Health Advice in a Digital World: Quality and Content of Online Information about the Prevention of Alzheimer’s Disease

Abstract

Background: As the number of older adults turning to the Internet for health information increases, so does the potential for online information to have a substantial impact on the patient-physician relationship and on their health. Inaccurate information may weaken patient-physician relationships or result in increased physician visits and health anxiety, while high quality information may allow Internet users to make better decisions about their health.

Objective: To assess the quality and content of available online resources about the prevention of Alzheimer’s disease (AD).

Methods: A sample of 308 articles related to the prevention of AD was collected from the first three pages of location-independent keyword searches on Google.com between September 17–30, 2014. Content analysis was applied to articles that met criteria (n = 298) and a quality evaluation tool was developed to generate a quality score for each of the articles (n = 290).

Results: We found that articles on the high end of the quality spectrum focused on modifiable risk factors and tended to present balanced information, while articles of low quality emphasized nutrition as a method of prevention and were more likely to be in conflict of interest.

Conclusion: This study provides the first insight into the content and quality of prevention information for AD currently available online and highlights the importance of future research to better understand the impact of this information on the patient-physician relationship and health decision-making of older adults.

Keywords

Alzheimer’s disease consumer health information Internet prevention

INTRODUCTION

The Internet is rapidly reshaping healthcare, and 72% of Internet users report searching for health information online [1]. Widely accessible online resources act as a new variable in the patient-physician relationship, and one study found 91% of physicians surveyed had patients who made inquiries about Internet health information in the last year [2].

The increase in the use of the Internet as a source of health-related information may allow for the rapid dissemination of information and can be beneficial to Internet users by allowing them to become better informed about their health and promoting their independence and autonomy [2]. In addition, online health information may help users feel more empowered about their health decisions, and may lead to increased social support as users are able to interact with others experiencing similar situations in online discussion groups [3 –5].

Alongside these benefits, however, are potential harms: inaccurate information may weaken patient-physician relationships and result in inappropriate use of health services, such as an increased number of physician visits and requests for unnecessary tests or treatment [2, 6]. Online resources may also contribute to excessive health-related anxiety [7].

Contrary to popular belief, the increase in the popularity of the Internet as a source of health information is not limited to younger generations; recent survey data show that the number of older adults spending time online has been rapidly increasing over the past decade, and six in ten adults aged 65 and older currently report using the Internet [8]. Among older Internet users, over half report using the Internet to retrieve health information, with most of these inquiries starting from a search engine [9]. As older adults are less likely to respond to credibility cues of webpages compared to young adults, they are especially vulnerable to inaccurate information [10].

As the number of individuals of all ages seeking online health information increases, so does the potential for this information to have a large impact on health decision-making in an aging population, especially in the context of the rising tide of dementia [11], one of the most feared conditions for older adults. Studies have shown that 92% of respondents with parents who have been diagnosed with Alzheimer’s disease (AD) are concerned about developing it themselves, and even among older adults with no family history of dementia, 47% are concerned about AD [13]. It is estimated that there were 44 million people with dementia worldwide in 2014, and that this number will more than triple by 2050 [12]. The estimated worldwide costs of dementia in 2010 were $604 billion US, and this number is also expected to rise with the increasing prevalence of dementia. As there are currently no disease-modifying therapies available for AD, the most prevalent form of dementia, there is growing emphasis on the importance of lifestyle modifications as way to potentially prevent or delay the onset of the disease [14]. Dissemination of high-quality information about the prevention of AD is therefore critical, and the online environment plays an important role in this process.

Previous work from our group has demonstrated that an extensive conversation about dementia is taking place online, and that websites hosting dementia-related information receive up to several million visitors per month [15, 16]. As there are many popular misconceptions regarding the prevention of AD, the Internet has the potential to be a powerful positive communication tool to clarify these myths and misunderstandings [17]. However, the Internet’s ease of use and broad reach could also serve to catalyze the spread of misinformation, and concerns have been raised over the quality of the information being communicated in the online environment [15, 18]. Previous research evaluating online health information has shown that the quality of information varies greatly across websites [19]. In a meta-analysis evaluating online health information on a variety of topics, 70% of the studies used indicators of quality to assess websites such as authorship, attribution, currency, readability, and comprehensiveness and concluded that low quality was a salient concern with online health information [18]. Bonnar-Kidd and colleagues examined websites related to physical activity using similar benchmarks and found that overall, 78% of the articles were rated as low quality [19]. For example, fewer than half of the sites displayed the authorship and currency of theinformation [19].

