Health Promotion International Advance Access originally published online on December 13, 2007
Health Promotion International 2008 23(1):70-77; doi:10.1093/heapro/dam039
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
PERSPECTIVES |
Relations between Internet use, socio-economic status (SES), social support and subjective health
1Norwegian Centre for Telemedicine, University Hospital of Northern Norway, PO Box 35, N-9038 Tromsø, Norway 2Department of Medical Informatics, Friedrich-Alexander-University Erlangen-Nürnberg, Erlangen, Germany 3 Universidade de Aveiro, Aveiro, Portugal 4 Foundation for Research and Technology, Heraklion, Greece
* Corresponding author. E-mail: silje.camilla.wangberg{at}telemed.no
 |
SUMMARY |
|---|
 TOP SUMMARY INTRODUCTIONMETHODRESULTSDISCUSSIONCONCLUSIONFUNDINGREFERENCES |
|---|
This study aimed to explore relations between Internet use, socio-economic status (SES), social support and subjective health. Participants were from representative samples between 15 and 80 years of age from seven different European countries. Two different survey datasets were used: (i) eHealth trends (eHT; N = 7934) and (ii) the European social survey (ESS2; N = 11248). Internet users who had used the Internet for health purposes were compared with Internet users who had not used it for health purposes. Structural equation modelling was used to assess the relationships between SES, Internet use, social support and subjective health. Use of other media was compared to Internet use in relation to social support and subjective health. Internet use was found to be more closely related to social support and subjective health than use of other media. Internet use was also found to be a plausible mediator between SES and subjective health, especially through interacting with social support.
Key words: Internet; SES; social support; subjective health
|
INTRODUCTION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODRESULTSDISCUSSIONCONCLUSIONFUNDINGREFERENCES |
|---|
There is a call for more research on the role of information and communication technology within health promotion, especially focusing on the societal and community perspective (Lintonen et al., 2007). We do not currently know enough about what role the Internet may play in relation to social disparities in health.
Some studies indicate that interventions to close the digital divide can lead to health benefits (Kreps, 2005; Shaw et al., 2006). In the absence of such interventions, however, it seems likely that Internet use may exacerbate existing SES differences in health. Similar effects of the Internet further empowering the empowered, have been found in political participation in the USA (Weber et al., 2003).
The Internet is not only an information highway. As a medium, the Internet might be considered more like a telephone than a TV, as it tends to increase social interactions instead of ‘stealing’ time away from it (Robinson et al., 2000). As a computer network, which is inherently social in nature, the Internet have the potential of increasing social capital (Wellman, 2001).
Furthermore, there seems to be some special qualities of the Internet such as the possibility for anonymity and asynchronous communication, freed from the limitations of time and space, that facilitates the formation of intimate personal relationships (Bargh et al., 2002). This indicates that the Internet may be a good facilitator of social support. Several studies suggest that is the case (e.g. Wellman et al., 2001; Barrera et al., 2002; Shaw et al., 2006), although a review concluded that there is a lack of robust outcome evidence on interventions trying to use the Internet to deliver social support (Eysenbach et al., 2004).
Internet-based interventions have a large potential for reaching people. In a recent survey, 44% of Europeans had used the Internet for health purposes (Andreassen et al., 2007). The following factors were associated with Internet use for health purposes: youth, being female, higher education, white-collar or no paid job, more visits to the GP, long-term illness or disabilities, and good subjective health. No potentially mediating mechanisms such as social support or self-efficacy were assessed in the survey. Furthermore, no comparisons were made within Internet users between those who had used the Internet for health purposes and those who had not.
The current study seeks to generate causal hypothesis about the interplay between SES, Internet use, social support and subjective health by: (i) in the same dataset as Andreassen et al. (eHT; Andreassen et al., 2007), compare Internet users who had not used the Internet for health purposes with those who had used the Internet for health purposes; (ii) by drawing upon additional data from the European Social Survey 2 (ESS2; Jowell and the Central Co-ordinating Team, 2005) to compare the use of Internet with use of other media in relation to social support and health and (iii) to model the relations between SES, Internet use, social support and subjective health.
