Body mass index and the use of the Internet for health information

Abstract

Objective:

Individuals who experience or anticipate negative interactions from medical providers related to conditions such as obesity may preferentially use the Internet for health information. Our objectives in this study were to (1) examine the association between body mass index (BMI) and Internet health information–seeking and (2) examine whether the association between patient-centred communication and Internet health information–seeking differed by BMI category.

Method:

We used data from the Health Information National Trends Survey 3 (n = 7,674), a cross-sectional nationally representative survey of US adults. Using weighted data, we determined associations between BMI category and Internet health information–seeking, adjusting for sociodemographic and health-related variables. We then determined the adjusted associations between patient-centred communication and Internet health information–seeking stratified by BMI category.

Results:

Most (78.7%) sought health information from the Internet. The odds of Internet health information–seeking among individuals with overweight (adjusted odds ratio [AOR] = 0.96) and individuals with obesity (AOR = 1.45) did not differ significantly from individuals with normal weight. The association between perceived levels of patient-centred communication and using the Internet for health information did not differ by BMI category.

Conclusion:

Study findings suggest that most individuals seek health information from the Internet. Despite non-significant results, this study points to the need for research on health information choices specific to weight loss among individuals with overweight and obesity using more comprehensive measures of the frequency, preference and motivation for information-seeking.

Keywords

Consumer health information health information–seeking Internet obesity patient–provider communication

Introduction

Health information–seeking has been associated with a variety of behavioural outcomes, including changes in diet (Brodie et al., 1999; Freisling et al., 2009; Lewis et al., 2012), physical activity (Brodie et al., 1999) and cancer screening and cancer prevention behaviours (Shim et al., 2006). The Internet is becoming an increasingly utilised source of health information. Nationally representative studies in the USA suggest that Internet health information–seeking has increased over time (Baker et al., 2003; Ybarra and Suman, 2006), with approximately 50%–60% of US adults seeking health information from the Internet within the past 12 months (Koch-Weser et al., 2010; Ybarra and Suman, 2006).

Certain populations are more likely to use the Internet for health information. Internet health information–seekers are more likely to be female, white and young; have high levels of education; have high levels of income and have current health problems than those who do not seek health information from the Internet (Koch-Weser et al., 2010; Renahy et al., 2008; Ybarra and Suman, 2006). The Internet has also been shown to be an especially attractive source of health information for individuals with stigmatised conditions, primarily due to the ease and anonymity permitted during the information search (Berger et al., 2005; Gallagher and Doherty, 2009; Powell et al., 2003).

Obesity is a highly stigmatised health condition. People with overweight and obesity have reported stigma, negative interactions and discomfort in medical settings, including embarrassment about being weighed, fear of disrespectful treatment and negative attitudes from providers (Amy et al., 2006); concerns with disrobing and embarrassment about body weight (Olson et al., 1994) and embarrassment and discomfort with physicians (Wee et al., 2000). Patients with obesity are viewed by health professionals as possessing poor motivation (Kristeller and Hoerr, 1997) and as noncompliant and unattractive (Bocquier et al., 2005; Foster et al., 2003). Negative attitudes towards obesity and towards patients with overweight and obesity have been suggested in numerous studies involving healthcare professionals, including physicians, nurses and registered dietitians, as well as among students studying to be health professionals (Bagley et al., 1989; Berryman et al., 2006; Brown and Thompson, 2007; Chambliss et al., 2004; Hare et al., 2000; Hoppe and Ogden, 1997; McArthur and Ross, 1997; Maddox and Liederman, 1969; Maroney and Golub, 1992; Wear et al., 2006). In contrast, individuals with overweight and obesity have reported finding the Internet a safe, non-judgmental space for support and information about health and weight loss (Lewis et al., 2010).

