Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Cancer. Author manuscript; available in PMC 2012 July 1.
Published in final edited form as:
Published online 2011 January 24. doi:  10.1002/cncr.25849
PMCID: PMC3119780

Minimal Social Networks Effects Evident in Cancer Screening Behavior

Nancy L. Keating, M.D., M.P.H., A. James O’Malley, Ph.D., Joanne M. Murabito, MD ScM, Kirsten P. Smith, Ph.D., and Nicholas A. Christakis, M.D., Ph.D.



Social networks may influence screening behaviors. We assessed whether screening for breast, prostate, or colorectal cancer is influenced by the actual screening behaviors of siblings, friends, spouse, and coworkers.


Observational study using Framingham Heart Study data to assess screening for eligible individuals during the late 1990s. We used logistic regression to assess if the probability of screening for breast, prostate, or colorectal cancer was influenced by the proportion of siblings, friends, and coworkers who had the same screening, as well as spouse’s screening for colorectal cancer, adjusting for other factors that might influence screening rates.


Among 1660 women aged 41–70, 71.7% reported mammography in the past year; among 1217 men aged 51–70, 43.3% reported prostate specific antigen testing in the past year; and among 1426 men and women aged 51–80, 46.9% reported stool blood testing and/or sigmoidoscopy in the past year. An increasing proportion of sisters who had mammography in the past year was associated with mammography screening in the ego (odds ratio [OR]=1.034, 95% confidence interval [CI]=1.000–1.065 for each 10% increase). A spouse with recent screening was associated with more colorectal cancer screening (OR 1.65, 95% CI=1.39–1.98 vs. unmarried). Otherwise, screening behaviors of siblings, friends, and coworkers were not associated with screening in the ego.


Aside from a slight increase in breast cancer screening among women whose sisters were screened and colorectal cancer screening if spouses were screened, the screening behavior of siblings, friends, or coworkers did not influence cancer screening behaviors.

Keywords: cancer screening, mammography, prostate specific antigen, social networks

Screening for cancer has the potential to save lives by identifying cancers at earlier stages, when they may be more amenable to treatment and cure. Nevertheless, many individuals who may benefit do not undergo routine screening.15 Research suggests that individuals’ social supports, including family and friends, or their perception that screening is normative among their peers, may positively encourage screening,612 and such findings have led to interventions using peers and/or other community members or worksite interventions to increase rates of screening.1317

Social contacts can strongly influence a variety of behaviors, such as smoking, weight gain, and drinking.1820 They might also influence screening behaviors by several mechanisms. Social contacts might provide information or advice about the purpose of specific tests, the benefits of testing, or the need for evaluation of symptoms. They might also provide encouragement to someone who has avoided screening and they may provide emotional support to someone concerned about abnormal screening results. They might share their own experiences with screening, which may be more powerful at influencing behavior than sharing less direct knowledge about the tests. Finally, social contacts might assist a person in finding a doctor or getting to appointments.

Although other studies have suggested benefits of generic social support in encouraging screening behaviors, we are unaware of studies examining whether actual screening behavior of one’s peers influence an individual’s likelihood of cancer screening. We studied a large network of individuals to assess whether screening for breast cancer, prostate cancer, or colorectal cancer is influenced by the actual screening behaviors of one’s siblings, friends, and coworkers, who may differ in their likelihood of discussing screening and their influence on the ego. We hypothesized that the screening behavior of individuals would be positively influenced by the screening behaviors of their contacts, with the effects greater for siblings and friends than coworkers. In this study, we were able to measure actual behavior of individuals’ social contacts, rather than rely on an index case to inform us about the behavior of others to whom they were connected.


Data and subjects

This study used data from the Framingham Heart Study, which, in 1948, enrolled 5209 individuals in the Original cohort.21 Children of the original cohort and their spouses were recruited in 1971 to form the Offspring cohort; this cohort included 5124 individuals.22 In 1994, a minority oversample of 508 individuals was initiated, and in 2002, the Third-Generation cohort, consisting of 4095 children of the offspring cohort, was initiated.23 The study protocol was approved by the institutional review boards of Harvard Medical School and the Boston University Medical Center.


