one’s own health situation, in contrast, for family carers care literacy refers to care of their parents. At the same time, the care needs of parents also reflect the health of the parents. Hence, family carers’ more general understanding of health should potentially support a higher level of care literacy: this would imply some degree of link from health literacy to care literacy even while conceptually distinct. In addition, the care needs and more general care context potentially affect family carers’ need for and use of information related to care and hence their care literacy. Thus, a first purpose of this study is to understand the variation in care literacy across family carers.

In turn, such care literacy is conceptualized as enabling better care provision through understanding care needs and deciding how best to address these. Family carers have been recognized as having the potential to learn to address gaps in knowledge and capabilities [5,6]. Furthermore, this potentially encompasses a diverse spectrum of issues related to care, including: understanding how to adapt care to address the risks due to the care receiver having dementia [7]; adapting nutrition in general [8] as well as specifically for those with dementia [9]; and addressing the risk of falling and how such risks also depend on the pressures on carers [10]. Addressing such issues may entail taking proactive, anticipatory preventive measures, which would be in line with an emphasis in addressing literacy in such a care context as entailing the gathering of relevant information to inform decisions and take corresponding action. Such understanding has the potential to affect carers themselves, such as their quality of life and time spent on care [11], as well as their care provision, which is the focus of this study. Thus, the second purpose of this study is to understand the implications of care literacy on the use of preventive care measures.

This study focuses on family carers in Japan who care for their parents and also work. Such working family carers are an important group to understand as rapidly becoming the norm as aging of society progresses, in contrast to past practices in which non-working daughters-in-law were the main carers for their parents-in-law [12]. Notwithstanding the increased need for family carers to address care and work, the Japanese care system for older people, the Community-based Integrated Care Systems, is premised on the continued importance of family care [13]. Indeed, family members are the main providers of care to older adults: of the 4.9 million people receiving care, 1.0 million are in care facilities and 3.9 million receive care at home [14]. Further, considering family carers in general, which includes also other types of carers, in particular those who care for a spouse, for 67.8% of those receiving care the primary caregiver is a family member. Amongst these, 54.4% have the main care provider and care receiver live together, of which 20.7% are adult children care providers and just over half of care providers are below 70 years old [15]. Broadening beyond primary carers to all those providing care to family members (but not including childcare), based on a government survey run every five years [16], as of 2017 the total number of family carers was 6.28 million (5.7% of the population over 15 years old), of which 3.46 million worked (5.2% of all workers). As compared to 2012, the total number of working family carers increased by 0.55 million, a 19% increase, which is a 3.5% increase per year. As of 2017, working family carers were 43% men and 57% women. The age distribution of these working family carers was 11% below 40 years old, 19% 40–49 years old, 65% 50–69 years old, and 5% 70 years old and over, with very similar age distribution for men and women, resulting in an average age of around 55 years old for working family carers. The increasing prevalence of family carers who work points to the importance of better understanding this key group of carers for older relatives in the community.

Thus, the objectives of this study are to understand, based on a cross-sectional online survey of family carers, first the patterns in care literacy relative to health literacy and other demographic and care characteristics so as to explore the determinants of care literacy, and second the association of care literacy with undertaking preventive measures related to care to provide evidence of the impact of care literacy on care provision.

2. Materials and Methods

2.1. Participants

The cross-sectional survey was conducted through an online research service company, running from 15 May–2 June 2021. To identify a sample of family carers who take care of their parents and who also work, for the first step the researcher drew on a nationally representative database of adults aged 20–70 years old (e.g., age, gender, income, geographic location) to invite participation in the survey. This yielded a panel of 5000 respondents who provided answers to the questions that enabled the identification of the relatively small proportion of family carers who work and take care of their parents, which resulted in n = 329 respondents. These respondents were asked to complete a longer survey on care literacy, health literacy, use of preventive measures, and sources of media and information used to understand care. Their survey responses were screened, for example removing respondents whose survey answers indicated potentially unreliable responses through review of open-text answers and no variation in entry key used. A substantial majority completed the survey with usable entries, in particular on their care literacy, health literacy, use of preventive measures, and sources of media and information used to understand care, yielding a sample used of n = 281. Comparing the 281 respondents used versus the 48 respondents screened out, respectively, the demographic characteristics are: average age 53 vs. 44 years old, 42% vs. 30% women, 66% vs. 29% married, 62% vs. 57% with children, and 720 vs. 697 income (10,000 Yen per year). Thus, those screened out comprised more younger, single, childless men, though with similar income. All these differences are less than one standard deviation of the sample retained. Thus, to the extent that the screening inadvertently removed valid respondents the resulting shift in the sample is not so significant. The main risk of bias is that a subset of single male carers for their parents is excluded, which would point to the potential value of follow-on research on this subset of carers. Comparing the 281 respondents used to the aggregate 2017 data on working carers [16], the sample has similar average age (53 vs. 55 years old) and a higher proportion of men (58% vs. 43%). Thus, there is a risk of potential bias as compared to the overall population of working family carers in relation to gender. In turn, the generalizability of findings needs to bear in mind the difference in gender mix, while noting a closer match in age.