Concerns regarding negative interactions with providers may lead to reluctance among patients with overweight and obesity to seek health information from their providers. One study determined that nearly half (45%) of participants with obesity reported that they did not rely ‘at all’ on their physicians for help with weight control, and an additional 30% reported that they relied on their physicians only a ‘slight amount’ (Wadden et al., 2000). Instead, individuals with overweight and obesity may preferentially turn to a competing source of health information such as the Internet. Although weight bias has been demonstrated on online discussion boards in a recent study (De Brun et al., 2014), the Internet may provide a space where individuals avoid judgment based on physical appearance (McKenna and Bargh, 2000) and may discuss and disclose health-related information on the Internet with less risk than would occur in face-to-face contexts (Wright and Bell, 2003).

Uses and Gratifications Theory (Katz et al., 1973a) provides a theoretical basis for understanding this behaviour. The theory assumes that individuals are active information-seekers who select from competing information sources (e.g. providers, family members, friends, the Internet or other media sources) to fulfil their perceived unmet needs (Katz et al., 1973a, 1973b). In the context of this theory, if health professionals are unable to fulfil the information needs of patients, competing sources of information such as the Internet may be preferentially utilised. Indeed, studies have suggested that patients who are not satisfied with their interactions with service providers may be more likely to turn to the Internet for health information. Internet health information–seeking has been associated with time constraints with providers (Anderson, 2004), the perception of the need for more information or unanswered questions following a visit with providers (Anderson, 2004; Rice, 2006), perceptions of lower levels of patient-centred communication with providers (Hou and Shim, 2010) and lower levels of empathy and quality of time spent with providers (Tustin, 2010).

Because the Internet may be a preferred source of health information for those with stigmatised conditions such as obesity, body mass index (BMI) may be an important predictor of Internet health information–seeking behaviour. Additionally, patient–provider relationship dynamics may impact on health information–seeking patterns (Anderson, 2004; Hou and Shim, 2010; Rice, 2006; Tustin, 2010), and patient–provider interactions and communication may differ for individuals with overweight and obesity compared to other individuals. Therefore, BMI may be an effect modifier that impacts the association between patient–provider communication (including levels of patient-centred communication) and Internet health information–seeking.

In the USA, few studies have examined the relationship between Internet health information–seeking behaviours and BMI, especially using nationally representative data. The purpose of this study was therefore to examine (1) BMI as a possible predictor of Internet health information–seeking and (2) whether the association between patient-centred communication and Internet health information–seeking differs between BMI categories. We hypothesise that BMI category will be a significant predictor of Internet health information–seeking, with individuals with obesity having significantly higher odds of Internet information–seeking than individuals without obesity. Our second hypothesis is that the association between patient-centred communication and Internet information–seeking will be different for individuals with obesity than for individuals without obesity. The results of this study will provide a greater understanding of Internet use for health information among individuals of different BMI categories and how using this source of health information might be associated with characteristics of patient–provider communication.

Methods

Data source

The data are from the Health Information National Trends Survey (HINTS). HINTS is a nationally representative survey that is conducted approximately every 2 years by the National Cancer Institute. The main objective of HINTS is to examine the ‘health information environment’ and its relationship with health behaviours. In this study, we used data from HINTS 3 because this iteration of the survey included several items related to patient–provider communication. HINTS 3 data consist of self-reported responses to 117 questions that were primarily closed-ended. The data were collected between January 2008 and April 2008 by computer-assisted telephone interview and by mail. Participants were selected using random-digit dialling for the telephone component and using random sampling of households from the US Postal Service for the mail component. Census blocks with a higher proportion of minorities were oversampled. The final sample for HINTS 3 consists of 7,674 individuals, and, when weighted, are representative of the US population aged 18 years and over.

Study variables

The first independent variable is BMI. BMI was calculated from self-reported height and weight and was classified into four ordered categories based on National Heart, Lung, and Blood Institute (NHLBI, 1998) classifications: (1) 18.5 and below as ‘underweight’, (2) 18.5–24.9 as ‘normal weight’, (3) 25.0–29.9 as ‘overweight’ and (4) 30.0 and above as ‘obese’.