We focused on the 3807 so-called “egos” in the Offspring cohort known to have participated in waves 6 and/or 7 (data collection occurred during 3-year periods centered in 1997 for wave 6 and 1999 for wave 7). The “ego” is the person whose screening behavior is being analyzed. Any persons to whom these subjects are linked (in any of the Framingham Heart Study cohorts) can serve as social contacts, referred to as “alters.” Details of the ascertainment of alters are described in detail elsewhere.18 Briefly, information was derived from archived, administrative tracking sheets used to identify people close to the study participants to facilitate follow-up. The tracking sheets provided complete information about all first-order relatives (parents, spouses, siblings, and children) and at least one “close friend” and these names were linked with the Framingham data to identify ties between egos and alters. Information on address and place of work were used to identify neighbors and coworkers. We classified alters as parents, full sisters, full brothers, friends, and coworkers. We restricted analyses to individuals and their alters who were eligible for screening based on sex- and age-specific screening recommendations (described below). In sensitivity analyses, we repeated all analyses including a small number of half siblings, step siblings, adopted siblings, and foster siblings (5.8% of all siblings). Results were similar and are not presented. We also identified spouses for analyses of colorectal cancer screening.

Screening behaviors

Data on screening were collected in two waves: wave 6 (1995–1998) and wave 7 (1998–2001). Women were asked the year of their last mammogram, men were asked the year of their last blood test for prostate cancer, and men and women were asked the year when stool was last tested for blood and when sigmoidoscopy was last done.

We assessed screening behavior based on recommendations from national guidelines. For breast cancer screening, we assessed report of mammography within the past year for women ages 41 to 7024, 25 (results were similar when we restricted to ages 51 to 70, because some guidelines did not recommend mammography for average risk women in their 40s before 199726 or later,27, 28 and results were similar when we assessed mammography within the last 2 years). For prostate cancer screening, we assessed prostate specific antigen (PSA) testing within the past year among men aged 51–70. 25, 29 For colorectal cancer screening, we assessed receipt of stool blood testing within the past year and/or sigmoidoscopy within the past 5 years for individuals aged 51–80. Screening guidelines at the time recommended yearly stool occult blood testing or flexible sigmoidoscopy every 3 to 5 years with stool blood testing or colonoscopy;25, 27, 30 the survey did not ask specifically about colonoscopy.

Control variables

We identified factors likely to be associated with cancer risk and/or screening behavior. Specifically, we documented each participant’s sex and age and we used data from the prior survey wave (including wave 5 information for subjects whose screening behavior was ascertained in wave 6) to characterize additional participant characteristics, including self-rated health status (excellent/very good, good, fair/poor, unknown), number of years of education, marital status, number of children living (whether in the FHS or not), current employment, number of times per week of intense physical activity, current smoking status, number of alcoholic drinks per week, and presence of diabetes, cardiovascular disease, and/or pulmonary disease (asthma or chronic obstructive pulmonary disease/emphysema). We used validated data from the Framingham Study to document history of cardiovascular disease and diabetes; presence of pulmonary disease was based on the clinical impression of the Framingham Study clinic examiner. Continuous variables were not categorized.


We evaluated testing among eligible participants for each of the waves; hence participants eligible in both waves contributed two sets of data to the analyses. We used logistic regression models to assess the proportion of siblings, friends, and coworkers who had the same type of screening test on screening for eligible egos (calculating the effect for each 10% increase in the proportion screened). We also controlled for the number of siblings, friends, and coworkers who were eligible for screening. Models for mammography were limited to female siblings, friends, and coworkers, and those for PSA testing were limited to male siblings, friends, and coworkers. Models also included all control variables described above and survey wave (wave 7 vs. 6). The colorectal cancer screening model included men and women, and we included a variable for sex and a variable reflecting marital status and if married whether spouse was screened or not screened. All models used generalized estimating equations, clustering on participants, to account for the possibility that a participant may contribute up to two dependent variable observations (one each wave in the role of ego) or be involved in multiple observations of the predictor variable (in the role of alter).