Whether online information is beneficial or detrimental to the health of Internet users hinges largely on the quality and the content of the available information. However, to date, there has been little research on the content and quality of freely available information about AD, and none on the prevention of AD specifically. We addressed this knowledge gap with the present study.

METHODS

Sample

Websites containing articles of interest were retrieved by conducting a search on Google, the most popular search engine in the USA at the time of the study [20]. Google.com/ncr (no country redirect) was used to maintain a location-disabled search to avoid localized results [21]. In addition, browsing history and cookies were cleared before each search, and searches were conducted on an application that prevents the tracking of browsing history and cookies during the search. The articles were retrieved using 105 different combinations of search terms related to AD prevention. The keywords used were: Alzheimer, Alzheimer’s, Alzheimers, prevent, prevention, avoid, avoidance, risk, advice, lifestyle, reduce, steps, tips, tricks, ways. We limited the search to the first three pages of results, as evidence suggests most Internet users do not look beyond three pages when using search engines [22]. During the article collection period, each article was entered into a database and briefly reviewed for an initial screening of relevance, and search terms were refined. After the articles were collected, they were subject to more rigorous inclusion criteria for the articles: 1) the article is in the English language; 2) no payment or login is required to access the article; and 3) prevention of AD is the main focus of the article as determined by the content of the headline and lead paragraph.

Quality analysis

We developed a customized quality evaluation tool to assess the articles collected based on previous tools by Chumber and colleagues [23] and Sandvik [24] and following recommendations by Silberg and colleagues [25], commonly referred to as the “JAMA benchmarks”. Each article was evaluated based on seven criteria informed by the above-mentioned studies, producing a numerical score for each criterion. The criteria included: authorship, attribution, type of study, conflict of interest, currency, complementarity, and tone. The individual scores were weighted and summed to produce a total score for each article out of 28, with each score’s contribution to the total score based on its relative importance (Table 1).

One investigator (TF) applied the quality evaluation tool to the entire sample, while a second coder (EC) applied the tool to 20% of the articles to ensure inter-rater agreement. Agreement was initially 93%; disagreements were resolved through discussion.

Content analysis

To compress the extensive amount of text into more meaningful categories, content analysis was rigorously and systematically applied to the sample. The analysis was based on explicit rules established in an emergent coding guide, which was developed by two coders (JR and TF) from a pilot analysis of a random sample of 10% of the articles retrieved which met the inclusion criteria. The final coding guide encompassed both the features and the content of the article, and was comprised of the following major categories: a) quality indicators (e.g., author, date, product endorsement), b) article features (e.g., tone, evidence listed), and c) action items (e.g., advice; Supplementary Table 1). The unit of analysis was defined as each individual article, and a rich coding strategy allowed for multiple categorizations of individual articles. The contents of the website outside of the text of the article were not included in the analysis. The entire sample was coded for content by one investigator (TF), and a second coder (JR) coded 20% of the articles to ensure inter-rater reliability. Reproducibility was initially 98% and remaining disagreements were settled through discussion to achieve consensus. Descriptive statistics were used to characterize the sample: for each category, a percentage of the total sample it was identified in was calculated.

To test relationships within the data set, Mann-Whitney tests were performed [26].

RESULTS

Sample

A total of 397 unique articles were retrieved from the period of September 17–30, 2014, with 308 of the articles meeting criteria for inclusion in the analysis. To assess the longevity of the sample, the search strategy was revisited at a second time point (November 5, 2014–March 10, 2015). Ten articles were removed as the links no longer redirected to the correct site at the later time point, which resulted in a final sample of 298 units of analysis for the content analysis portion of the study (Supplementary Table 2). The final sample consisted of 290 articles, as a further eight of the original 298 articles could no longer be accessed at a third time point between May 22–27, 2015 for quality analysis.

Overall quality

Articles fell on the entire spectrum of quality, ranging from very poor to excellent (Fig. 1). The boxed areas on Fig. 1 containing the two tails of the score distribution represent the portion of the sample with the best scores (score greater than 23/28, n = 31) and with the worst scores (score lower than 12/28, n = 34).