The second independent variable involves respondents’ perceived levels of patient-centred communication. In HINTS 3, patient-centred communication was measured using six items representing six dimensions of patient-centred communication as conceptualised by Epstein and Street (2007). Therefore, to provide an overall measure of patient-centred communication, we averaged responses to these six items available in HINTS 3. Participants were first told that they would be asked a series of questions related to communication with health professionals. Participants were then asked, within the past 12 months: (1) ‘How often did doctors, nurses, or other providers provide you with the chance to ask all of the health-related questions that you had?’ (2) ‘How often did doctors, nurses, or other providers give you the attention that you needed for your feelings and emotions?’ (3) ‘How often did doctors, nurses, or other providers involve you in decisions about your general health care?’ (4) ‘How often did doctors, nurses, or other providers ensure that you understood the things you needed to do to take care of yourself?’ and (5) ‘How often did doctors, nurses, or other providers help you deal with feelings of uncertainty about your health?’ Finally, participants were asked how often, over the past 12 months, they felt they could rely on doctors, nurses or other health professionals to take care of their healthcare needs. Possible responses for all six questions included the following: (1) ‘never’, (2) ‘sometimes’, (3) ‘usually’ and (4) ‘always’.

Responses to these six questions were averaged to create a scale score representing the overall degree of patient-centredness of provider communication; scale scores ranged from 1 to 4, with higher scores representing higher levels of patient-centred communication. Respondents who were missing responses to three or more of the individual questions were treated as missing for the scale score. For those missing one or two responses, the remaining responses were averaged; for example, for those respondents with two questions missing, the scale score average was based on responses to the remaining four questions. The scale had an acceptable level of reliability (Cronbach’s alpha = .89).

The dependent variable was a dichotomous measure of the use of the Internet for health information. This variable was created by combining responses from two questions: (1) ‘The most recent time you looked for information about health or medical topics, where did you go first?’ and (2) ‘Did you look or go anywhere else?’ For the first question, respondents could choose one source from 14 listed choices; for the second question, respondents could choose multiple sources of information from the 14 listed choices. Respondents who selected ‘Internet’ as an information source for either question were coded as Internet health information–seekers; those who did not select ‘Internet’ for either question were coded as non-Internet health information–seekers.

Potential covariates included age (18–34, 35–49, 50–64, 65–75 or 75 years and over) and sex (male or female). Because of small sample sizes in some racial groups, race and ethnicity were recoded into five categories: non-Hispanic White, Hispanic, non-Hispanic Black, non-Hispanic Asian or other/multi-racial. Other potential covariates included education level (less than high school, high school, some college/technical school or college graduate), employment status (unemployed, employed, retired or other) and nativity (foreign-born or US-born). Health and healthcare-related variables included health insurance status (insured or not insured), having a regular source of care (yes or no) and perceived health status (fair/poor, good or very good/excellent).

Data analysis

All analyses were conducted in Stata version 12.1. The data were weighted using provided sampling weights to make the results of the study generalisable to the non-institutionalised US population aged 18 years and over. For all variables, ‘don’t know’ and ‘refused’ responses were coded as missing.

We used cross-tabulations and unadjusted logistic regression analyses to determine covariates that were significantly associated with Internet health information–seeking. Those covariates with associations significant at the p < .1 level were retained for the adjusted model. Multiple logistic regression was then used to determine adjusted associations between BMI and the use of the Internet as a health information source.

Small cell sample sizes limited our ability to obtain reliable estimates of the interaction term between BMI category and patient-centred communication. Therefore, we examined effect modification of BMI on the relationship between patient-centred communication and Internet health information–seeking by stratifying these analyses by BMI category (normal weight group, overweight group and obese group).

Results

Most (78.7%) individuals reported using the Internet for health information. On average, patient-centred communication scores were high (3.39 on a scale of 1–4). Participant characteristics are shown in Table 1 overall and by Internet use for health information. Cross-tabulation analyses suggested that four covariates were significantly associated with Internet health information–seeking; those who used the Internet for health information were younger (p < .001), more highly educated (p < .001), more likely to be US-born than foreign-born (p < .001) and more likely to be in excellent or very good health (p < .001).

