Abstract
Dementia has become a major public health concern globally. With no cure available and strong evidence for modifiable risk factors, it is imperative that the public are knowledgeable about dementia and to reduce their risk. The aim of this study was to measure the knowledge of dementia and its risk factors in the Australian public, as well as the number and type of information sources used. An online survey promoted through various social media platforms was completed by 596 Australian adults aged 18–78 years (59% aged 18–44 years; 78% female). Eighty-eight percent of respondents were able to recognise dementia from a vignette, more so from a moderate than from a mild symptom vignette. Only 19% of respondents had a good understanding of dementia, that is describing impairment in both cognition and function. Ninety-five percent of respondents recognised that genetics and old age contributed to a person getting dementia. However, respondents had poor knowledge of empirically supported modifiable risk factors, with most viewed as likely contributors to dementia at chance levels (50%) or below. Respondents reported using informal information sources more often than formal information sources to learn about dementia. The public appear to be able to recognise the symptoms of dementia but lack understanding of how dementia impacts both a person’s cognitive functioning and ability to perform everyday tasks. Furthermore, the public remain largely unaware of empirically supported modifiable risk factors that contribute to the development of dementia. It is imperative that the public are educated on how to access and evaluate dementia-related information sources in order to increase their knowledge and understanding of dementia.
Introduction
Dementia is defined as a progressive loss of cognitive function, resulting from brain changes, which significantly impairs social and occupational functioning (American Psychiatric Association, 2013; Dementia Australia, 2018a; Dening & Sandilyan, 2015; Rone-Adams et al., 2013). While there are many forms of dementia, Alzheimer’s disease is the most common, accounting for 50–75% of all dementias (Hudson et al., 2012; Livingston et al., 2017). Clinically, dementia is characterised as either early- or late-onset, with 65 years of age used universally to differentiate between the two (Ferreira et al., 2018; Koedam et al., 2010). With an ageing population and no cure for dementia in the foreseeable future, it is imperative that the public understand what dementia is and how to reduce their risk, given rising costs associated with dementia (Carpenter et al., 2011; Cations et al., 2018; Livingston et al., 2017; Low & Anstey, 2009).
Dementia has become a major public health concern globally and is the fifth highest cause of death (World Health Organisation, 2018). In 2010, it was estimated that over 44 million people globally were living with dementia (Prince et al., 2008). In 2020, this number increased to 50 million (Livingston et al., 2020). By 2050, the number of people globally with dementia is expected to rise to 152 million (Livingston et al., 2020). The direct and indirect costs of dementia place a large burden on individuals, families, communities, healthcare systems and government budgets (Carpenter et al., 2011; Hudson et al., 2012; Livingston et al., 2017; Low & Anstey, 2009). In 2015, it was estimated that the cost of dementia globally was US$818 billion, accounting for formal and informal costs of care (Prince et al., 2008). By 2030, the cost of dementia is expected to rise to US$2 trillion (Prince et al., 2016).
The economic impacts of dementia are significant, as are the health, social and psychological costs to people living with dementia, their families and carers. Not only do people living with dementia face the prospect of a reduced lifespan, but they also typically live with a greater number of years of reduced quality of life and increased disability (Prince et al., 2015). The families of people living with dementia often experience a prolonged grieving process for their loved one that is like grieving a death (Chan et al., 2013). Denial and anger at the diagnosis, guilt for perhaps having not done more and sadness for ‘loss’ of the person now living with dementia are commonly experienced emotions (Chan et al., 2013). Additionally, it can be difficult for family members to express and explain their grief to others while the person living with dementia is alive (Papastravrou et al., 2007; Rudd et al., 1999). Family members, particularly female relatives, are heavily impacted as they typically devote large amounts of time and resources into providing care (Broadaty & Donkin, 2009; Papastravrou et al., 2007). As a result, many carers are unable to fulfil workplace and other family responsibilities, which can result in un- or underemployment, increased mental and physical health concerns and increased social isolation. These factors can place further strain on the relationship between family members and loved ones living with dementia (Broadaty & Donkin, 2009; Papastravrou et al., 2007).
Dementia recognition is the identification of dementia as a diagnosis, when presented with a case study or vignette that describes the symptoms of dementia (Blay & Piza Peluso, 2008; Low & Anstey, 2009; Low et al., 2011). Dementia understanding relies on a person’s knowledge of dementia and the ability to comprehend the impact that it has on cognitive and occupational functioning. Recognising and understanding dementia is critical for early diagnosis, the exploration of treatment options, the empowerment of individuals to participate in managing the progression of their symptoms and increasing the quality of life for both the individuals and their caregivers (Bradford et al., 2009; Breining et al., 2014; Diamond & Woo, 2014; Riva et al., 2012). No quantitative studies to date have evaluated the general public’s understanding of dementia, that is they have not explicitly asked respondents to describe the characteristics of dementia.