Quality indicators

Articles about the prevention of AD were found on a variety of different types of websites as determined by their “About” pages: 24% of articles were hosted on health information websites, 24% on news sites, 7% on advocacy group websites, 6% on science or medical news sites, and 4% on government sites (Supplementary Figure 1; for examples of each type of website, refer to Supplementary Table 3). The lowest scoring articles were more likely to come from health (35%) and lifestyle websites (15%), while the highest scoring articles were more likely to come from general news sites (29%), science or medical news websites (13%), or government websites (6%; Fig. 2A). Overall, 49% of the articles were hosted on websites that sold a product, service, subscription, or membership. Among the highest scoring articles and lowest scoring articles, 45% and 62% of articles, respectively, came from commercial websites. Higher scoring articles were more likely to have a neutral headline (42%), mention research in the headline (23%), or include a headline in the form of a question (16%), while lower scoring articles were more likely to display an optimistic headline (36%), a headline in the form of a list (24%), or a directive headline (15%; Fig. 2B; for examples of each type of headline, refer to Supplementary Table 3). Most articles in the sample included a date of creation or update (76%), and 68% of the entire sample of articles were written or updated in the past five years. The highest scoring articles nearly always displayed a date (97%), compared to 59% of the lowest scoring articles. Two thirds (67%) of the complete sample of articles displayed authorship, with 19% of those including author credentials (for examples of types of author, refer to Supplementary Table 3). 94% of the highest scoring articles and 53% of the lowest scoring articles indicated an author. Of the entire sample, 20% of articles recommended talking to a physician about the information, with 10% of the top scoring articles and 28% of the lowest scoring articles recommended doing so (for examples, refer to Supplementary Table 3). Nineteen percent of all the articles endorsed a product or service within the text of the article (for examples of product endorsements, refer to Supplementary Table 3). None of the highest scoring articles contained such an endorsement, while half (50%) of the lowest scoring ones did show a conflict of interest.

Sixty percent of articles in the whole sample cited general research as evidence for their claims (Fig. 2B). For example: “Research suggests that regular aerobic exercise can reduce your risk for dementia by 40 to 50 percent” [27].

The “attribution” criteria of the quality evaluation tool referred to how articles supported their claims about the prevention of AD. Fifty-eight percent of the articles supported their claims with a specific, traceable study (e.g., by disclosing the name of the authors or of the journal): “According to a new study to be published in the Journal of Alzheimer’s Disease, drinking around three cups of coffee every day may help older people avoid the onset ofAlzheimer’s” [28].

Articles also used quotes or information from a doctor (30%) or from a professor (14%) to support their claims: “Dr. Lipton suggests following a Mediterranean-style diet full of fish, fruits, and vegetables loaded with antioxidants, whole grains, and olive oil,” [29] while other articles relied on unsupported biological rationale (27%), a description of physiological processes with no reference to a specific study: “Stress is like poison for your hippocampus – the part of brain responsible for memory – as it leads to chronically high levels of the hormone cortisol. High cortisol levels can [ ... ] kill brain cells [ ... ]. Busting stress, therefore, is vital to preserve your memory” [30].

Other sources of evidence were reference lists (25%), links to webpages (19%), reference to an advocacy or government body (17%), a conference (8%), or a book (6%; (for examples of each, refer to Supplementary Table 3). Figure 2B illustrates that most of the best-scoring articles cited a specific study as evidence, while the lowest-scoring articles tended to use general research or unsupported biological rationales to support theirclaims.

When evaluating the tone of the recommendations contained in the sample of articles, we found that 22% of articles made strong recommendations (Fig. 2C): “If you’re serious about the prevention of Alzheimer’s and improving memory loss, you should definitely take a high potency multiple vitamin and mineral capsule” [30]. Sixty three percent of articles made milder recommendations, for example: “Findings indicate that an active social life and strong network of friends may help prevent Alzheimer’s in later life” [31]. 13% presented balanced advice: “Although some research has shown that light to moderate alcohol use may protect against dementia, the results aren’t clear” [32]. Finally, one percent of articles did not provide any advice: “A comprehensive analysis by an independent government panel has found that there is not enough scientific evidence to date to support the advice doctors currently give — such as exercising, doing crossword puzzles or eating a Mediterranean-style diet — for preventing or controlling symptoms of Alzheimer’s disease and dementia” [33].

When characterizing high- and low-quality articles, we found that 58% of the top scoring articles offered balanced advice, while 79% of the lowest scoring articles made strong recommendations (Fig. 2C).