We identified 1660 women aged 41–70, who had 597 sisters, 175 female friends, and 174 female coworkers aged 41–70 enrolled in the Framingham Heart Study. A total of 1269 women participated in both waves of the survey, so the total number of observations was 2929 women eligible for mammography; of these 71.7% had undergone mammography in the past year. We identified 1217 men aged 51–70, who had 337 brothers, 142 male friends, and 99 male coworkers aged 51–70. A total of 804 men participated in both waves of the survey, so the total number of observations was 2021; 43.3% had undergone PSA testing in the past year. We identified 3045 men and women aged 51–80. These individuals had 1426 siblings, 364 friends, and 299 coworkers aged 51–80, and 1530 had spouses aged 51–80. A total of 2260 participated in both waves, so the total number of observations was 5305; 46.9% had undergone stool blood testing in the past year and/or flexible sigmoidoscopy in the last 5 years. Characteristics of each cohort are included in Table 1.

Table 1
Characteristics of the study cohorts

Table 2 demonstrates the influence of the screening behaviors of siblings, friends, and coworkers on screening behavior of the ego. For mammography screening, an increasing proportion of sisters who had undergone screening mammography in the past year was slightly associated with mammography screening in the ego (odds ratio [OR] 1.034, 95% confidence interval [CI] 1.000 to 1.065 for each 10% increase in the proportion of sisters screened). At an average rate of 71.7% of women being screened, this odds ratio corresponds to a risk ratio31 of 1.009, suggesting a very small 0.9% increase in screening rates to 72.4% for a 10% increase in the proportion of sisters screened. Women with a greater number of sisters were less likely than women with fewer sisters to undergo mammography (OR 0.85, 95% CI 0.72 to 1.01), although this finding was of borderline statistical significance (P=.06). The proportion of female friends and female coworkers who had undergone mammography was not associated with the probability of the ego undergoing mammography screening.

Table 2
Factors associated with recommended screening in egos

The extent of PSA testing in the past year among siblings, friends, and coworkers was not associated with PSA testing among egos, nor was the number of brothers, male friends, or male coworkers.

For colorectal cancer screening with stool blood tests and/or sigmoidoscopy, individuals married to a spouse that had been screened were more likely to be screened than those who were unmarried (OR 1.65, 95% CI 1.39 to 1.98), with a risk ratio31 of 1.296. With 42.6% of unmarried individuals screened, this corresponds to a 12.6% absolute increased risk of screening, to 55.2% for married individuals whose spouses were screened. Individuals married to a spouse who had not been screened or for whom screening status was unknown did not differ from unmarried individuals in screening. The non-overlapping confidence intervals for married patients whose spouses were or were not screened suggest that screening status of the spouse is a more important factor than marital status itself. The proportion of siblings, friends, or coworkers who had been screened was not associated with screening. Individuals with more friends were less likely to undergo colorectal cancer screening (OR 0.73, 95% CI 0.58 to 0.91 for each additional friend in the cohort).

In each model we adjusted for several control variables, some of which had significant effects. Other characteristics of the egos associated with mammography screening included older age (OR 1.02, 95% CI 1.01 to 1.04 for each year of age), being married (OR 1.63, 95% CI 1.33, 2.00), and being physically active 4 or more times per week (OR 1.05, 95% CI 1.00, 1.09). Smokers were much less likely to undergo mammography than nonsmokers (OR 0.56, 95% CI 0.44, 0.71) as were women with more comorbid illnesses (OR 0.80, 95% CI 0.65 to 0.99). Participants surveyed in the wave centered in 1999 were more likely to have mammograms than those surveyed in the wave centered in 1997 (OR 1.40, 95% CI 1.21 to 1.62).

For PSA testing, older men were more likely to be screened than younger men (OR 1.05, 95% CI 1.03 to 1.08). Married men had more PSA tests than unmarried men (OR 1.73, 95% CI 1.29, 2.31), and men with more years of education were more likely to have PSA testing (OR 1.08, 95% CI 1.04, 1.12). Smokers were less likely than nonsmokers to have PSA testing (OR 0.71, 95% CI 0.54, 0.94) as were men with more comorbid illnesses (OR 0.76, 95% CI 0.62 to 0.94). Men surveyed in wave 7 were more likely to report recent PSA testing than those surveyed in wave 6 (OR 1.59, 95% CI 1.35, 1.87), reflecting the general increase in the use of screening over this time-frame.