It is important to know whether the public recognise and understand what dementia is as misconceptions and stigma affect the public’s willingness to accept empirically supported dementia information (Arai et al., 2008; Breining et al., 2014; Cations et al., 2018; Jang et al., 2010; Purandare et al., 2007; Roberts et al., 2014; Werner, 2003). For example a commonly held misconception is that dementia is a part of the normal ageing process (Hudson et al., 2012; Low & Anstey, 2009). Those who hold this belief are (1) less likely to engage in preventative efforts to reduce their risk of dementia; (2) less likely to recognise the symptoms of early dementia, leading to delayed diagnosis which significantly limits an individual’s ability to manage the diagnosis; and (3) less likely to engage in help-seeking behaviours (Arai et al., 2008; Breining et al., 2014; Cations et al., 2018; Jang et al., 2010; Prince et al., 2008; Purandare et al., 2007; Roberts et al., 2014; Smith et al., 2014; Werner, 2003).
Low and Anstey (2009), along with Low et al. (2011) and Blay & Piza Peluso (2008), measured dementia recognition in the public by asking respondents to identify dementia from a clinical vignette. In their study of 2000 Australian adults, Low and Anstey (2009) reported that 82% of respondents were able to recognise dementia from a written vignette. No differences in recognition were found between vignettes presenting mild and moderate symptoms, nor between the genders of the characters. However, being female, having known or cared for someone with dementia and having a higher household income were associated with higher rates of dementia recognition. Using a sample of 350 Italian, 414 Greek and 437 Chinese Australians, Low et al. (2011) found that recognition of dementia was lower in these cultural groups (61%, 58% and 72%, respectively) than in third-generation Australians (n = 500, 85%). Low et al. (2011) proposed that misconceptions about dementia being a normal part of ageing contributed to the lower recognition rate in the ethnic minority groups. Blay and Piza Peluso (2008) reported that less than 5% of their 500 Brazilian respondents identified dementia or Alzheimer’s disease from a vignette. Furthermore, 22% of respondents believed that the symptoms presented were part of the normal ageing process.
There is strong evidence that we can prevent late-life dementia by around 30–35% (Livingston et al., 2017; Norton et al., 2014); therefore, it is important that the general public become educated about its risk factors (Cations et al., 2018; Livingston et al., 2017; Norton et al., 2014). There are complicated public health messaging considerations, as modifiable risk factors such as obesity and hearing loss appear to exert greater influence at specific lifespan stages. While the public are generally able to correctly identify non-modifiable risk factors, such as age and genetics (Diamond & Woo, 2014; Low & Anstey, 2009; Nielsen & Waldemar, 2016; Roberts et al., 2003; Zheng & Woo, 2016), modifiable risk factors are not well recognised (Blay & Piza Peluso, 2008; Breining et al., 2014; Carpenter et al., 2011; Cations et al., 2018; Hudson et al., 2012; Low & Anstey, 2009; Riva et al., 2012; Yang et al., 2015). Knowledge of modifiable risk factors is significant to individuals as it demonstrates that they can decrease (or increase) their risk of dementia throughout their lifespan and that developing dementia is not an inevitable part of ageing (Dementia Australian, 2018a; Livingston et al., 2017; Norton et al., 2014). Good knowledge of dementia and its risk factors could facilitate risk reduction behaviours and guide health and medical decisions (Carpenter et al., 2011; Hudson et al., 2012; Livingston et al., 2017; Low & Anstey, 2009; Lüdecke et al., 2016; Matioli et al., 2011).
Identifying the sources of information that the public use to gain knowledge of dementia could assist policymakers to tailor educational and awareness initiatives. Commonly used sources may also be more accessible, which may assist in engaging and informing more people about dementia (Carpenter et al., 2011; Cations et al., 2018; Livingston et al., 2017; Zhang et al., 2017). Identifying sources that are disseminating incorrect information is also useful as they could be updated to reflect empirically supported information and challenge misconceptions and stigma surrounding dementia (Carpenter et al., 2011; Lee et al., 2010; Sun et al., 2014). Previous studies have shown that those who utilise multiple (four or more) and various information sources tend to be more knowledgeable about dementia (Carpenter et al., 2011; Roberts et al., 2003; Sun et al., 2014).