2.2. Care Literacy

The notion of care literacy is taken from Costantini et al. [1]. The specific measures to assess care literacy were developed for this study. The form and structure of the care literacy survey was based on the established health literacy survey by Ishikawa et al. [3] and [4]. Thus, the wording was aimed to address care (versus health) while keeping a similar survey approach. As a base from which to adapt to care, the Ishikawa et al. [3,4] health literacy nine-item scale, in Japanese, was used with the nine-item scale for care literacy also in Japanese, as the survey was administered in Japanese. The English translation of the question asked for care literacy is: ‘Since you started caring for your parent(s), have you done any of the following things to learn more about caring for them or anything related to care for them?’, with the nine follow-on items as follows: I gathered knowledge and information from many sources; I selected what I wanted from a wide range of knowledge and information; I understood the knowledge and information that I have gathered; I communicated my feelings and thoughts about care of my parents to the doctor, care managers, or other people close to me; I adapted the ways of my providing care and aspects of my parents’ everyday life based on the knowledge and information I have gathered; I thought about whether the knowledge and information I have gathered applies to my situation; I questioned the reliability of the knowledge and information I have gathered; I asked or checked whether the knowledge and information I have gathered is true or not; I researched to make decisions about day care facilities and care services in order to support my decision making. The wording of the scales reflects experience in conducting fieldwork on care with family carers. Each item was rated on a four-point score, from not at all (=1) to very well (=4). To compute an aggregate score for care literacy, the average of each item was first computed and then the average of the nine items. This approach was also taken to be in line with the computation of health literacy. Thus, the items used for care literacy are in line with general notions of literacy and health literacy, adapted to reflect that while health literacy is focused on understanding own health situation, care literacy is focused on understanding the care situation of parents’ as well as care provision by oneself, others and the care and health systems.

2.3. Health Literacy

Health literacy was assessed based on the approach developed by Ishikawa et al. [3,4], with original scale items available in Japanese and versions of this instrument used in different health care contexts. This instrument has been used in online surveys, based on a version in Italian to study COVID-19 vaccination [14]. The instrument is based on considering different aspects of health literacy: functional literacy, which is primarily about ability to read and understand information; communication literacy, which focuses on accessing information from diverse sources and explaining to others; and critical literacy, which is about assessing the quality and applicability of information to inform decision-making [2]. As in Japan the literacy rate is very high, functional literacy is taken as given. Hence, the nine-item scale used enables self-assessment of health communication literacy (five items) and health critical literacy (four items), each rated on a four-point score, from not at all (=1) to very well (=4), with the average score computed as a measure of health literacy. Participants were asked to answer the nine-item scale for their level of health literacy in general, not specific to the current moment.

2.4. Use of Preventive Measures

To understand respondents’ use of preventive measures taken to address their parents’ care needs, a broad set of measures was included. The set of preventive measures spanned various aspects, specifically asking about prevention: related to dementia; related to frailty; through partial renovation of their house; bathroom renovation; related to nutrition; and a residual category to cover other measures taken. The questionnaire for preventive measures was developed leveraging experience from prior qualitative fieldwork related to the family carers for their parents, which informed the choice of which items to include and the terms used in the questions. Within the survey used in the manuscript, the responses received were screened for anomalous or inconsistent answers, as well as a check that answers to the six questions were not highly correlated (i.e., respondents rated their answers differently across the scales). Respondents were asked to rate the extent of their use of each measure on a five-point scale, from ‘not at all’ = 1 through to ‘very much’ = 5. To provide a summary measure of the extent of preventive measures taken, the average rating of the six preventive measures was calculated.