Type of advice

Overall, 84% of all the articles discussed nutrition, 78% addressed lifestyle factors, 73% promoted exercise, and 66% emphasized cognitive engagement (Table 2).

Nutrition advice broadly was found in 65% and 97% of the highest- and lowest-scoring articles, respectively, lifestyle in 77% and 88%, exercise in 65% and 71%, and cognitive engagement in 68% and 65%. Regarding nutrition, the lowest-scoring articles were more likely to recommend eating nuts and getting more antioxidants, while the highest-scoring articles were more likely to recommend foods rich in whole grains and vitamin E, and avoiding refined carbs (Fig. 3A). For lifestyle factors, the lowest-scoring articles were more likely to endorse decreasing stress, meditating, and avoiding aluminum, while a greater proportion of the highest-scoring articles endorsed quitting smoking, and controlling cases of diabetes and diabetes if these are present (Fig. 3B). Of the highest- and lowest-scoring articles that mentioned exercise, more of the highest-scoring articles recommended general exercise, while more lowest-scoring articles specified aerobic exercise (Fig. 3C). In terms of cognitive engagement, the lowest-scoring articles were more likely to discuss learning new things and using a computer, while the highest-scoring articles were more likely to discuss higher education and doing puzzles and board games (Fig. 3D). In the supplements category, 10% of the highest-scoring articles recommended taking vitamin E supplements and folic acid, 6% recommended vitamin B12 supplements, and none recommended taking fish oil. Among the lowest-scoring articles, 13% recommended vitamin E, 16% recommended vitamin B12, 13% recommended fish oil, and 6% mentioned folic acid. The most frequently discussed complementary and alternative interventions as defined by the National Institutes of Health [34] were the Mediterranean diet (16%), folic acid (10%), and vitamin E supplements (10%) for the highest-scoring articles and the Mediterranean diet (21%), turmeric or curcumin (15%), and vitamin E supplements (15%) for the lowest-scoring articles.

Two-tailed Mann-Whitney tests were performed to determine whether the highest- and lowest-scoring articles differed significantly in the frequency with which they mentioned advice in a specific category. According to the tests, the lowest-scoring articles were significantly more likely to discuss advice pertaining to nutrition (U = 1180, p < 0.05) and complementary and alternative medicine (U = 32, p < 0.05).

DISCUSSION

The results from this study of online articles about AD prevention show that while a wide variety of articles about the prevention of AD are accessible on the Internet, the quality of these articles varies greatly, ranging from articles with unsupported claims and conflicts of interests to high quality articles based on recent, peer-reviewed evidence. When characterizing the high quality articles and the low quality articles, we found that: 1) different types of evidence are used to support claims in high and low quality articles, 2) lower quality articles tend to make stronger recommendations, 3) lower quality articles are more likely to endorse a product or service, and 4) articles of lower quality emphasize nutrition as a method of prevention.

We appreciate the limitations of the present study. The article retrieval was performed over the course of two weeks, and due to the dynamic nature of the Internet, articles may change. Also, while Google was chosen as the search engine based on its popularity, and a preliminary analysis showed no new websites using other search engines such as Bing, the possibility remains that other sites may have been uncovered through different search engines. Despite these limitations, the results show that the information about the prevention of AD Internet to which consumers are exposed is highly variable in quality, binary at the tails, and, in the case of lower quality articles, potentially harmful due to conflicts of interest and unbalanced advice. Since there is yet to be a reliable cure for AD, the focus on prevention is of utmost importance [14].

Research has shown that online information on preventive health, such as the focus of this study, tends to be of lower quality than information on the diagnosis and treatment of injuries and illnesses [35]. Though there were subtle differences in the specific characteristics of different types of online health information in our sample, the findings confirm the overarching theme that the quality of online health information is heterogeneous in nature. However, there are marked differences between articles that scored high on the quality spectrum and those that scored low. Our results also show that in general, lower quality articles placed more emphasis on nutrition, while higher quality articles tended to discussmodifiable risk factors. These differences may be explained by the fact that nutritional changes are perceived as easier to implement than lifestyle modifications such as controlling one’s weight [36]. Top scoring articles did not endorse products or services, while over half of the low scoring articles did, which constitutes a clear conflict of interest in the presentation of seemingly unbiased information. Similarly, a study on online weight loss information also found that commercial websites ranked lowest in terms of content quality [37]. Beyond commercial interests, we found that higher scoring articles tend to be more cautious in the tone of their advice, while lower scoring articles use strong vocabulary to support their claims. These findings are consistent with a study evaluating the quality of online information on the MMR vaccine that found that articles that took an anti-vaccination stance were much more likely to provide unbalanced information [38]. In addition, our results showed that high quality articles were more likely to cite traceable studies to support their claims, as seen in a previous study of the quality of online articles on the treatment of depression [39].