We expected the Internet to be the only medium positively related to both social support and to subjective health. Further, we hypothesized that there would be an increasing SES gradient from overall Internet usage, through use of Internet but not for health purposes, to use of the Internet for health purposes.
|
METHOD |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODRESULTSDISCUSSIONCONCLUSIONFUNDINGREFERENCES |
|---|
EHT
Stratified samples of the populations (age group 15–80 years) of seven European countries (Norway, Denmark, Germany, Greece, Poland, Portugal and Latvia) participated in computer-assisted telephone interviews during the period October–November 2005 (N = 7934). For further details on the survey see Andreassen et al., 2007.
Data were gathered on the following variables: gender, age, education as measured through the International Standard Classification of Education (ISCED; UNESCO, 1997), living rurally or in a city, number of children, Internet use, subjective health status, number of visits to the GP and rated importance of different sources of health information.
ESS2
The European Social Survey is a biennial multi-country survey covering over 20 nations. The data in the second round was collected in 2004/2005 [ESS2; for more details see (Jowell and the Central Co-ordinating Team, 2005)]. Data from six of the same countries as in the eHT survey and in the same age group were used (N = 11248). Data from Latvia were not available. The variables used were on demographics, media use, social support and subjective health.
Measures
Internet use was measured in the ESS2 with the item: ‘how often do you use the internet, the World Wide Web or e-mail – whether at home or at work – for your personal use?’ Response alternatives was: ‘No access at home or work’, ‘Never use’, ‘Less than once a month’, ‘Once a month’, ‘Several times a month’, ‘Once a week’, ‘Several times a week’ or ‘Every day’. Use of TV and Newspapers was rated on an 8-point scale ranging from ‘No time at all’ to ‘more than 3 hours’ on ‘an average weekday’.
Internet use for health purposes was in the eHT measured with the item: ‘How often do you use the Internet to get information about health or illness?’ The response alternatives were: ‘every day’, ‘every week’, ‘every month’, ‘every six months’, ‘every year’, ‘less than once a year’ and ‘never’. Only those answering that they had used the Internet in general was asked this question.
Subjective health was measured with the same item in both surveys, ‘How is your health in general? Would you say it is 5 = very good, 4 = good, 3 = fair, 2 = bad, or 1 = very bad?’. The scale was reversed before analysis.
Social support was in the ESS2 measured with three items: (i) ‘how often do you meet socially with friends, relatives or work colleagues?’ (Sclmeet). The response alternatives were the same as for Internet use. (ii) ‘Do you have anyone with whom you can discuss intimate and personal matters?’ (intmdisc). The response alternatives were ‘yes’ or ‘no’. (iii) ‘Compared to other people of your age, how often would you say you take part in social activities?’ (Sclact). The response alternatives were: ‘Much less than most’, ‘Less than most’, ‘About the same’, ‘More than most’ or ‘Much more than most’.
Statistical analyses
Weighting was used in the analyses of both datasets in order to correct for differences in sizes of the countries' populations, as well as for minor skewedness in the variables gender, age and education for the countries that had not succeeded in obtaining a sample according to stratified sampling plan. No variables had more than 5% missing data.
ANOVA was used to compare differences between groups on continuous variables and chi-square to compare groups on categorical variables. MANOVA was used for rated importance of different sources of health information. Linear regression was used to look at the contribution of various media use to explain variance in subjective health.
Pearson's correlations were used to examine relations among individual items from the ESS2. Furthermore, structural equation modelling (SEM) was used to assess more complex relations between media use, SES, social support and subjective health in the ESS2 data. This allowed us to control for the effect of all the other variables on the respective relations simultaneously, look at mediational effects, and to have the constructs SES and social support represented by latent variables. The three items concerning social support (frequency of social meetings, social activities and having someone to discuss intimate issues with) were loaded onto a latent variable for use in SEM, and the same with the three SES variables (gender, education and income). Model specification was done in a confirmatory manner based on our hypothesized relations between the constructs. Only one post hoc modification was made, the addition of one error covariance.
All confidence intervals (CI) are 95%.
|
RESULTS |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODRESULTSDISCUSSIONCONCLUSIONFUNDINGREFERENCES |
|---|
ESS2
Those who had used the Internet (44.6%) had a higher educational attainment, better subjective health, more social interactions and fewer consultations with health professionals (Table 1).
|
Correlations between all variables can be seen in Table 2. As for relations to other media, Internet use was negatively correlated to TV use (r = –0.20), very weakly positively related to use of radio (r = 0.05), and not significantly related to reading newspapers.
|
All correlations between use of radio, TV and newspapers and the individual social support items were r < 0.09. For Internet use, the correlations ranged from 0.11 to 0.27.