For stool blood testing or flexible sigmoidoscopy, older participants were more likely to report screening (OR 1.02, 95% CI 1.01 to 1.03). Participants who worked were less likely than those who did not to be screened (OR 0.85, 95% CI 0.74, 0.98), as were those who smoked (OR 0.62, 95% CI 0.51 to 0.74) and those with more comorbidities (OR 0.83, 95% CI 0.73 to 0.94). Participants surveyed in the wave centered in 1999 were more likely to undergo colon cancer screening than those surveyed in the wave centered in 1997 (OR 1.52, 95% CI 1.38, 1.67).


We examined whether screening behavior of siblings, friends, coworkers, and spouses influenced analogous screening behaviors of individuals. We found that mammography screening increases with an increasing proportion of sisters who have had a mammogram (although women with more sisters were less likely to get mammograms, a finding of borderline statistical significance). PSA testing did not vary by the proportion of brothers, friends, or coworkers who had the test. Colorectal cancer screening was strongly associated with screening among one’s spouse, but not with the proportion of friends who were screened.

Several studies suggest that support of others increases an individual’s likelihood of participating in cancer screening. For example, women with higher scores on the social network index6, 7 or who report social support from physicians, family, and friends8 are more likely to undergo mammography and Pap smears. In addition, a study of employed women found that women who perceived that screening is normative among their peers were more likely to undergo regular mammography, although the extent of social support and the size of one’s social network was not associated with screening behavior.9 On the other hand, two studies have observed that women reporting explicit encouragement to undergo mammography by social network members were less likely to be screened,9, 32 suggesting that the women who avoid mammography may be more likely to be offered encouragement from others. Perceived risk of cancer is also associated with mammography screening, and this perceived risk is often due to a family history of cancer.33 In a previous study using data from the Framingham study, reporting a family history of breast cancer was strongly associated with reporting a mammogram in the last 2 years.34

In the current study, we were able to broaden the scope of social contacts examined (to include friends and coworkers), broaden the nature of cancers considered, and, most important, trace out direct ties between people and directly query alters about screening behavior rather than merely surveying egos about alters.

Past research suggests that friends can influence mammography behavior with direct efforts. One study randomized individuals to call or not call friends to encourage them to get a mammogram. Friends who received a call had a 15% increase in mammography compared with those who did not receive a call. This effect remained after controlling for demographic characteristics, was effective for black and white women of all ages, and was most pronounced among women with lower household incomes.35 In addition, women reporting close friends with whom they could discuss their health were more likely to have ever had a Pap smear.11 Programs have thus been developed that successfully use social support to improve screening for cervical cancer and breast cancer.36

Fewer data are available about the impact of interventions on social contacts on prostate cancer screening or colorectal cancer screening. The value of PSA testing for prostate cancer remains controversial,37 yet research suggests that patients deciding about PSA testing value anecdotes about the decisions of friends, family, or media celebrities.38 Thus, we had expected that the prostate screening behaviors of alters would influence those of the egos in our study. Consistent with other research,39 we found that married men were more likely than unmarried men to undergo PSA screening. Men may be encouraged by their wives to get more routine and preventive care, or may be persuaded to get PSA testing specifically.

Colorectal cancer screening can be inconvenient and invasive, and, for colonoscopy, requires time off from work and someone to accompany the individual to the procedure. These factors may lead to negative attitudes about screening.40 Nevertheless, support from friends and family has been associated with screening, as have positive attitudes about the screening and beliefs that it is safe.41 We found a strong association of colorectal cancer screening among spouses of individuals who have been screened, but no associations based on the proportion of siblings, friends, or coworkers who were screened.