The wide availability of online health information is dynamically changing the nature of the patient-physician relationship [46]. When provided with high quality information, patients are better informed and require less of their physician’s time to have a new diagnosis or surgery procedure explained to them [47]. On the contrary, discussing poor quality online information at physician visits may be time consuming and increase the length of consultations [3]. Online health information may also result in requests for unnecessary tests or treatment, which can lead to decreased patient satisfaction when physicians refuse them [3, 6]. In addition, some physicians felt their authority was challenged by patients who sought health information online [3, 6]. Not only can widely accessible articles with low quality health information be detrimental to the patient-physician relationship, they can also be dangerous to Internet users’ health. A study conducted by Walji and colleagues showed that 25% of websites collected regarding three popular herbal supplements contain recommendations that could directly lead to physical harm [48]. In another example, a patient with maxillary sinus cancer died from complications caused by self-medication with hydrazine sulphate obtained from an alternative medicine website [49]. Low quality online health information can also lead to financial expenditures on complementary and alternative medicine that is based on little to no scientific evidence. In 2012, the expenditure on alternative therapies in the USA was estimated to be $13.9 billion [50]. The high use of alternative therapies is a salient issue for AD, as a study conducted by Palmour and colleagues on dietary supplements for AD in websites found that only 16% of the sites provide peer-reviewed, scientific support for their claims of benefit. Their study and others highlight the need for better regulation on the online marketing of dietary supplements [51].

Many of the potential harms associated with low quality online health information can be at least partially addressed through effective education. Moving forward, health care professionals who interact with older adults, their families and caregivers ought to consider the following recommendations for discussing online health information:

Develop an awareness of the information available online, such as the latest news stories related to AD research, to anticipate and prepare for upcoming discussions;

Educate members of the patient community about the hallmarks of high quality online health information such as a balanced tone, traceable evidence, and an emphasis on sustainable and long-lasting lifestyle changes over “quick fixes”;

Suggest varied sources of high quality online information;

Advise advocacy groups or government associations if you feel information provided on their websites is wrong or outdated.

Effective engagement and evidence-based education about online resources related to the prevention AD are critical to ensure that the benefits of these resources are maximized and promote positive health decision-making in older adults.

Footnotes

ACKNOWLEDGMENTS

We gratefully acknowledge Dr. Judy Illes, PhD, FRSC, FCAHS, Professor of Neurology and Director, National Core for Neuroethics at the University of British Columbia and Canada Research Chair in Neuroethics, for providing insightful comments on the manuscript, and Emanual Cabral, BA, undergraduate research assistant at the National Core for Neuroethics, for his assistance in developing the coding guide and his role as a second coder.

This study was funded by the Canadian Institutes for Health Research, the British Columbia Knowledge Development Fund, the Canadian Foundation for Innovation and the Vancouver Coastal Health Research Institute.

Authors’ disclosures available online ().

Appendix

References

Pew Research Center: Internet, Science & Tech (2013) Health Online 2013, Available at http://www.pewinternet.org/2013/01/15/health-online-2013/

Kim

, Kim

(2009) Physicians’ perception of the effects of internet health information on the doctor–patient relationship. Inform Health Soc Care 34, 136–148.

van Uden-Kraan

, Drossaert

CHC

, Taal

, Smit

, Seydel

, van de Laar

MAFJ

(2010) Experiences and attitudes of Dutch rheumatologists and oncologists with regard to their patients’ health-related Internet use. Clin Rheumatol 29, 1229–1236.

van Uden-Kraan

, Drossaert

CHC

, Taal

, Seydel

, van de Laar

MAFJ

(2009) Participation in online patient support groups endorses patients’ empowerment. Patient Educ Couns 74, 61–69.

Idriss

, Kvedar

, Watson

(2009) The role of online support communities: Benefits of expanded social networks to patients with psoriasis. Arch Dermatol 145, 46–51.