A regression was performed with subjective health as the dependent variable and media use as predictors. Overall, we found that all media use was positively related to subjective health, with Internet use being related to an average increase in subjective health of 0.55 (0.52–0.58). As shown in Table 3, Internet explained an additional 8.3% of shared variance in subjective health after taking use of TV and newspapers into account.
|
The regression was also run separately amongst those who had not used the Internet (we thank an anonymous reviewer for pointing out the need to test this; Table 3). Overall use of TV and newspapers accounted for 0.6% of shared variance in subjective health amongst those who had not used the Internet, showing that also among those who had not used the Internet, other media is explaining less variance in subjective health. For TV watching, however, the CI's for the unstandardized coefficient (B) for Internet use in the total sample and for TV watching among non-users are overlapping, so we cannot say for sure that Internet use predicts greater increases in subjective health. Furthermore, the CI's for B in the total sample and in the sample of those who had not used the Internet are overlapping, and we can therefore not say that TV watching is a greater predictor of subjective health among those who do not use the Internet either.
The full structural model (Figure 1) showed that SES is a strong predictor of Internet use (β = 0.61, CI: 0.59–0.63), and have both a direct effect on subjective health (β = 0.13, CI: 0.09–0.17), as well as indirect effects mediated through Internet use and social support, resulting in a total effect of β = 0.28 (0.26–0.31). After controlling for SES, Internet use both have a direct effect on subjective health (β = 0.10, CI: 0.07–0.13), as well as an indirect effect mediated through social support (β = 0.06, CI: 0.05–0.08), resulting in a total effect of β = 0.16 (0.13–0.19).
|
eHT
Characteristics of those who had used the Internet but not for health purposes compared to Internet users who had used it for health can be seen in Table 4. Although most differences were found to be small, those who had used the Internet but not for health purposes tended to lower educational attainment, more children, visit their GP less often and were more likely to be male.
|
The respondents were asked to rate eight sources of health information for importance. Of these, CI's were overlapping for rated importance of three sources. Those who had not used the Internet for health purposes rated not only Internet, but four additional listed sources of health information as less important than the ones who had used the Internet for health purposes. The effect size for the difference between the groups on all items was
p2 = .14. The largest difference in rated importance was for the Internet, while differences for other media were small. All means can be seen in Table 4. |
DISCUSSION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODRESULTSDISCUSSIONCONCLUSIONFUNDINGREFERENCES |
|---|
Internet use was found to have both a direct positive relation to subjective health, as well as an indirect positive relation, mediated through social support.
As expected, the Internet differed from other media, and the most from TV, which in some ways can be characterized as the most passive medium. The communicative feature of the medium was also evident in that Internet use was more highly correlated with social support. The Internet did not differ so much from other media with regard to rated importance as an information medium. We think this reflects underlying constructs related to SES, such as health consciousness or health literacy.
Socio-economic status (SES) was related to subjective health, to Internet use in general, as well as health-related Internet use. The SES gradient in health has been well established through previous research (Graham, 2002). Studies have also previously linked SES to access to, and use of, the Internet (Fox, 2005) and health information off- and online (Lorence et al., 2006; Andreassen et al., 2007). Despite the straight-forward model, we assume complex mechanisms connect SES and the dependent variables.
A plausible mediating mechanism linking Internet use to health is self-efficacy. Self-efficacy has been found to partly mediate the relation between education and health behaviour (Leganger and Kraft, 2003). A recent study did not find a mediating effect for self-efficacy, but did find that self-efficacy moderated the effect of an Internet-based intervention on health behaviours (Wangberg, 2007).
A second and related mechanism is that of adaptive coping, such as changing risk behaviour versus dysfunctional coping, for instance avoiding further information. Iversen and Kraft found that coping styles mediated the relationship between SES and health information (Iversen and Kraft, 2006).
A third plausible mediating mechanism, that is also reflected in our data, is the Internet's potential to facilitate social support ( Wellman et al., 2001; Barrera et al., 2002; Boase et al., 2006). Social support have both a direct effect on health, as well as an indirect effect through influencing health behaviours [for an overview, see (Cohen et al., 2000)]. SES has previously been related to social support, but the precise role social support play in the relation between SES and health is still unclear. Some studies like the current, support the notion that social support is a mediator between SES and health (Mickelson and Kubzansky, 2003), while a recent study found no mediating effect either on subjective or objective health (Gorman and Sivaganesan, 2007). A more complex relationship seems to be plausible, wherein SES moderates for whom social support acts as a mediator between SES and health, e.g. that social support may act as a buffer against stress experienced by low SES individuals (Mickelson and Kubzansky, 2003; Gorman and Sivaganesan, 2007).