Table 1.

Respondents’ sociodemographic and health-related characteristics (n = 5,568).

Characteristic	Frequency distribution unweighted, n (weighted %)	Used the Internet for health information unweighted, n (weighted %)	Did not use the Internet for health information unweighted, n (weighted %)
Age (years)*
18–34	775 (23.8)	681 (25.8)	94 (16.1)
35–49	1,801 (41.8)	1,607 (47.1)	194 (22.3)
50–64	1,687 (21.3)	1,275 (19.8)	412 (26.9)
65–74	791 (7.7)	478 (5.6)	313 (15.4)
75+	473 (5.5)	138 (1.7)	335 (19.3)
Sex
Male	1,911 (46.7)	1,414 (47.1)	497 (45.1)
Female	3,653 (53.3)	2,785 (52.9)	868 (55.0)
Race and ethnicity
White, non-Hispanic	4,155 (72.3)	3,234 (74.3)	921 (64.3)
Hispanic	286 (7.2)	175 (5.8)	111 (12.7)
Black, non-Hispanic	513 (15.0)	352 (14.5)	161 (17.1)
Asian, non-Hispanic	121 (3.7)	90 (3.7)	31 (3.6)
Other/multi-racial	145 (1.8)	98 (1.7)	47 (2.3)
Education level*
Less than HS	240 (6.8)	68 (3.2)	172 (20.5)
HS graduate	1,262 (26.9)	887 (26.2)	375 (29.6)
Some college/tech	1,642 (3.9)	1,241 (40.0)	401 (33.4)
College graduate	2,148 (27.6)	1,794 (30.6)	354 (16.5)
Nativity*
Foreign-born	427 (10.0)	270 (8.3)	157 (16.4)
US-born	4,863 (90.0)	3,718 (91.7)	1,145 (83.7)
Health insurance status
Not insured	539 (16.6)	388 (16.8)	151 (16.0)
Insured	4,959 (83.4)	3,772 (83.2)	1,187 (84.0)
Regular source of care
Does not have a regular source of care	927 (26.5)	699 (27.4)	228 (23.3)
Has a regular source of care	4,592 (73.5)	3,482 (72.6)	1,110 (76.8)
Perceived health status*
Very good/excellent	2,974 (57.5)	2,421 (61.8)	553 (41.3)
Good	1,684 (31.2)	1,210 (29.4)	474 (37.7)
Fair/poor	627 (11.3)	357 (8.7)	270 (20.9)
Body mass index
Underweight	71 (1.6)	45 (1.5)	26 (1.8)
Normal weight	2,086 (41.3)	1,624 (43.1)	462 (34.6)
Overweight	1,767 (30.1)	1,312 (28.2)	455 (37.3)
Obese	1,325 (27.0)	976 (27.2)	349 (26.3)
Patient-centred communication	Unweighted mean (SD)	Unweighted mean (SD)	Unweighted mean (SD)
Patient-centred communication	3.39 (0.63)	3.40 (0.62)	3.44 (0.62)

SD: standard deviation.

χ² association: *p < .001.

The results for the unadjusted and adjusted logistic regression models are shown in Table 2. In adjusted analyses (including retained covariates), BMI (compared to normal weight, overweight adjusted odds ratio [AOR] = 0.96, confidence interval [CI]: 0.64, 1.46; obese AOR = 1.45, CI: 0.84, 2.50) and patient-centred communication (AOR = 0.93; CI: 0.69, 1.25) were not significantly associated with use of the Internet for health information. The final model (excluding BMI) was then applied in a BMI-stratified analysis (Table 3). Adjusting for covariates, patient-centred communication was not significantly associated with use of the Internet for health information among individuals with normal weight, among individuals with overweight or among individuals with obesity.

Table 2.

Unadjusted and adjusted associations between BMI category and covariates and using the Internet for health information.