The aim of this study was to investigate what the Australian public know and understand about dementia and its risk factors. The current study will extend and update the findings of Low and Anstey (2009) by replicating their study design (a decade later) and assessing knowledge of dementia and its risk factors. Since the study by Low and Anstey (2009) was published, there has been increased publicity of dementia in health campaigns and in the popular media. Therefore, it would be expected that the public will demonstrate a greater recognition and understanding of dementia. Furthermore, we assess newly identified modifiable risk factors that have not been assessed previously, such as hearing loss and physical activity. It is hypothesised that the majority (80% or more) of respondents will recognise dementia and that being female, personal dementia experience and having a higher household income will be associated with increased dementia recognition. Furthermore, the public will identify empirically supported non-modifiable risk factors at greater than chance (50%) levels but not empirically supported modifiable risk factors. Finally, the public will utilise informal information sources to inform their knowledge of dementia more than formal information sources, and those who utilise four or more information sources will be knowledgeable about dementia. An exploratory analysis will be conducted to investigate the public’s dementia understanding.
Method
Procedure
An online survey was conducted via SurveyMonkey between 1 June and 10 August 2018. To prevent priming of answers to the written vignettes presented, the survey was introduced as ‘Understanding Chronic Health Conditions in Adulthood’. Respondents were provided with a participant information sheet at the start of the survey and were informed of the true aims of the study at the end of the survey. The respondent’s completion and submission of the survey was classified as informed consent. The survey was distributed via the Cognitive Ageing and Impairment Neurosciences laboratory’s (at the University of South Australia) Facebook and Twitter accounts, as well as through Gumtree and SONA (an online research participation sign-up system).
Facebook and Twitter paid advertisements were used to distribute the survey in various community, sporting, health and hobby/leisure groups that included people aged 18 and older. Facebook automatically placed advertisements on Instagram. All participants who completed the survey were eligible to enter a gift card draw in which they could receive one of five AU$200 gift cards.
Participants
A total of 719 Australian adults attempted to complete the survey. Respondents were excluded from participation if they (1) had a diagnosed cognitive impairment, such as an intellectual disability or dementia; (2) were under the age of 18 years; (3) resided outside of Australia; and (4) had already completed the survey. Thirty-four respondents were excluded from participating in the survey based on these criteria. A further 89 respondents were excluded as they did not complete the survey. In total, 596 respondents were included in the analyses. Ethical approval was obtained for this study (University of South Australia Human Ethics Committee).
A power analysis was performed using G*Power (Faul et al., 2007) for sample size estimation based on data from Low and Anstey (2009) (N = 2000) on variables that were predictors of dementia recognition. The variables included female gender, personal dementia experience and higher household income. The power analysis revealed that using a power of 0.80 at a significance level of α = 0.05, an estimated sample size of 268 would be needed for female gender (Cohen’s d = 0.34), 114 for having known someone with dementia (Cohen’s d = 0.53), 164 for having cared for someone with dementia (Cohen’s d = 0.44) and for household income, 1436 (US$15,600–US$52,000), 252 (US$52,000–US$104,000) and 166 (>US$104,000) (Cohen’s d = 0.15, 0.36 and 0.44, respectively) when compared to the reference category (<US$15,600).
Measures
The survey utilised in this study was adapted from Low and Anstey (2009). Our survey questions are in the Supplementary Material. The survey was composed of three sections. The eligibility section of the survey contained four questions (e.g. do you have a diagnosed cognitive impairment such as an intellectual disability or dementia?). The demographic section of the survey contained 11 questions (pertaining to, e.g. gender, marital status and previous experience in a health professional role). The dementia section contained 11 questions relating to dementia recognition, dementia understanding, knowledge of dementia risk factors and information sources used by the public (e.g. most specialists would agree that the person in the previous vignette has dementia – do you know what dementia is?). The subsections dementia recognition, dementia understanding, knowledge of dementia risk factors and information sources used by the public are parts of the dementia section mentioned above. The survey contained 26 questions in total and took an average of 11.57 minutes to complete (SD = 7.02 minutes). Measures were presented in the order in which they are described here.