Murray

, Lo

, Pollack

, Donelan

, Catania

, Lee

, Zapert

, Turner

(2003) The impact of health information on the Internet on the physician-patient relationship: Patient perceptions. Arch Intern Med 163, 1727–1734.

Baumgartner

, Hartmann

(2011) The role of health anxiety in online health information search. CyberPsychology Behav Soc Netw 14, 613–618.

Smith

(2014) Older adults and technology use. Pew Research Center, Available at: http://www.pewinternet.org/2014/04/03/older-adults-and-technology-use/

Levy

, Janke

, Langa

(2014) Health literacy and the digital divide among older Americans. J Gen Intern Med 30, 284–289.

10.

Liao

, Fu

(2014) Age differences in credibility judgments of online health information. ACM Trans Comput-Hum Interact 21, 2:1–2:23.

11.

Dudgeon

(2010) Rising tide: The impact of dementia on Canadian society, Alzheimer Society of Canada, Toronto.

12.

Prince

, Albanese

, Guerchet

, Prina

(2014) World Alzheimer’s Report, Alzheimer’s Disease International, London.

13.

Cutler

(2001) Correlates of personal concerns about developing Alzheimer’s disease among middle-aged persons. Am J Alzheimers Dis Other Demen 16, 335–343.

14.

Imtiaz

, Tolppanen

A-M

, Kivipelto

, Soininen

(2014) Future directions in Alzheimer’s disease from risk factors to prevention. Biochem Pharmacol 88, 661–670.

15.

Robillard

, Johnson

, Hennessey

, Beattie

, Illes

(2013) Aging 2.0: Health information about dementia on Twitter. PLoS One 8, 1–5.

16.

Robillard

, Illes

, Arcand

, Beattie

, Hayden

, Lawrence

, McGrenere

, Reiner

, Wittenberg

, Jacova

(2015) Scientific and ethical features of English-language online tests for Alzheimer’s disease. Alzheimers Dement (Amst) 1, 281–288.

17.

Lowe

, Balsis

, Hughes

, Shine

, Carpenter

(2015) Misconceptions of Alzheimer’s disease. Clin Gerontol 38, 149–156.

18.

Eysenbach

, Powell

, Kuss

, Sa

(2002) Empirical studies assessing the quality of health information for consumers on the world wide web: A systematic review. JAMA 287, 2691–2700.

19.

Bonnar-Kidd

, Black

, Mattson

, Coster

(2009) Online physical activity information: Will typical users find quality information? Health Commun 24, 165–175.

20.

Purcell

, Brenner

, Rainie

(2012) Search engine use 2012. Pew Research Center, Available at: http://www.pewinternet.org/2012/03/09/search-engine-use-2012/

21.

Kent (2014) How to always use Google.com and stop Google from sending you to country specific website [Tip].available at:. DotTech,–https://dottech.org/158839/prevent-google-from-redirecting-you-to-its-local-homepage-tip/.

22.

Beitzel

, Jensen

, Chowdhury

, Frieder

, Grossman

(2007) Temporal analysis of a very large topically categorized web query log. J Am Soc Inf Sci Technol 58, 166–178.

23.

Chumber

, Huber

, Ghezzi

(2015) A methodology to analyze the quality of health information on the Internet: The example of diabetic neuropathy. Diabetes Educ 41, 95–105.

24.

Sandvik

(1999) Health information and interaction on the Internet: A survey of female urinary incontinence. BMJ 319, 29–32.

25.

Silberg

, Lundberg

, Musacchio

(1997) Assessing, controlling, and assuring the quality of medical information on the Internet: Caveant lector et viewor— let the reader and viewer beware. JAMA 277, 1244–1245.

26.

Zaiontz

(2015), Real Statistics Resource Pack.

27.

Eckelkamp

(2013) You can prevent Alzheimer’s disease, Posted 24 July 2013, Prevention, http://www.prevention.com/health/brain-health/you-can-prevent-alzheimers-disease Accessed 10 July 2015.

28.

Munzenrieder

(2012) UM Researchers: Caffeine May Help You Avoid Alzheimer’s, Miami New Times, http://www.miaminewtimes.com/news/um-researchers-caffeine-may-help-you-avoid-alzheimers-6547451, Posted 5 June 2012, Accessed 10 July 2015.

29.