2.5. Media and Information Sources

To understand what informs participants’ sense of care literacy, participants were asked to indicate the main sources of information they used spanning information from media, health and care services, and personal connections to understand overall care issues and for local area information related to care. In particular, options included national and local versions of broadcast and print media, social media, podcasts, medical and care service staff and professionals, as well as family members, friends, colleagues, neighbors, and ‘others’ for participants to define. To assess the breadth of information accessed, the measure is the total number of sources used summed across sources to access overall information and local information.

2.6. Other Variables

The other variables describe aspects of the respondents’ individual, family, and care context. These demographic and care-related variables are: age; household income; and hours of care provided to parents; gender; marital status (either married or not married); whether or not they have children; educational level; whether they are full-time employees, versus part-time or taking care leave; parental care needs, assessed based on the government scale of seven care needs categories, from needing light daily-living support to requiring full-time assistance; and distance from their parents, including whether living together, within walking distance, or a range of travel times from within 30 min to over 2 h away.

2.7. Statistical Analysis

To assess each of care literacy and health literacy on the respective nine-item scales, the average of the four-point score was used. The hours spent in care was taken to be the mid-point of each of the six categorical ranges up to 35 h, and 35 h for those reporting above 35 h, and thus treated as a continuous variable. The continuous variables were: age; household income; and hours of care provided to parents. The categorical variables were: gender; marital status; children; education; whether full-time employees; parental care needs; and distance from parents. First, the set of variables is analyzed with focus on correlations across related subsets of variables, in particular between care literacy and health literacy, and across the six measures of prevention and on comparing the means of variables, based on Student’s t-test, across sub-samples with high versus low care literacy. Next, the single-variable analysis was complemented by multivariable analysis. The first regression has care literacy as dependent variables, with as independent variables health literacy, access to information sources and the demographic and care-related variables. The second multivariable analysis has the preventive measures as dependent variables, with as independent variables care literacy and the demographic and care-related variables. A p-value less than 0.05 was considered to indicate statistical significance for all analyses. The data was analyzed using STATA (Statistics Data Analysis, Version MP-13.1 for Windows, StataCorp LP, Texas, TX, USA).

3. Results

The sample of family carers for older parents who also work or are taking leave from work due to provision of care (n = 281),

Table 1. Care literacy, heath literacy, demographic, work and care characteristics, and extent of preventive measures taken for sample.

Variable	Mean	Median	Std. Dev.
Care Literacy (1 to 4)	2.75	2.89	0.60
Health Literacy (1 to 4)	2.86	3.00	0.53
Age	52.7	54.0	9.27
Gender (Woman = 1; Man = 0)	0.42	0	0.49
Marital status (Married = 1; Other = 0)	0.66	1	0.47
Children (Yes = 1; No = 0)	0.62	1	0.49
Household Income (10,000 Yen per year)	720	650	449
Education level (% college or higher education)	0.75	1	0.44
Employment (% full-time not taking care leave)	0.88	1	0.33
Care provided to parents (hours per week)	8.54	5	9.71
Parents care needs (% with lighter need: support at levels 1 or 2, or care at level 1 of 5, on government scale)	0.46	0	0.50
Distance from home to parents’ home: % living together or within walking distance	0.46	0	0.50
Number of sources used to access information on care	5.28	4	4.41
Average of prevention measures (1 = not at all to 5 = very much)	3.22	3.17	0.73
Extent of prevention related to dementia (1 = not at all to 5 = very much)	3.28	3	1.03
Extent of prevention related to frailty (1 = not at all to 5 = very much)	2.89	3	1.09
Extent of prevention through partial renovation of home (1 = not at all to 5 = very much)	3.44	4	1.15
Extent of prevention through bathroom renovation (1 = not at all to 5 = very much)	3.06	3	1.23
Extent of prevention through nutrition (1 = not at all to 5 = very much)	3.60	4	0.93
Extent of other prevention measures (1 = not at all to 5 = very much)	3.04	3	1.00

, comprises 42% women and 58% men, with mean age 53 years (9.3 standard deviation). Participants’ living arrangements with respect to their parents receiving care vary, including 46% living with or close by their care-receiving parents. Additionally, their care-receiving parents span the range of the official seven care needs categories, from needing light daily-living support to requiring full-time assistance [17]. Participants’ self-reported hours spent on care average 8.5 h per week (9.7 standard deviation), with 41% under 3 h, 47% between 3–20 h, and 12% 20 h or more.