Overall, this work again reinforces the distinction between social support and social network effects.42 The existence of social ties, and the willingness of others to help with health care can affect screening, as suggested by prior work. However, this is a different effect than that of the specific influence whereby an alter’s actual behavior influences a similar behavior in an ego. By analogy, it is the difference between the impact on a person’s happiness of having many friends versus the impact on a person’s happiness of having friends who are themselves happy.43 We found that the screening behaviors of one’s contacts, at least among those contacts included in the study, had little relevance to screening behaviors. Screening behaviors may be less “contagious” because they are not easily observed (unlike smoking, alcohol, obesity, and happiness)1820, 43 and may not be comfortable topics to discuss. New evidence suggests that ties among friends are influenced by observable characteristics such as obesity and smoking, but not by less easily observed traits, such as blood pressure and depression score.44

Our findings should be interpreted in light of some limitations. First, information on screening was only collected in two waves of the Framingham Study, both during the late 1990s. Thus, we cannot be certain that the findings are relevant to current screening behaviors; screening rates for colorectal cancer have increased since this time,45 although our study period corresponded with the peaking of mammography rates, which declined in the early 2000s.46 Second, our study focused on a single community that was lacking in racial and ethnic diversity, so the generalizability of our findings to other populations requires further study. Rates of prostate cancer screening in our cohort were lower than those of colorectal cancer screening, which has not been observed nationally.47 Third, we could only assess screening behaviors among alters who were included in the Framingham cohorts and of ages that would make them eligible for screening themselves, and our cohort of egos had relatively few alters in the study, limiting our power to observe effects. Moreover, if an individual had many friends but few were in the Framingham cohorts, then our study would likely underestimate the effects of the other friends’ behaviors. Fourth, the survey question about stool blood testing did not distinguish in-office or at-home testing. Finally, self-report of screening may overestimate use.48

In conclusion, mammography receipt among sisters and colorectal screening among spouses slightly influenced personal screening behaviors, but otherwise screening behaviors of siblings, friends, and coworkers were not associated with increased rates of cancer-specific screening. These observations suggest that while many health behaviors may spread across social ties, not all health behaviors necessarily do. Some behaviors may be intrinsically more “contagious”, just like some fashions are easier to adopt and some germs are more contagious than others.


This study was funded by the National Institute of Aging (R-01 AG-24448 and P01 AG-031093), and National Heart, Lung and Blood Institute Framingham Heart Study (contract N01-HC-25195).

The authors would like to thank Laurie Meneades, M.S., for expert programming assistance and Garrett Kirk, B.S., for administrative assistance.


The authors have no conflicts of interest to report.