Dementia recognition
Dementia recognition was measured by randomly allocating respondents to one of two vignettes, originally designed by Low and Anstey (2009). Notably, the conceptualisation of clinical dementia being characterised by cognitive and functional impairments has not changed. Each vignette described symptoms and behaviours of a person who met the diagnostic criteria for Alzheimer’s disease according to the Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-V) (American Psychiatric Association, 2013). Vignette A described mild symptoms and contained the lead character, John. Vignette B described moderate symptoms and contained the lead character, Mary. After reading the vignette, respondents were asked to state what was wrong with John/Mary (open response). As there are many forms of dementia, all forms were considered correct (Brown et al., 2017; Hudson et al., 2012; Livingston et al., 2017; Zhang et al., 2017).
Dementia understanding
Respondents were advised that most specialists would agree that the person in the previous vignette had dementia and were asked if they knew what dementia was (forced choice). Respondents who answered ‘yes’ to this question were asked to describe their understanding of dementia (open response). Responses were placed into three categories: no understanding, some understanding and good understanding. ‘No understanding’ was defined as responses describing an incorrect understanding of dementia and for those who answered that they did not know what dementia was. ‘Some understanding’ was defined as describing either cognitive impairment or the ability to perform everyday tasks. ‘Good understanding’ was defined as describing impairment in both cognitive functioning and the ability to perform everyday tasks.
Knowledge of dementia risk factors
Respondents were asked to rate on a four-point Likert scale (very likely to very unlikely) which factors contributed to a person developing dementia; 31 potential risk factors were presented. Nine modifiable and two non-modifiable risk factors supported with empirical evidence from Livingston et al. (2017) and Norton et al. (2014), which represented the most up-to-date understanding of dementia risk factors, were included. Two risk factors with emerging empirical support, head injury and stroke, were also included (Fann, et al., 2018; Kuźma et al., 2018; Savva & Stephan, 2010). Seven risk factors with some empirical support from Livingston et al. (2017) and Norton et al. (2014) were also added. These included hearing loss in late-life, hypertension in late-life, smoking in midlife, obesity in late-life, depression in midlife, physical inactivity in midlife and diabetes in midlife. Furthermore, nine risk factors with no empirical evidence were included, which were identified from previous literature and informal consultation with researchers in the field of dementia and websites (Alzheimer’s Association, 2018; Blay & Piza Peluso, 2008; Breining et al., 2014; Carpenter et al., 2011; Cations et al., 2018; Hudson et al., 2012; Low & Anstey, 2009; Riva et al., 2012; Yang et al., 2015); these included brain disease, aluminium, stress, virus/infection, laziness, dental fillings, flu shots, aspartame and lack of mental activity.
Information sources used by the public
Respondents were asked to select as many of the 20 listed information sources that they had remembered using to inform their knowledge of dementia. The information sources were categorised as popular media, academic, educational, health/advocacy, personal, no sources and other (for all other responses). These categories were then further divided into formal (academic, education and health/advocacy) and informal (popular media and personal) sources.
Design
A cross-sectional, between-subjects design was utilised in this study. Descriptive statistics (e.g. percentage of people who reported knowing someone with dementia) were used to describe the data. The independent variables in this study were gender, age, first language, education, state/territory resided in, location, marital status, place of birth, annual household income, employment status, previous experience in a health professional role and personal dementia experience. The dependent variables in this study were dementia understanding, dementia recognition, dementia risk factors and total information sources.
Statistical analysis
The data were processed and analysed using Statistical Product and Service Solutions (version 22) to form the variables in the data analysis. Descriptive statistics were performed to obtain percentages for dementia recognition and dementia understanding. Chi-square tests of independence were performed to investigate associations between gender, personal dementia experience, annual income and dementia recognition. Variables that were statistically significant at the 0.05 level in the chi-square tests of independence were entered into a binomial logistic regression (enter method) to investigate which, if any, of the variables were the best predictors of dementia recognition. The data from nine respondents were removed from these analyses as there was not enough power to run the analyses for the gender variable in the ‘other’ category.
Binomial tests were conducted to investigate if the proportions of correct answers for each of the dementia risk factors were significantly different from chance (0.50). Chi-square tests of independence were also conducted to investigate associations between dementia recognition, dementia understanding and total information sources reported. Bonferroni adjustments for multiple comparisons of the three independent groups were made (0.05/6), and alpha was set to <0.008.
Results
Sample characteristics
The final sample was predominantly female (78%), aged 18–44 years (59%), had completed university (55%), resided in South Australia (44%), were married or de facto (51%), spoke English (96%), were born in Australia (84%), resided in a major urban area (75%) and had known but not cared for someone with dementia (58%) (see Supplementary Material).