Considering the correlation amongst these variables, the correlation amongst the nine questions related to care literacy ranges from 0.50 to 0.73. By comparison, the correlation amongst the nine questions related to health literacy ranges from 0.45 to 0.74. The correlation of care literacy and health literacy is 0.6. The highest correlation between any pair formed by a question for care literacy and a question for health literacy is 0.49 (and the lowest is 0.28). This indicates the measure for care literacy is distinct to that for health literacy.

Amongst the variables relates to prevention, the correlations between pairs of measures range from 0.2 to 0.6, with the three highest correlations at: 0.6 between home renovation and bathroom renovation, 0.6 between measures related to dementia and frailty; and 0.4 between measures related to frailty and bathroom renovation. Thus, there is considerable variation across respondents in the pattern of preventive measures takes.

To consider how single variables vary with care literacy, first the subsamples with care literacy above versus below the mean care literacy of 2.75 are compared. Based on single-variable analysis, the statistically significant differences in means are: 0.9 for care literacy (p < 0.001), 0.5 for health literacy (p < 0.001), 12% more women (p = 0.042), 2.6 h more hours of care (p = 0.025), and 0.3 increase in average of prevention measures (p = 0.001) as well as, amongst the single prevention dimensions, increases of 0.3 for dementia related (p = 0.020), 0.3 for frailty related (p = 0.036), 0.4 for partial renovation (p = 0.006), and 0.6 for nutrition related (p = 0.001).

These patterns of significance were analyzed through two multivariable analyses. The first (

Table 2. Results from multivariable regression analysis with care literacy as dependent variable.

Independent Variable	Coef.	p-Value	95% Conf. Interval
Health Literacy (1 to 4)	0.65	0.000	0.54	0.76
Age	−0.002	0.536	−0.009	0.005
Gender (Woman = 1; Man = 0)	0.11	0.044	0.00	0.23
Marital status (Married = 1; Other = 0)	0.004	0.956	−0.14	0.15
Children (Yes = 1; No = 0)	0.06	0.424	−0.08	0.19
Household Income (10,000 Yen per year)	−0.0001	0.053	−0.0003	0.0000
Education level (% college or higher education)	0.06	0.385	−0.07	0.19
Employment (% full-time not taking care leave)	0.19	0.035	0.01	0.36
Care provided to parents (hours per week)	0.01	0.000	0.01	0.02
Parents care needs (% with lighter need: support at levels 1 or 2, or care at level 1 of 5, on government scale)	0.02	0.718	−0.09	0.13
Distance from home to parents’ home: % living together or within walking distance	0.08	0.161	−0.03	0.20
Number of sources used to access information on care	0.01	0.029	0.00	0.03
Constant	0.59	0.013	0.12	1.06

) has as care literacy as a dependent variable and as independent variables health literacy, the number of sources used to access information on care, and the demographic and care-related variables: age; gender; marital status; whether or not have children; household income; educational level; whether full-time employees or taking care leave; hours of care provided to parent; parental care needs; distance from parents.

The significant variables are: health literacy (p < 0.001); gender (p = 0.044), with higher for women; employment (p = 0.035), with higher employed full-time; hours of care provided (p < 0.001); and number of sources used to access information on care (p = 0.029). As a comparison, for these variables the crude coefficients and 95% confidence intervals are: health literacy (0.68 ([0.57 to 0.79)), gender (0.16 (0.02 to 0.30)), employment (0.17 (−0.05 to 0.38)), hours of care (0.01 (0.01 to 0.02)), and number of sources used to access information on care (0.03 (0.02 to 0.05)). Hence, other than for employment, the other variables significant in multivariable regression are also significant in respective crude regressions, namely gender, employment, number of information sources used, hours of care, and health literacy. In terms of the magnitude of the coefficients of these variables, these imply a higher care literacy of: 0.34 for a one standard deviation increase in health literacy; 0.11 for a one standard deviation increase in hours of care; 0.06 for a one standard deviation increase in number of information sources; 0.04 for women versus men; and 0.19 if full-time employment. Thus, care literacy is associated primarily with, as expected, higher health literacy, and as reflected in results for whether employed full-time and hours of care provided with increased claims on time for work and care.

The second multivariable analysis instead focuses on how care literacy is associated with preventive care measures: in particular, with (

Table 3. Results from multivariable regression analysis with average of preventive measures as dependent variable.