Characteristic	Unadjusted OR (95% CI)	Adjusted OR (95% CI)
Age (years)
18–34	Referent	Referent
35–49	1.32 (0.72, 2.40)	0.98 (0.61, 1.56)
50–64	0.46 (0.30, 0.70)**	0.32 (0.21, 0.48)***
65–74	0.23 (0.10, 0.50)***	0.14 (0.07, 0.26)***
75+	0.06 (0.04, 0.08)***	0.04 (0.03, 0.06)***
Sex
Male	Referent	–
Female	0.92 (0.48, 1.75)	–
Race/ethnicity
White, non-Hispanic	Referent	Referent
Hispanic	0.39 (0.26, 0.60)***	0.47 (0.27, 0.84)*
Black, non-Hispanic	0.73 (0.11, 4.70)	0.23 (0.13, 0.42)***
Asian, non-Hispanic	0.89 (0.45, 1.75)	0.83 (0.42, 1.66)
Other/Multi-racial	0.64 (0.35, 1.17)	0.53 (0.26, 1.07)
Education level
Less than HS	0.18 (0.07, 0.42)***	0.32 (0.17, 0.59)**
HS graduate	Referent	Referent
Some college/tech	1.35 (0.53, 3.41)	1.07 (0.69, 1.68)
College graduate	2.10 (0.99, 4.46)	2.58 (1.50, 4.46)**
Nativity
Foreign-born	Referent	Referent
US-born	2.15 (1.43, 3.24)***	1.74 (1.02, 2.96)*
Health insurance status
Not insured	Referent	–
Insured	0.95 (0.22, 4.08)	–
Regular source of care
Does not have a regular source of care	Referent	–
Has a regular source of care	0.80 (0.35, 1.86)	–
Perceived health status
Very good/excellent	Referent	Referent
Good	0.52 (0.33, 0.83)**	0.75 (0.48, 1.16)
Fair/poor	0.28 (0.18, 0.44)***	0.67 (0.44, 1.01)
Patient-centred communication	0.94 (0.69, 1.28)	0.93 (0.69, 1.25)
BMI category
Underweight^a	0.68 (0.24, 1.93)	1.45 (0.71, 2.98)
Normal weight	Referent	Referent
Overweight	0.61 (0.37, 1.00)	0.96 (0.64, 1.46)
Obese	0.83 (0.37, 1.88)	1.45 (0.84, 2.50)

BMI: body mass index; OR: odds ratio; CI: confidence interval.

Estimates may be statistically unreliable due to small sample size in this BMI category (n = 71).

p < .05; **p < .01; ***p < .001.

Table 3.

Adjusted associations between patient-centred communication and using the Internet for health information: stratified by BMI category.

Characteristic	Adjusted^a OR (95% CI) among individuals with normal weight (n = 1,749)	Adjusted^a OR (95% CI) among individuals with overweight (n = 1,581)	Adjusted^a OR (95% CI) among individuals with obesity (n = 1,192)
Patient-centred communication	0.85 (0.56, 1.29)	1.19 (0.74, 1.91)	0.72 (0.52, 1.02)

BMI: body mass index; OR: odds ratio; CI: confidence interval.

Adjusted for covariates: age, race/ethnicity, education, perceived health status and nativity.

Because Internet information–seeking may be influenced by access to the Internet, we repeated our analyses among individuals who were Internet users. We limited our sample to those who responded ‘yes’ to the following question: ‘Do you ever go on-line to access the Internet or World Wide Web, or to send and receive e-mail?’ Findings were similar using this limited sample (data not shown).

Discussion

This nationally representative study in the USA examined Internet health information–seeking in the context of BMI. The results of our study suggest that contrary to our hypotheses, overweight and obesity were not significantly associated with using the Internet for health information; furthermore, stratified analyses do not provide evidence that the association between patient-centred communication and the use of the Internet for health information differs based on BMI classification. Our results indicate that BMI may not be an important characteristic to consider when predicting or examining the use of the Internet for health information. Regardless of BMI classification, patient-centred communication did not predict the use of the Internet for health information.