Understanding of dementia
Twenty percent (117) of respondents had no understanding of dementia. Sixty-one percent (368) of respondents had some understanding of dementia. All respondents in the some understanding category detailed cognitive impairments, primarily memory loss, that is, none included functional impairments (activities of daily living and instrumental activities of daily living). Nineteen percent (111) of respondents had a good understanding of dementia (i.e. detailed both cognitive and functional impairments) (see Supplementary Material).
Recognition of dementia
Eighty-eight percent (525) of respondents correctly identified that the character in the clinical vignette was presenting with dementia. A chi-square test of independence was conducted between dementia recognition and vignette allocation. There was a statistically significant association between dementia recognition and vignette allocation: Those allocated to the moderate symptom vignette (257 of 301) were significantly more likely to correctly recognise dementia than those allocated to the mild symptom vignette (268 of 295), χ2(1) = 4.24, p = 0.04. The association had a small effect size (Cohen, 1988), Cramer’s V = 0.08 (see Supplementary Material).
Predictors of dementia recognition
None of the univariate analyses were significant at the alpha <0.05 level (refer to Supplementary Material for cross tabulations) including gender, personal dementia experience and income. A binomial logistic regression (enter method) was performed to ascertain if the univariate analyses were masking any multivariable significant effects (i.e. once covariance was considered). The logistic regression model was not statistically significant, χ2(6) = 9.03, p = 0.17; the model explained 2.9% (Nagelkerke R2) of the variance in dementia recognition. Neither gender and personal dementia experiences nor income were statistically associated with correct dementia recognition (see Supplementary Material).
Knowledge of dementia risk factors
Empirically supported non-modifiable risk factors
The binomial test indicated that the proportions for genetics, 0.95, and age, 0.95, were significantly greater than 0.50 (chance), both p < 0.001 (see Supplementary Material).
Empirically supported modifiable risk factors
The binomial tests indicated that the proportions for less education (in early life), 0.25; hearing loss (in midlife), 0.24; and obesity (in midlife), 0.43, were significantly lower than 0.50 (chance), all p < 0.001. The proportions for hypertension (in midlife), 0.48 (p = 0.306); diabetes (in late-life), 0.47 (p = 0.130); and heart disease, 0.47 (p = 0.204), did not significantly differ from chance (0.50). The proportions for smoking (in late-life), 0.57; depression (in late-life), 0.69; physical inactivity (in late-life), 0.65; and social isolation (in late-life), 0.80, were significantly higher than 0.50 (chance), all p < 0.001 (see Supplementary Material).
Modifiable risk factors with emerging empirical support
The binomial tests indicated that proportions supporting stroke (0.91) and head injury (0.88) were significantly higher than 0.50 (chance), both p < 0.001 (see Supplementary Material).
Risk factors with some empirical support
The binomial tests indicated that the proportions for hearing loss (in late-life) (0.28) were significantly lower than 0.50 (chance), p < 0.001. The proportions for smoking (in midlife), 0.59; depression (in midlife), 0.65; and physical inactivity (in midlife), 0.63, were significantly higher than 0.50 (chance), all p < 0.001. The proportions for diabetes (in midlife), 0.45 (p = 0.030), and hypertension (in late-life), 0.50 (p = 0.967), did not significantly differ from 0.50 (see Supplementary Material).
Popular beliefs
The binomial tests indicated that the proportions for dental fillings (0.14), flu shots (0.04) and aspartame (0.16) were significantly lower than 0.50 (chance), all p < 0.001. The proportions for brain disease (0.92), aluminium (0.67), stress (0.76), virus/infection (0.60), laziness (0.75) and lack of mental activity (0.83) were significantly higher than 0.50 (chance), all p < 0.001 (see Supplementary Material).
Dementia information sources
On average, respondents reported using 4.62 information sources (SD = 2.94 sources). Informal sources of information (1606) (popular media (1008) and personal (598)) were endorsed more often than formal sources of information (1089) (health/advocacy (462), educational (328) and academic (299)). No sources (12) and other (50) were also reported by respondents. A chi-square test of independence was conducted between total information sources reported and dementia recognition. There was no statistically significant association between total information sources reported and dementia recognition, χ2(1) = 3.41, p = 0.07. The association was small (Cohen, 1988), Cramer’s V = 0.08 (see Supplementary Material).