Graham

(2014) 5 things you can do now to avoid Alzheimer’s disease, Good Housekeeping, http://www.goodhousekeeping.com/health/womens-health/prevent-alzheimers-disease, Posted 28 July 2014, Accessed on 10 July 2015.

30.

5-step Alzheimer’s prevention plan, the Dr. Oz Show, http://www.doctoroz.com/article/5-step-alzheimers-prevention-plan, Posted 30 May 2012, Accessed 10 July 2015.

31.

8 ways to prevent Alzheimer’s disease, Prevention, http://www.prevention.com/health/health-concerns/8-ways-prevent-alzheimers-disease, Posted 5 June 2013, Accessed 10 July 2015.

32.

Prevention of Alzheimer’s Disease, Web MD, http://www.webmd.com/alzheimers/guide/understanding-alzheimers-disease-prevention, Posted 25 April 2015, Accessed 22 September, 2015.

33.

Park

(2010) Study: Lifestyle May Not Prevent Alzheimer’s, Time, http://content.time.com/time/health/article/0,8599,1996235,00.html, Posted 14 June 2010, Accessed 22 September 2015.

34.

NIH>, National Center for Complementary and Integrative Health (2008) Complementary, alternative, or integrative health: What’s in a name? Available at: https://nccih.nih.gov/health/integrative-health. Date Created: October 2008, Accessed 11 May 2016.

35.

Kitchens

, Harle

, Li

(2014) Quality of health-related online search results. Decis Support Syst 57, 454–462.

36.

Barrie

, Jones

(2011) “As long as you have some of that, it cancels it out”: How advertisers use guilt to sell us quick-fixes for our unhealthy behaviour.pp. Australian and New Zealand Marketing Association Conference (ANZMAC) 1–6.

37.

Modave

, Shokar

, Peñaranda

, Nguyen

(2014) Analysis of the accuracy of weight loss information search engine results on the Internet. Am J Public Health 104, 1971–1978.

38.

Abbott

(2000) Web page quality: Can we measure it and what do we find? A report of exploratory findings. J Public Health 22, 191–197.

39.

Griffiths

, Christensen

(2000) Quality of web based information on treatment of depression: Cross sectional survey. BMJ 321, 1511–1515.

40.

Charnock

, Shepperd

, Needham

, Gann

(1999) DISCERN: An instrument for judging the quality of written consumer health information on treatment choices. J Epidemiol Community Health 53, 105–111.

41.

Boyer

, Baujard

, Geissbuhler

(2011) Evolution of health web certification through the HONcode experience. Stud Health Technol Inform 169, 53–57.

42.

Minervation (2007) LIDA tool: The Minervation validation instrument for healthcare websites.

43.

Morr

, Shanti

, Carrer

, Kubeck

, Gerling

(2010) Quality of information concerning cervical disc herniation on the Internet. Spine J 10, 350–354.

44.

Langille

, Veldhuyzen van Zanten

, Shanavaz

, Massoud

(2012) Systematic evaluation of obstructive sleep apnea websites on the internet. J Otolaryngol Head Neck Surg 41, 265–272.

45.

van der Marel

, Duijvestein

, Hardwick

, van den Brink

, Veenendaal

, Hommes

, Fidder

(2009) Quality of web-based information on inflammatory bowel diseases. Inflamm Bowel Dis 15, 1891–1896.

46.

McMullan

(2006) Patients using the Internet to obtain health information: How this affects the patient–health professional relationship. Patient Educ Couns 63, 24–28.

47.

Ferguson

(2000) Online patient-helpers and physicians working together: A new partnership for high quality health care. BMJ 321, 1129–1132.

48.

Walji

, Sagaram

, Meric-Bernstam

, Johnson

, Mirza

, Bernstam

(2004) Efficacy of quality criteria to identify potentially harmful information: A cross-sectional survey of complementary and alternative medicine web sites. J Med Internet Res 6, e21.

49.

Hainer

, Tsai

, Komura

, Chiu

(2000) Fatal hepatorenal failure associated with hydrazine sulfate. Ann Intern Med 133, 877–880.

50.

Davis

, Weeks

(2012) The concentration of out-of-pocket expenditures on complementary and alternative medicine in the United States. Altern Ther Health Med 18, 36–42.

51.

Palmour

, Vanderbyl

, Zimmerman

, Gauthier

, Racine

(2013) Alzheimer’s disease dietary supplements in websites. HEC Forum 25, 361–382.