As for other moderating relations, literacy in general and health literacy in particular, is likely to influence for whom Internet use is related to subjective health (Nutbeam, 2000; McCray, 2005; Norman and Skinner, 2006). Literacy and SES are also intertwined (McCray, 2005).
Limitations to the study
The cross-sectional data means that causal attributions cannot be made. The same is true for strict mediational analysis, which requires longitudinal data with at least three time-points.
Our main outcome measure was of subjective health, meaning we cannot generalize the findings to objective health measures. SES have been found to be more closely related to subjective than to objective health measures (Gorman and Sivaganesan, 2007).
The structural model did not show a good fit according to chi-square, although the other measures indicated a good fitting model. It is known that the chi-square can be overly sensitive in large samples. Furthermore, our measurement models for both SES and social support were not perfect and may thus explain some lack of fit, as well as weaker than expected relations between some of the latent variables. It has also been suggested that surveys may underestimate the relationship between SES and subjective health (Lorant et al., 2007).
For the same reason (lack of ideal measurement), we suspect, the modification indices suggested adding an error covariance between the error terms for education and social meetings. Another plausible explanation is that this negative covariance could reflect some kind of underlying shortage of time construct.
Nevertheless, considering the overall picture emerging from our data in light of previous research, we find it warranted to draw the general conclusions that we do in terms of plausible relationships, i.e. that Internet could act as a mediator between SES and subjective health, and that one of the ways Internet use could positively influence health is through increasing availability of social support. We therefore believe that the correlational model we have presented could be useful for guiding causal hypothesis generation and model specification in future longitudinal research.
|
CONCLUSION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODRESULTSDISCUSSIONCONCLUSIONFUNDINGREFERENCES |
|---|
Internet use was found to be a mediator between socio-economic variables and health. This suggests that the digital divide and social disparities in health are overlapping. Hence a bridging of the digital divide and better access to the Internet in the European population cannot be expected to influence social disparities in health alone. It does, however, seem likely that the Internet may play an important role in this. Therefore, we believe that one should aim further research at exploiting the health promoting potential of the Internet alongside other efforts to flatten out the current social gradient in health.
|
FUNDING |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODRESULTSDISCUSSIONCONCLUSIONFUNDINGREFERENCES |
|---|
S.C.W. was funded by a doctoral fellowship from the Norwegian Research Council, grant no. 167063/V50. The e-health consumer trends survey (eHT) was in part financed by the Programme of Community Action in the Field of Public Health E-health (2003–2008) of the Health and consumer protection directorate general, directorate C, EC. Funding to pay the Open Access publication charges for this article was provided by the author.
|
ACKNOWLEDGEMENTS |
|---|
The ESS data was downloaded from the Norwegian Social Science Data Services' (NSD) data archive. We acknowledge the following for their efforts in eHT design and data collection: Roxana C. Dumitru, Laurence Esterle, Maria M. Bujnowska-Fedak, Per Hjortdahl, Angelina Kouroubali, Per Egil Kummervold, Antònio Sousa Pereira, Iveta Pudule, Birgitte Lolan Ravn, Andrzej Staniszewski, Henning Voss, Manolis Tsiknakis and Rolf Wynn. We also thank Maurice B. Mittelmark for useful suggestions for improving the manuscript.
|
REFERENCES |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODRESULTSDISCUSSIONCONCLUSIONFUNDINGREFERENCES |
|---|
Andreassen H. K., Bujnowska-Fedak M. M., Chronaki C. E., Dumitru R. C., Pudule I., Santana S., et al. European citizens' use of E-health services: a study of seven countries. BMC Public Health (2007) 7:53.[CrossRef][Medline]
Bargh J. A., McKenna K. Y. A., Fitzsimons G. J. Can you see the real me? The activation and expression of the ‘true self’ on the Internet. Journal of Social Issues (2002) 5:33–48.