A chi-square test of independence was conducted between total information sources reported and dementia understanding. There was a statistically significant association between total information sources reported and dementia understanding, χ2 (2) = 13.92, p = < 0.001. The association was small to moderate (Cohen, 1988), Cramer’s V = 0.15 (see Supplementary Material). Post hoc comparisons (Bonferroni correction) revealed that those who had no understanding of dementia were significantly more likely to report using three or less information sources (p < 0.001). No significant differences were found among those who had some understanding of dementia (p = 0.76) and those who had a good understanding of dementia (p = 0.01) (see Supplementary Material).
Discussion
The current study suggests that the Australian public have good recognition of the symptoms of dementia, with 88% of participants being able to do so. However, their understanding of dementia and its risk factors is limited. These findings are important as the risk of dementia can be reduced through the modification of lifestyle factors (Livingston et al., 2017; Norton et al., 2014).
Only 19% (111) of respondents demonstrated a good understanding of dementia, in which they accurately described dementia as a disorder that impairs cognitive functioning and an individual’s ability to perform everyday tasks. The majority (62% (368)) of the respondents demonstrated some understanding of dementia by identifying cognitive impairments in one area of functioning. Interestingly, all respondents in this category reported cognitive impairments (commonly described as memory impairments) but no functional symptoms. While this knowledge assists with recognition, it lacks the depth of understanding that impairments, resulting from dementia, reach other cognitive functions (beyond memory) and significantly impair an individual’s ability to perform everyday tasks (American Psychiatric Association, 2013; Dementia Australia, 2018a; Dening & Sandilyan, 2015; Rone-Adams et al., 2013). This heavy emphasis on memory loss could also be reflective of misconceptions about memory loss, dementia and normal ageing (Ayalon & Areán, 2004; Blay & Piza Peluso, 2008; Jang et al., 2010; Zheng & Woo, 2016).
The findings of this study revealed that 20% (117) of respondents demonstrated no understanding of dementia. It appears that while the public can recognise dementia as a diagnosis based on symptom presentation, a minority were unable to define dementia themselves without prompts. This new finding indicates that the public may not have an adequate understanding of both cognitive and functional symptoms of dementia. A lack of understanding could have serious implications, including delays in help-seeking behaviours and a reluctance to accept empirically supported dementia information and treatment (Arai et al., 2008; Breining et al., 2014; Cations et al., 2018; Jang et al., 2010; Purandare et al., 2007; Roberts et al., 2014; Werner, 2003).
The majority, 88% (524), of respondents were able to recognise dementia from the written clinical vignettes, as expected, consistent with Low and Anstey’s (2009) reported 82% correct recognition. Recently, dementia has become a more publicised issue in Australia, with campaigns including ‘Dementia Awareness Month’ and ‘Become a Dementia Friend’ raising the profile of dementia in the public (Dementia Australia, 2018b; Dementia Friendly Communities, 2018). Popular US and European TV programmes (e.g. Grey’s Anatomy, Neighbours and The Sopranos) and films (e.g. The Notebook and Still Alice) have also featured characters presenting with symptoms of dementia (Dementia Australia, 2018a; Hill, 2018; Stevenson, 2013). Contrary to expectation, the public’s recognition of dementia and its symptoms did not appear to have substantially increased in the past decade, despite public health campaigns and popular media.
Although the increase in recognition (Low and Anstey, 2009) is small, 82% compared to 88% in the current study, this could still have a considerable impact on real-world recognition of dementia at scale. Increased recognition of dementia in the public can assist to encourage help-seeking behaviours, in early diagnosis, and in exploration of treatment options; empower individuals to participate in managing the progression of their symptoms and increase quality of life for both the individuals and their caregivers (Bradford et al., 2009; Breining et al., 2014; Diamond & Woo, 2014; Jang et al., 2010; Riva et al., 2012). However, it is important to note that recognition in real life may be poorer than the results in this study indicate. Family members may not notice subtle changes in cognitive functioning or may attribute changes to normal ageing (Diamond & Woo, 2014; Jang et al., 2010; Low & Anstey, 2009). Individuals experiencing symptoms of dementia may be delayed in obtaining a diagnosis, meaning this can significantly restrict treatment options and place significant burden on both the individuals and their families (Bradford et al., 2009; Dubois et al., 2016).