Consistent with previous literature, education level (Koch-Weser et al., 2010; Ybarra and Suman, 2006) and age (Koch-Weser et al., 2010) were associated with Internet health information–seeking, with older individuals and those with lower levels of education having lower odds of Internet health information–seeking. Also consistent with the previous literature (Koch-Weser et al., 2010; Ybarra and Suman, 2006), we found that those who are non-Hispanic Black were less likely to seek health information from the Internet compared to those who are non-Hispanic White. Our study also suggested that those of Hispanic ethnicity were less likely to seek health information from the Internet.

The overall results of our study indicate that overweight and obesity are not important predictors when examining Internet health information–seeking. However, our study has limitations that should be considered. It may be difficult to capture all relevant characteristics of health information–seeking with a single outcome variable. As previously noted, approximately 75% of individuals in our weighted sample reported seeking health information from the Internet. In analyses not shown, however, 87.9% of health information–seekers reported seeking information from more than one source. For the majority of participants, the outcome variable does not reflect exclusive use of one source of health information, suggesting that the Internet is often used as a complement to health information from other sources rather than as an exclusive source of health information.

In the context of Uses and Gratifications Theory (Katz et al., 1973a, 1973b), this finding may indicate that multiple sources of health information are used by many individuals to fulfil their perceived unmet information needs. Available measures only address which source an individual uses ‘first’ (which may reflect the most easily accessible source rather than the preferred source), as well as which other sources an individual has used, which does not allow us to comment on ‘preferred’ sources of information. Our measures were also limited to the broad definition of seeking ‘health information’, which may not necessarily reflect seeking obesity- or weight-related information. If we had the ability to measure seeking health information specific to weight or obesity, it is possible that our results would have differed; future studies would benefit from examining health information–seeking as it relates to these specific topics. Finally, we were unable to assess motivations for seeking health information. For example, some participants may have sought information from the Internet based on a recommendation or referral by a provider or to supplement information provided by a provider. Future studies may benefit from assessing the frequency, preference or motivation of seeking specific types of health information to provide greater insight into decisions that individuals make regarding where to seek health information.

Our study also has limitations related to available measures of patient–provider interactions. Available questions measured six aspects of patient-centred communication, as opposed to the overall quality of the relationship or communication between patients and providers. Available questions also refer to interactions with ‘doctors, nurses or other providers’ as opposed to the relationship with one provider in particular. It is possible, then, that some individuals may have experienced negative attitudes or stigma from their primary providers, but they may have experienced more positive interactions with other types of providers, leading them to report more positive overall perceived patient-centred communication. Additionally, nearly 40% of the weighted sample answered the highest available category (‘always’) for all five questions. The ceiling effects of our scale score may have also made it difficult to accurately capture the degree of positivity or negativity to which patients view their interactions with their providers.

The HINTS 3 dataset is also cross-sectional, which limits the ability to examine temporal relationships between the study’s variables. HINTS 3 also had a low response rate (24.23% for the telephone survey mode and 30.99% for the mail survey mode) (Cantor et al., 2007), which may have introduced the potential for non-response bias. All responses were also self-reported in HINTS 3, including participant height and weight. Finally, data for HINTS 3 were collected in 2008; it is likely that Internet use, in general and for health information, has become more common over time.

Despite these limitations, to our knowledge, this study is the first to examine relationships between BMI category and general Internet health information–seeking in a US nationally representative dataset. Other studies have examined similar relationships using HINTS data, including various predictors (including BMI) of use of the Internet for assistance with diet, weight and physical activity (McCully et al., 2013) and various predictors (not including BMI) of Internet health information–seeking and other e-health behaviours (Kontos et al., 2014). We examined BMI as a predictor of general health information–seeking on the Internet, as well as possible effect modification of BMI on the relationship between patient-centred communication and Internet health information–seeking. Our non-significant findings suggest implications for public health practice. Data from this iteration of HINTS (not shown) suggest that approximately 75% of respondents used the Internet for health information, which has potential implications related to health behaviours, healthcare utilisation and health communication with providers. Future studies examining the influence of health information–seeking from the Internet on these factors would be beneficial and would provide greater insight on the impact of this health information source on health-related behaviours. Our study’s results also suggest that regardless of patients’ BMI status, providers should promote open conversations regarding the use of the Internet for health information and should encourage the use of reputable web-based resources to complement information provided in the appointment, including information related to weight and weight loss.