1. Rimer BK, Conaway MR, Lyna PR, Rakowski W, Woods-Powell CT, Tessaro I, et al. Cancer screening practices among women in a community health center population. Am J Prev Med. 1996;12:351–7. [PubMed]
2. Cooper GS, Doug Kou T. Underuse of colorectal cancer screening in a cohort of Medicare beneficiaries. Cancer. 2008;112:293–9. [PubMed]
3. Ko CW, Kreuter W, Baldwin LM. Persistent demographic differences in colorectal cancer screening utilization despite Medicare reimbursement. BMC Gastroenterol. 2005;5:10. [PMC free article] [PubMed]
4. McBean AM, Yu X. The underuse of screening services among elderly women with diabetes. Diabetes Care. 2007;30:1466–72. [PubMed]
5. Demark-Wahnefried W, Strigo T, Catoe K, Conaway M, Brunetti M, Rimer BK, et al. Knowledge, beliefs, and prior screening behavior among blacks and whites reporting for prostate cancer screening. Urology. 1995;46:346–51. [PubMed]
6. Kang SH, Bloom JR, Romano PS. Cancer screening among African-American women: their use of tests and social support. Am J Public Health. 1994;84:101–3. [PubMed]
7. Suarez L, Lloyd L, Weiss N, Rainbolt T, Pulley L. Effect of social networks on cancer-screening behavior of older Mexican-American women. J Natl Cancer Inst. 1994;86:775–9. [PubMed]
8. Taylor VM, Thompson B, Montano DE, Mahloch J, Johnson K, Li S. Mammography use among women attending an inner-city clinic. J Cancer Educ. 1998;13:96–101. [PubMed]
9. Allen JD, Sorensen G, Stoddard AM, Peterson KE, Colditz G. The relationship between social network characteristics and breast cancer screening practices among employed women. Ann Behav Med. 1999;21:193–200. [PubMed]
10. Suarez L, Ramirez AG, Villarreal R, Marti J, McAlister A, Talavera GA, et al. Social networks and cancer screening in four U.S. Hispanic groups. Am J Prev Med. 2000;19:47–52. [PubMed]
11. Seow A, Huang J, Straughan PT. Effects of social support, regular physician and health-related attitudes on cervical cancer screening in an Asian population. Cancer Causes Control. 2000;11:223–30. [PubMed]
12. Allen JD, Stoddard AM, Sorensen G. Do social network characteristics predict mammography screening practices? Health Educ Behav. 2008;35:763–76. [PMC free article] [PubMed]
13. Stockdale SE, Keeler E, Duan N, Derose KP, Fox SA. Costs and cost-effectiveness of a church-based intervention to promote mammography screening. Health Serv Res. 2000;35:1037–57. [PMC free article] [PubMed]
14. Allen JD, Stoddard AM, Mays J, Sorensen G. Promoting breast and cervical cancer screening at the workplace: results from the Woman to Woman Study. Am J Public Health. 2001;91:584–90. [PubMed]
15. Derose KP, Duan N, Fox SA. Women’s receptivity to church-based mobile mammography. J Health Care Poor Underserved. 2002;13:199–213. [PubMed]
16. Katz ML, Tatum C, Dickinson SL, Murray DM, Long-Foley K, Cooper MR, et al. Improving colorectal cancer screening by using community volunteers: results of the Carolinas cancer education and screening (CARES) project. Cancer. 2007;110:1602–10. [PMC free article] [PubMed]
17. Paskett E, Tatum C, Rushing J, Michielutte R, Bell R, Long Foley K, et al. Randomized trial of an intervention to improve mammography utilization among a triracial rural population of women. J Natl Cancer Inst. 2006;98:1226–37. [PMC free article] [PubMed]
18. Christakis NA, Fowler JH. The spread of obesity in a large social network over 32 years. N Engl J Med. 2007;357:370–9. [PubMed]
19. Christakis NA, Fowler JH. The collective dynamics of smoking in a large social network. N Engl J Med. 2008;358:2249–58. [PMC free article] [PubMed]
20. Rosenquist JN, Murabito J, Fowler JH, Christakis NA. The spread of alcohol consumption behavior in a large social network. Ann Intern Med. 2010;152:426–33. [PMC free article] [PubMed]
21. Dawber T. The Framingham Study: the epidemiology of atherosclerotic disease. Cambridge, MA: Harvard University Press; 1980.
22. Feinleib M, Kannel WB, Garrison RJ, McNamara PM, Castelli WP. The Framingham Offspring Study. Design and preliminary data. Prev Med. 1975;4:518–25. [PubMed]
23. Splansky GL, Corey D, Yang Q, Atwood LD, Cupples LA, Benjamin EJ, et al. The Third Generation Cohort of the National Heart, Lung, and Blood Institute’s Framingham Heart Study: design, recruitment, and initial examination. Am J Epidemiol. 2007;165:1328–35. [PubMed]
24. ACOG committee opinion. Routine cancer screening. Number 185, September 1997 (replaces no. 128, October 1993). Committee on Gynecologic Practice. American College of Obstetricians and Gynecologists. Int J Gynaecol Obstet. 1997;59:157–61. [PubMed]