There was a statistically significant association between correct recognition of dementia and the vignette that the respondent was allocated to, although the effect size was only small to medium. Respondents who were allocated to the moderate (Mary) vignette were significantly more likely to correctly identify dementia than those allocated to the mild (John) vignette. This contrasts with the findings of Low & Anstey (2009) who reported no statistically significant differences in recognition between the two vignettes. There are two possible explanations for these findings. The first is that the public may be more familiar with the symptoms of moderate dementia than mild. The second explanation is that more often, respondents may expect females (Mary) to have dementia, rather than males (John). Females over the age of 65 are twice as likely to develop dementia after age 85 compared to males (Andersen et al., 1999; Podcasy & Epperson, 2016). Dementia affects females disproportionately as females tend to live longer and increasing age is a significant risk factor for dementia (Andersen et al., 1999; Dementia Australia, 2018a; Podcasy & Epperson, 2016). Furthermore, the presentation of dementia symptoms in males is often subtler and progresses slower than in females (Peres et al., 2011). If symptoms of dementia in males are not as easily recognised by members of the public, this could significantly delay diagnosis and impact treatment options (Andersen et al., 1999; Peres et al., 2011; Podcasy & Epperson, 2016). Future studies should counterbalance gender and symptom severity.
Being female, higher annual income and knowing or caring for someone with dementia were not statistically significant predictors of dementia recognition in this study. These findings did not support the hypothesis and were in contrast to Low and Anstey (2009) who found these factors significantly associated with dementia recognition. This may be due to the sample not being representative (biased to young, highly educated females from South Australia). Or it may be the case that these factors no longer related to increased recognition of dementia due to recent health campaigns and dementia in the popular media.
Australian adults are not well informed about modifiable risk factors for dementia. Head injury and stroke, which are dementia risk factors with emerging empirical support, were correctly supported at greater than chance levels. Empirically supported non-modifiable risk factors, genetics and old age, were believed to be likely contributors of dementia by 95% of respondents (for both). This is a positive indication as genetics and old age play a major role in late-onset (Livingston et al., 2017; Norton et al., 2014). Four of the 10 modifiable risk factors with empirical support – smoking, depression, physical inactivity and social isolation in late-life – were identified by participants at greater than chance levels.
However, six out of the 10 empirically supported modifiable risk factors were not supported by participants (i.e. proportion supporting them was significantly less than chance): less education in early life; hearing loss, hypertension and obesity in midlife; diabetes in late-life and heart disease. It is important to note that participants were not just unsure about this set of risk factors, as would be indicated if no significant difference from chance. This demonstrates that the public are not fully aware of the influence of modifiable risk factors at particular stages of life. For example, the influence of depression in midlife on dementia risk has strong empirical evidence but depression in late-life does not (Livingston et al., 2017; Norton et al., 2014). Public health campaigns should target campaigns around these risk factors.
Knowledge of modifiable risk factors by the public is crucial for reducing the prevalence of dementia (Cations et al., 2018; Daviglus et al., 2010; National Institute for Health and Care Excellence, 2015; Livingston et al., 2017; Norton et al., 2014). 30–35% of all dementias are attributable to to modifiable risk factors (Livingston et al., 2017; Norton et al., 2014). Crucially, less education in early life and hearing loss in midlife are two of the greatest modifiable risk factors for dementia, accounting for 8–19% and 9% of attributable risk, respectively. These were the least understood by Australian adults. However, this strong empirical evidence for modifiable risk factors has only been demonstrated relatively recently (Livingston et al., 2017; Norton et al., 2014). This may partly explain the lack of knowledge of modifiable risk factors in the public; that is, there may be a significant lag in communication. Also contributing to this lack of knowledge is that dementia risk factor information is often conflicting depending on the studies published (Carpenter et al., 2011; Cations et al., 2018). For example, Livingston et al. (2017) identified nine modifiable risk factors associated with specific life stages. In contrast, Norton et al. (2014) identified seven risk factors, of which only three were associated with a specific life stage. It is difficult for policy makers to design health campaigns that promote the influence of modifiable risk factors at one stage of life and not another. Furthermore, it is difficult for the public to identify which information they should follow and which they should ignore (Carpenter et al., 2011; Cations et al., 2018). This issue is further exacerbated by the use of informal information sources which may disseminate incorrect information about dementia risk factors (Carpenter et al., 2011; Lee et al., 2010; Sun et al., 2014).
Participants typically did not support non-empirically supported factors (‘popular beliefs’ such as laziness and weakness of character) as being dementia risk factors, with rates typically lower than those reported in Low and Anstey (2009). For example, only 33% believed that aluminium was a likely risk factor for dementia. However, the fact that such factors are still believed to increase the risk of dementia is still a cause for concern and must be addressed (Blay & Piza Peluso, 2008; Low & Anstey, 2009; Low et al., 2011). Poor knowledge of dementia risk factors can contribute to the perpetuation in the belief of non-empirically supported risk factors and inhibit preventative and help-seeking behaviours (Blay & Piza Peluso, 2008; Low & Anstey, 2009; Low et al., 2011).