Independent Variable	Coef.	p-Value	95% Conf. Interval
Care Literacy (1 to 4)	0.29	0.002	0.11	0.48
Health Literacy (1 to 4)	−0.07	0.529	−0.27	0.14
Age	−0.01	0.062	−0.02	0.00
Gender (Woman = 1; Man = 0)	0.13	0.153	−0.05	0.30
Marital status (Married = 1; Other = 0)	0.12	0.285	−0.10	0.35
Children (Yes = 1; No = 0)	0.09	0.403	−0.12	0.30
Household Income (10,000 Yen per year)	0.0001	0.353	−0.0001	0.0003
Education level (% college or higher education)	0.16	0.119	−0.04	0.36
Employment (% full-time not taking care leave)	−0.12	0.394	−0.39	0.15
Care provided to parents (hours per week)	0.002	0.608	−0.007	0.012
Parents care needs (% with lighter need: support at levels 1 or 2, or care at level 1 of 5, on government scale)	−0.13	0.139	−0.30	0.04
Distance from home to parents’ home: % living together or within walking distance	−0.19	0.036	−0.37	−0.01
Constant	2.96	0.000	2.24	3.69

) the average of prevention measures as a dependent variable and as independent variables care literacy, health literacy, and the demographic and care-related variables: age; gender; marital status; whether or not they have children; household income; educational level; whether full-time employees or taking care leave; hours of care provided to parent; parental care needs; distance from parent.

The significant variables are for more preventive measure with: higher care literacy (p = 0.002) and shorter distance from home to parents’ home (p = 0.036). As a comparison, for these variables the crude coefficients and 95% confidence intervals are: care literacy (0.27 (0.14 to 0.42)) and distance from home to parents’ home (−0.14 (−0.31 to −0.03)). In terms of the magnitude of the coefficients of these variables, these imply an increase in average of preventive measures by: 0.17 for a one standard deviation increase in care literacy and 0.13 if living closer to or with parents. Thus, care literacy is associated with taking of more preventive measures.

As the set of preventive measures have varying degrees of correlation, a follow-on multivariable analysis has each preventive measure in turn as dependent variable and the same independent variables as in Table 3. The results indicate which aspects of prevention are associated with higher care literacy and other independent variables. In particular, for measures related to nutrition the significant variables are care literacy (p < 0.001), health literacy (p = 0.040), and gender (p = 0.027). For measures related to frailty the significant variables are care literacy (p = 0.028), household income (p = 0.031), and less severe parent care needs (p = 0.006). For measures related to home renovation the significant variables are care literacy (p = 0.018), marital status (p = 0.035), and education level (p = 0.019). For other measures, each has one significant variable: for measures related to dementia the shorter distance from home to parents’ home (p = 0.014); for bathroom renovation the age of respondents (p = 0.007), and for other measures whether have children (p = 0.013). Thus, in line with the relatively limited correlation across preventive measures, there is variation in which independent variables are significant. In this sample, increased care literacy is not associated with preventive measures related to dementia, bathroom renovation and other measures but is associated with measures related to nutrition, frailty, and home renovation.

4. Discussion

The notion of care literacy provides additional insights for understanding the situation of family carers who work and provide care for their older parents from which the survey respondents are drawn. Such working family carers are likely coordinate care with other family members as well as care and health services: for instance, the most accessed source for information on care are local care managers and care workers. Additionally, such working family carers, with variations across the sample in care hours provided as well as working hours, are likely to be under significant time pressure. Hence, for such working family carers, the provision of care to their parents is a significant commitment and likely to benefit from better understanding of care so as to shape how they provide care; thus, the notion of care literacy is potentially relevant to such participants.

The first set of results provide evidence of how care literacy varies systematically across the sample of working family carers. Importantly, in the results, while care literacy is related to health literacy, care literacy is distinct. Care literacy refers to the care of parents, whereas health literacy to one’s own health situation. Nonetheless, the care parents need is inextricably linked to the health of the parents; thus, a more general understanding of health, such as captured in one’s own health literacy, is likely to support higher level of care literacy. Moreover, health literacy centrally features the gathering of relevant information and using this for decision making, and is thus rooted in the literacy aspect, which is also relevant to literacy about care.