The Internet will likely continue to be an important source of health information, particularly among certain populations. Some populations have been shown to use the Internet for health information more frequently than others, but our study suggests that overweight and obesity may not be important predictors of general Internet health information–seeking. Further research may be warranted to better understand individuals’ health information–seeking patterns, including more comprehensive measures related to the frequency, preference and motivations of seeking different types of health information from various sources.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References

Amy

Aalborg

Lyons

. (2006) Barriers to routine gynecological cancer screening for White and African-American obese women. International Journal of Obesity 30: 147–155.

Anderson

(2004) Consumers of e-health: Patterns of use and barriers. Social Science Computer Review 22(2): 242–248.

Bagley

Conklin

Isherwood

. (1989) Attitudes of nurses toward obesity and obese patients. Perceptual and Motor Skills 68: 954.

Baker

Wagner

Singer

. (2003) Use of the Internet and e-mail for health care information: Results from a national survey. JAMA 289: 2400–2406.

Berger

Wagner

Baker

(2005) Internet use and stigmatized illness. Social Science & Medicine 61: 1821–1827.

Berryman

Dubale

Manchester

. (2006) Dietetic students possess negative attitudes toward obesity similar to nondietetics students. Journal of the American Dietetic Association 106: 1678–1682.

Bocquier

Verger

Basdevant

. (2005) Overweight and obesity: Knowledge, attitudes, and practices of general practitioners in France. Obesity Research 13: 787–795.

Brodie

Kjellson

Hoff

. (1999) Perceptions of Latinos, African Americans, and Whites on media as a health information source. Howard Journal of Communication 10: 147–167.

Brown

Thompson

(2007) Primary care nurses’ attitudes, beliefs, and own body size in relation to obesity management. Journal of Advanced Nursing 60: 535–543.

10.

Cantor

Coa

Crystal-Mansour

. (2007) Health Information National Trends Survey (HINTS) 2007: Final Report. National Cancer Institute. Available at: http://hints.cancer.gov/docs/HINTS2007FinalReport.pdf (accessed 15 December 2013).

11.

Chambliss

Finley

Blair

(2004) Attitudes toward obese individuals among exercise science students. Medicine and Science in Sport and Exercise 36: 468–474.

12.

De Brun

McCarthy

McKenzie

. (2014) Weight stigma and narrative resistance evident in online discussions of obesity. Appetite 72: 73–81.

13.

Epstein

Street

Jr (2007) Patient-centered communication in cancer care: Promoting healing and reducing suffering. NIH publication no. 07-6225, October. Bethesda, MD: National Cancer Institute.

14.

Foster

Wadden

Makris

. (2003) Primary care physicians’ attitudes about obesity and its treatment. Obesity Research 11(10): 1168–1177.

15.

Freisling

Haas

Elmadfa

(2009) Mass media nutrition information sources and associations with fruit and vegetable consumption among adolescents. Public Health Nutrition 13(2): 269–275.

16.

Gallagher

Doherty

(2009) Searching for health information online: Characteristics of online health seekers. Journal of Evidence-Based Medicine 2: 99–106.

17.

Hare

Price

Flynn

. (2000) Attitudes and perceptions of fitness professionals regarding obesity. Journal of Community Health 25: 5–21.

18.

Hoppe

Ogden

(1997) Practice nurses’ beliefs about obesity and weight related interventions in primary care. International Journal of Obesity and Related Metabolic Disorders 21: 141–146.

19.

Hou

Shim

(2010) The role of provider–patient communication and trust in online sources in Internet use for health-related activities. Journal of Health Communication 15: 186–199.

20.