Barrera M. Jr, Glasgow R. E., McKay H. G., Boles S. M., Feil E. G. Do Internet-bases support interventions change perceptions of social support?: An experimental trial of approaches for supporting diabetes self-management. American Journal of Community Psychology (2002) 30:637–654.[CrossRef][ISI][Medline]
Boase J., Horrigan J. B., Wellmann B., Rainie L. The Strength of Internet Ties: The Internet and email aid users in maintaining their social networks and provide pathways to help when people face big decisions. (2006) Washington, DC: PEW Internet & American Life Project.
Cohen S., Gottlieb B. H., Underwood L. G. Social relationships and health. In: Social Support Measurement and Intervention: A Guide for Health and Social Scientists—Cohen S., Underwood L. G., Gottlieb B. H., eds. (2000) Oxford: Oxford University Press.
Eysenbach G., Powell J., Englesakis M., Rizo C., Stern A. Health related virtual communities and electronic support groups: systematic review of the effects of online peer to peer interactions. BMJ (2004) 328:1166–1172.
Fox S. Digital Divisions (2005) Washington, DC: PEW Internet & American Life Project.
Gorman B. K., Sivaganesan A. The role of social support and integration for understanding socioeconomic disparities in self-rated health and hypertension. Social Science and Medicine (2007) 65:958–975.[CrossRef][Medline]
Graham H. Building an inter-disciplinary science of health inequalities: the example of lifecourse research. Social Science and Medicine (2002) 55:2005–2016.[CrossRef][Medline]
Iversen A. C., Kraft P. Does socio-economic status and health consciousness influence how women respond to health related messages in media? Health Education Research (2006) 21:601–610.
Jowell R., and the Central Co-ordinating Team. European Social Survey 2004/2005: Technical Report (2005) London: Centre for Comparative Social Surveys, City University.
Kreps G. L. Disseminating relevant health information to underserved audiences: implications of the Digital Divide Pilot Projects. J Med Libr Assoc (2005) 93((Suppl. 4)):S68–S73.[ISI][Medline]
Leganger A., Kraft P. Control constructs: do they mediate the relation between educational attainment and health behaviour? Journal of Health Psychology (2003) 8:361–372.[Abstract]
Lintonen T. P., Konu A. I., Seedhouse D. Information technology in health promotion. Health Education Research (2007) 10.1093/her/cym001.
Lorant V., Demarest S., Miermans P.-J., Oyen H. V. Survey error in measuring socio-economic risk factors of health status: a comparison of a survey and a census. International Journal of Epidemiology (2007) 36, 1292–1299.
Lorence D. P., Park H., Fox S. Assessing health consumerism on the Web: a demographic profile of information-seeking behaviors. Journal of Medical Systems (2006) 30:251–258.[CrossRef][ISI][Medline]
McCray A. T. Promoting health literacy. Journal of the American Medical Informatics Association (2005) 12:152–163.
Mickelson K. D., Kubzansky L. D. Social Distribution of Social Support: The Mediating Role of Life Events. American Journal of Community Psychology (2003) 32:265–281.[CrossRef][ISI][Medline]
Norman D. C., Skinner A. H. eHealth Literacy: Essential Skills for Consumer Health in a Networked World. Journal of medical Internet research (2006) 8.
Nutbeam D. Health literacy as a public health goal: a challenge for contemporary health education and communication strategies into the 21st century. Health Promotion International (2000) 15:259–267.
Robinson J. P., Kestnbaum M., Neustadtl A., Alvarez A. Mass media use and social life among Internet users. Social Science Computer Review (2000) 18:490–501.[Abstract]
Shaw B., Gustafson D. H., Hawkins R., McTavish F., McDowell H., Pingree S., et al. How underserved breast cancer patients use and benefit from eHealth programs: implications for closing the digital divide. American Behavioral Scientist (2006) 49:823–834.
UNESCO. International Standard Classification of Education [Electronic Version]. (1997) http://www.unesco.org/education/information/nfsunesco/doc/isced_1997.htm.
Wangberg S. C. An Internet-based diabetes self-care intervention tailored to self-efficacy. Health Education Research (2007) 10.1093/her/cym014.
Weber L. M., Loumakis A., Bergman J. Who participates and why?: an analysis of citizens on the Internet and the mass public. Social Science Computer Review (2003) 21:26–42.[Abstract]
Wellman B. Computer networks as social networks. Science (2001) 293:2031–2034.
Wellman B., Haase A. Q., Witte J., Hampton K. Does the Internet increase, decrease, or supplement social capital?: social networks, participation, and community commitment. American Behavioral Scientist (2001) 45:436–455.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||