25. Smith RA, Mettlin CJ, Davis KJ, Eyre H. American Cancer Society guidelines for the early detection of cancer. CA Cancer J Clin. 2000;50:34–49. [PubMed]
26. National Cancer Institute. National Cancer Advisory Board issues mammography screening recommendations. Vol. 2009. National Institutes of Health; 1997. [press release]
27. U.S. Preventive Services Task Force. Guide to clinical preventive services: report of the US Preventive Services Task Force. 2. Baltimore, MD: Williams & Wilkins; 1996.
28. U.S. Preventive Services Task Force. Screening for breast cancer: recommendations and rationale. Ann Intern Med. 2002;137:344–6. [PubMed]
29. Middleton RG, Thompson IM, Austenfeld MS, Cooner WH, Correa RJ, Gibbons RP, et al. Prostate Cancer Clinical Guidelines Panel Summary report on the management of clinically localized prostate cancer. The American Urological Association. J Urol. 1995;154:2144–8. [PubMed]
30. Levin B, Bond JH. Colorectal cancer screening: recommendations of the U.S. Preventive Services Task Force. American Gastroenterological Association Gastroenterology. 1996;111:1381–4. [PubMed]
31. Zhang J, Yu KF. What’s the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA. 1998;280:1690–1. [see comments] [PubMed]
32. Lerman C, Rimer B, Trock B, Balshem A, Engstrom PF. Factors associated with repeat adherence to breast cancer screening. Prev Med. 1990;19:279–90. [PubMed]
33. Calvocoressi L, Kasl SV, Lee CH, Stolar M, Claus EB, Jones BA. A prospective study of perceived susceptibility to breast cancer and nonadherence to mammography screening guidelines in African American and White women ages 40 to 79 years. Cancer Epidemiol Biomarkers Prev. 2004;13:2096–105. [PubMed]
34. Murabito JM, Evans JC, Larson MG, Kreger BE, Splansky GL, Freund KM, et al. Family breast cancer history and mammography: Framingham Offspring Study. Am J Epidemiol. 2001;154:916–23. [PubMed]
35. Calle EE, Miracle-McMahill HL, Moss RE, Heath CW., Jr Personal contact from friends to increase mammography usage. Am J Prev Med. 1994;10:361–6. [PubMed]
36. Gotay CC, Wilson ME. Social support and cancer screening in African American, Hispanic, and Native American women. Cancer Pract. 1998;6:31–7. [PubMed]
37. Barry MJ. Screening for prostate cancer--the controversy that refuses to die. N Engl J Med. 2009;360:1351–4. [PubMed]
38. Farrell MH, Murphy MA, Schneider CE. How underlying patient beliefs can affect physician-patient communication about prostate-specific antigen testing. Eff Clin Pract. 2002;5:120–9. [PubMed]
39. Ross LE, Taylor YJ, Richardson LC, Howard DL. Patterns in prostate-specific antigen test use and digital rectal examinations in the Behavioral Risk Factor Surveillance System, 2002–2006. J Natl Med Assoc. 2009;101:316–24. [PubMed]
40. Taskila T, Wilson S, Damery S, Roalfe A, Redman V, Ismail T, et al. Factors affecting attitudes toward colorectal cancer screening in the primary care population. Br J Cancer. 2009;101:250–5. [PMC free article] [PubMed]
41. Holden D, Harris R, Porterfield D, Jonas D, Morgan L, Reuland D, et al. RTI International-University of North Carolina Evidence-based Practice Center, Contract No. 290-2007-10056-I. Rockville, MD: Agency for Healthcare Research and Quality; Feb, 2010. Enhancing the Use and Quality of Colorectal Cancer Screening.
42. Smith KP, Christakis NA. Social Networks and Health. Annu Rev Sociol. 2008;34:405–29.
43. Fowler JH, Christakis NA. Dynamic spread of happiness in a large social network: longitudinal analysis over 20 years in the Framingham Heart Study. BMJ. 2008;337:a2338. [PMC free article] [PubMed]
44. O’Malley AJ, Christakis NA. Longitudinal analysis of large social networks: estimating the effect of health traits on changes in friendship ties. Stat Med. 2010 In press. [PMC free article] [PubMed]
45. Seeff L, King J, Pollack L, Williams K. Increased use of colorectal cancer tests--United States, 2002 and 2004. MMWR. 2006;55:308–11. [PubMed]
46. Ryerson AB, Miller J, Eheman C, White M. Use of mammograms among women aged ≥ 40, United States, 2000–2005. MMWR. 2007;56:49–51. [PubMed]
47. Sirovich BE, Schwartz LM, Woloshin S. Screening men for prostate and colorectal cancer in the United States: does practice reflect the evidence? JAMA. 2003;289:1414–20. [PubMed]
48. Rauscher GH, Johnson TP, Cho YI, Walk JA. Accuracy of self-reported cancer-screening histories: a meta-analysis. Cancer Epidemiol Biomarkers Prev. 2008;17:748–57. [PubMed]