Respondents reported acquiring information about dementia from informal information sources more often than formal information sources. These findings are consistent with Roberts et al. (2003) and Sun et al. (2014). The number of information sources reported did not have a significant effect on dementia recognition, contrary to expectation. Additionally, those who reported using four or more information sources did not have a statistically significant greater understanding of dementia than those who reported using three or less information sources. This suggests that it may be the quality of the information source used, rather than the number, that influences dementia knowledge.
Strengths, limitations and future directions
This study is one of few that have investigated dementia recognition and conceptual understanding of dementia in the public, as well as to include an extensive list of dementia risk factors. There are three significant limitations of the current study, the first of which is the unequal distribution of the sample. The sample was largely composed of young (18–44 years of age), university-educated females from South Australia. The sample obtained in this study does not align with the sample obtained by Low and Anstey (2009) as they employed a stratified sampling approach based on gender and geographical region (phone numbers randomly sampled from Whitepages; 23% response rate of eligible participants). Low and Anstey (2009) have suggested that over-representation of such groups may contribute to higher than average levels of dementia knowledge being reported. Our sample is therefore a limitation; however, social media platforms such as Facebook, Instagram and Twitter enable the recruitment of more representative samples than traditional convenience sample approaches (e.g. flyers on community noticeboards, advertisements in newspapers and university participant pools). However, our sample is not population-based and does not reflect the true population structure of Australia; that was beyond this study.
The second limitation of this study is that gender and vignette symptom severity were not counterbalanced. Therefore, it may not be more severe symptoms that respondents were responding to but instead were responding based on cultural and stereotypical expectations of females having dementia in late-life. This could have contributed to the higher number of correct responses in those allocated to the moderate symptom (Mary) group. However, using the same vignettes from Low and Anstey (2009) allowed for the direct comparison of dementia recognition between the two studies. Future studies should use a larger variety of vignettes, counterbalancing factors such as gender and symptom severity. It would also be interesting to include risk factors in vignettes, to assess if the public intuitively link these with the onset and progression of dementia symptomatology.
The third limitation of this study is the complexity surrounding the identification of dementia risk factors. For example, Livingston et al. (2017) consider smoking in late-life to be a risk factor for dementia, but not during midlife. Similarly, Livingston et al. (2017) identify and provide evidence for the contribution of midlife hearing loss and social isolation to dementia, whereas Norton et al. (2014) do not. Furthermore, neither study identify head injury or stroke as risk factors for dementia, where emerging research has provided evidence to support a link between head injury and stroke and the increased risk of dementia (Fann et al., 2018; Kuźma et al., 2018; Savva & Stephan, 2010). Notably, the study by Livingston et al. (2017) was updated recently (Livingston et al., 2020; published after this manuscript was first submitted), and head injury is now included as a modifiable dementia risk factor.
Dementia is and will remain a major public health concern for the foreseeable future. These findings show that the public are generally able to recognise dementia but have limited conceptual understanding of dementia. The public’s knowledge of empirically supported modifiable risk factors was limited, with most risk factors being viewed as likely contributors to dementia at less than chance levels. In order to reduce the burden of dementia, it is important that knowledge of such risk factors is increased through public health campaigns.
Supplemental Material
sj-pdf-1-dem-10.1177_1471301221997301 – Supplemental Material for What do the public really know about dementia and its risk factors?
Supplemental Material, sj-pdf-1-dem-10.1177_1471301221997301 for What do the public really know about dementia and its risk factors? by Alana K Nagel, Tobias Loetscher, Ashleigh E Smith and Hannah AD Keage in Dementia
Footnotes
Author Contributions
AN conceptualised the study design, set up and monitored the online survey, performed the statistical analyses and interpretation of the results and drafted the manuscript for publication. HADK assisted in the conceptualisation of the study design, supervised the study, assisted with the statistical analyses and interpretation of the results and reviewed drafts of the manuscript. AES and TL provided feedback during the conceptualisation of the study design and reviewed drafts of the manuscript. All authors approved the final document.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This study was funded by the NHMRC Dementia Research Leadership Fellowship funding to HADK (GNT1135676) and TL (GNT1136269) and ARC-NHMRC Dementia Research Fellowship (GNT1097397).
Ethical Approval
The current study was granted ethics approval by the University of South Australia’s Human Research Ethics Committees in accordance with the national guidelines (approval number 201089).
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