In addition to health literacy, the significant explanatory variables with the next highest impact on variation in care literacy are number of care hours provided and whether work full-time. This points to care literacy being more relevant to those more immersed into care and/or facing greater time pressure due to other commitments, notably work. The significance of gender could also indicate time pressure, as women in Japan are more likely to disproportionately fulfill domestic duties even if working and/or providing care to relatives [18]. These varied sources of time pressure indicate participants need to seek effective and efficient care, for which better information and the use of such information is likely central. Indeed, the breadth of information sources is also a significant explanatory variable, including gathering information from media sources as well as care and health professionals, family members, and others. This suggests the importance of understanding how each of the diverse set of stakeholders involved in care, in this study with focus on working family carers, access and use relevant information, as captured in the notion of care literacy.

The results point to the importance of care literacy in affecting care provision through the analysis of preventive measures of care respondents report undertaking. Care literacy is associated with a greater use in general of preventive measures, and in particular related to nutrition, frailty, and general home renovation. Nutrition measures are also associated with health literacy, which corresponds to the interplay of nutrition and health. Additionally, women undertake more preventive nutrition measures, which could reflect their involvement in meal preparation and hence scope for addressing nutrition. Frailty is associated with less severe parental care needs, as assessed by the government scale that determines extent of social care support. This could reflect that at higher levels of assessed need the increased support from care services also comprises advice to family carers, whereas at the lower levels of assessed need family carers receive less support and guidance and thus the family carer’s care literacy has a greater differential impact on care practices. Thus, such interventions to address frailty and nutrition, and potentially general home renovation, may be less triggered by specific recommendations from care and medical staff (as compared to for care of those with more severe care needs or with dementia) and thus more impacted by care providers’ level of care literacy.

As the survey is cross-sectional, these interesting patterns of association between care literacy and preventive measures is suggestive, warranting follow-up research to understand the temporal dynamics. For instance, understanding what triggers an increase in care literacy and how changes in care literacy are linked to subsequent changes in care provision and preventive measures. In addition, as care literacy depends also on care receivers’ condition this necessarily evolves over time with changing care needs. Moreover, care providers’ own health and care conditions and behaviors could affect their health literacy and care literacy as well as use of preventive measures, such as use of preventive measures for themselves affecting use of preventive measures for their parents. Thus, the cross-sectional form of the survey and scope of information gathered places limits on inferences drawn from the results.

Additionally, the pattern of results rests on the measures of care literacy and preventive measures. The measures for care literacy were developed to reflect the approach taken towards health literacy by Ishikawa et al. [3,4]. This points to the value of follow-up research to develop measures of care literacy for use not only with family carers but in a variety of contexts in line with the broader notion of care literacy in Costantini et al. [1]. Additionally, the preventive measures were developed for this survey, while respondents’ answers were screened for inclusion the measures have not been used in a broader context, the interesting pattern of results points to a valuable line of future research.

Further, the online survey format necessarily has some limitations, such as participants needing to be registered with the survey company and internet users, thus limiting external validity. The broad reach of the online survey does, on the plus side, provide the ability to address specific groups of interest, such as the working family carers focused on. The screening of survey respondents to form the final sample of n = 281 screened out n = 48 respondents based on usability of their survey responses. While the final sample and those screened out are not substantially different, those screened out are on average younger, single, childless men, though with similar income. To the extent that the screening inadvertently removed valid respondents, the main risk of bias is that a subset of single men who care for their parents is excluded. Given the increasing proportion of never-married men and the decrease in number of siblings in Japanese society, there is a possibility of such a subset of carers, which potentially merit follow-up focused research. As all the respondents are Japanese (which reflects that 98% of the population in Japan is Japanese [19]), there is no significant scope for race or cross-cultural subsamples to consider.

Bearing in mind the limitations highlighted, family carers for their parents who also work are an important group of carers to better understand. Broader government data on family carers who work have a similar average age to the sample and a higher proportion of women. The statistical analyses all included gender (and age) as an independent variable, thus mitigating the impact of the gender mix of the sample. In turn, the results contribute to understanding the over 3 million family carers who also work [16]. This sub-group of carers focused on is important as increasingly family carers also work and there is growing participation of men as well as women in care [12]. Thus, in a Japanese context the study contributes to understanding a significant and growing group of carers. Further, as super-ageing Japan is an important reference point to learn from for other countries with rapidly ageing populations, the findings are of broader interest.

Finally, care literacy is evidenced to be distinct from health literacy, associated with extent of involvement in care and work, and associated with greater use of selected preventive measures. These findings substantiate the relevance of the notion of care literacy and support further exploration for different stakeholders involved in care of the determinants and consequence of their care literacy, so as to contribute towards achieving a more sustainable society for those receiving care as well as those involved in care provision.