Katz

Blumler

Gurevitch

(1973a) Uses and gratifications research. Public Opinion Quarterly 37(4): 509–523.

21.

Katz

Gurevitch

Haas

(1973b) On the use of mass media for important things. American Sociological Review 38: 164–181.

22.

Koch-Weser

Bradshaw

Gualtieri

. (2010) The Internet as a health information source: Findings from the 2007 Health Information National Trends Survey and implications for health communication. Journal of Health Communication 15: 279–293.

23.

Kontos

Blake

Chou

. (2014) Predictors of eHealth usage: Insights on the digital divide from the Health Information National Trends Survey 2012. Journal of Medical Internet Research 16(7): e172.

24.

Kristeller

Hoerr

(1997) Physician attitudes towards managing obesity: Differences among six specialty groups. Preventive Medicine 26: 542–549.

25.

Lewis

Martinez

Freres

. (2012) Seeking cancer-related information from media and family/friends increases fruit and vegetable consumption among cancer patients. Health Communication 27(4): 380–388.

26.

Lewis

Thomas

Blood

. (2010) ‘I’m searching for solutions’: Why are obese individuals turning to the Internet for help and support with ‘being fat’? Health Expectations 14: 339–350.

27.

McArthur

Ross

(1997) Attitudes of registered dietitians toward personal overweight and overweight clients. Journal of the American Dietetic Association 97: 63–66.

28.

McCully

Don

Updegraff

(2013) Using the Internet to help with diet, weight, and physical activity: Results from the Health Information National Trends Survey (HINTS). Journal of Medical Internet Research 15(8): e148.

29.

McKenna

Bargh

(2000) Plan 9 from cyberspace: The implications of the Internet for personality and social psychology. Personality and Social Psychology Review 4(1): 57–75.

30.

Maddox

Liederman

(1969) Overweight as a social disability with medical implications. Journal of Medical Education 44: 214–220.

31.

Maroney

Golub

(1992) Nurses’ attitudes toward obese persons and certain ethnic groups. Perceptual and Motor Skills 75: 387–391.

32.

National Heart, Lung, and Blood Institute (NHLBI) (1998) Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. McLean, VA: International Medical Publishing, Inc.

33.

Olson

Schumaker

Yawn

(1994) Overweight women delay medical care. Archives of Family Medicine 3: 888–892.

34.

Powell

McCarthy

Eysenbach

(2003) Cross-sectional survey of users of Internet depression communities. BMC Psychiatry 3: 19.

35.

Renahy

Parizot

Chauvin

(2008) Health information seeking on the Internet: A double divide? Results from a representative survey in the Paris metropolitan area, France, 2005–2006. BMC Public Health 8: 69.

36.

Rice

(2006) Influences, usage, and outcomes of Internet health information searching: Multivariate results from the Pew surveys. International Journal of Medical Informatics 75(1): 8–28.

37.

Shim

Kelly

Hornik

(2006) Cancer information scanning and seeking behavior is associated with knowledge, lifestyle choices, and screening. Journal of Health Communication 11: 157–172.

38.

Tustin

(2010) The role of patient satisfaction in online health information seeking. Journal of Health Communication 15(1): 3–17.

39.

Wadden

Anderson

Foster

. (2000) Obese women’s perceptions of their physicians’ weight management attitudes and practices. Archives of Family Medicine 9: 854–860.

40.

Wear

Aultman

Varley

. (2006) Making fun of patients: Medical students’ perceptions and use of derogatory and cynical humor in clinical settings. Academic Medicine 81: 454–462.

41.

Wee

McCarthy

Davis

. (2000) Screening for cervical and breast cancer: Is obesity an unrecognized barrier to preventive care? Annals of Internal Medicine 132: 699–704.

42.

Wright

Bell

(2003) Health-related support groups on the Internet: Linking empirical findings to social support and computer-mediated communication theory. Journal of Health Psychology 8(1): 39–54.

43.

Ybarra

Suman

(2006) Help seeking behavior and the Internet: A national survey. International Journal of Medical Informatics 75: 29–34.