Annals of Oncology Advance Access originally published online on September 4, 2007
Annals of Oncology 2008 19(1):156-162; doi:10.1093/annonc/mdm413
epidemiology |
Influence of education level on cancer survival in Sweden
1 Center for Family and Community Medicine, Karolinska Institute, Huddinge
2 Department of Radiation Sciences, Oncology, Umeå University, Umeå, Sweden
3 Division of Molecular Genetic Epidemiology, German Cancer Research Center, Heidelberg, Germany
* Correspondence to: Dr S. K. Hussain, Center for Family and Community Medicine, Karolinska Institute, Alfred Nobels allé 12, SE-141 83 Huddinge, Sweden. Tel: +46-8-524-886-68; Fax: +46-8-524-887-06; E-mail: shehnaz.hussain{at}ki.se
 |
Abstract |
|---|
 Top Abstract introductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
Background: While cancer survival at several sites has historically been shown to vary by education level, a current comprehensive assessment of survival following a cancer diagnosis in Sweden, a country with universal health care and cancer screening, has yet to be carried out.
Methods: Using the 2006 update of the Swedish Family-Cancer Database and Cox's proportional hazards regression methods, we calculate the adjusted hazard ratio (HR) and 95% confidence interval to estimate the influence of education level on site-specific cancer survival.
Results: Significant positive associations between education level and cancer survival were observed following a diagnosis of upper aerodigestive track cancer, colon cancer, pancreatic cancer, lung cancer, kidney cancer, urinary bladder cancer, melanoma, non-Hodgkin's lymphoma, breast cancer, endometrial cancer, cervical cancer, prostate cancer, and testicular cancer. Although the HRs differed between cancer sites, compared with women and men completing <9 years of education, university graduates were associated with a significant 40% improved survival for all cancer sites combined.
Conclusions: Survival differences by education level were observed for both indolent and aggressive malignancies.
Key words: cancer, education, survival
|
introduction |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
Approximately 21 000 women and men died from cancer in Sweden in the year 2004. Education level may be related to behaviors, health conditions, or access to knowledge and resources that have direct and indirect impacts on cancer survival. Influences of survival that may be related to education include stage at diagnosis [1–7], timeliness and type of cancer treatment [1, 2], and level of psychosocial support [3–6]. There is ample literature describing socioeconomic differences in cancer survival in countries around the world with various systems of health care [7]. Despite the fact that Sweden, like many European countries, has a system of universal health coverage which collects nominal fees from its users and in theory delivers a standardized quality of health care to all users, there is some evidence that a socioeconomic gradient may exist in cancer survival.
A few studies have directly or indirectly addressed the association between occupation and cancer survival in Sweden, with discordant results. One study of Swedes born from 1896 to 1940 found that compared with blue-collar occupations, white-collar occupations were associated with increased survival from melanoma [8], cancers of the breast and cervix among women, and cancer of the rectum among men [9]. These authors observed no occupational differences in survival from lung, stomach, or pancreatic cancers [9]. Another Swedish study of younger men found that professional occupations, compared with manual occupations, were associated with increased respiratory cancer incidence and more pronounced increased respiratory cancer mortality, suggesting a negative association with survival. Gastrointestinal, genitourinary, skin, nervous, hematopetic, or lymphatic cancer incidence and mortality were largely unassociated with occupation in this study [10].
Education and occupation are not always interchangeable proxies of the same underlying phenomena in health outcomes research [11]. For example, occupation title may reflect exposures at the work place [12], while education level could reflect health literacy or one's ability to self-manage disease [13]. While no prior Swedish studies have assessed the role of education in cancer survival, in neighboring Norway and Finland, education was positively associated with survival from melanoma [14], prostate cancer [15], hematologic cancers [14], respiratory cancers [14], and breast cancer in some studies [6, 14] but not others [16]. It has yet to be determined to what extent a similar education gradient in cancer survival exists the Swedish population. Improvements in early detection and treatment of cancer in the last few decades in Sweden have led to improved survival across many cancer sites [17–22], however, it is unclear if these improvements have been experienced equally by individuals of all levels of education.
On account of limited data regarding the association between education level and cancer survival in Sweden, this study was conducted on a recent, near complete, cohort of the Swedish population for 25 common cancer sites, separate for women and men. Survival analysis techniques were used to account for several potential confounding factors. The objective was that this comprehensive assessment may help identify potential disparities in cancer survival in Sweden, and prompt future targeted investigations into possible factors influencing such disparities.
|
patients and methods |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
study population
The 2006 update to the Swedish Family-Cancer Database [23] was the cohort used for this study. This Database was first created in the mid-1990s by linking census information, death notifications, and the administrative family register at Statistics Sweden, to the Swedish Cancer Registry, using a technical identification number. Women and men included in this study were a subset of those captured by the Family-Cancer Database. This study population was restricted to cancer-free individuals who were residing in Sweden at the study start date of 1 July 1990, covered by the 1990 census. Furthermore, the population was restricted to individuals born from 1926 to 1960 (ages 30–64 years at the study start date), to help insure that their lifetime highest education level would be captured by the 1990 census.
education information
Education information was complete for 83% of eligible individuals for this study. The national census categorized education into seven categories which were collapsed in this study to form four meaningful groups according to total years of education: (i) <9 years, (ii) 9–11 years, (iii) 12–13 years, and (iv) university graduate, including medical and research doctorates.
cancer ascertainment
Invasive cancer diagnoses were based on the four-digit code according to the International Classification of Diseases (ICD) revision seven. The Swedish Cancer Registry is currently using the ICD revision 10 diagnostic coding system but for comparability with the earlier years, codes were translated to the ICD revision seven. Information on incident cancers is close to 100% complete for the eligible study population on account of compulsory nationwide cancer registration by clinicians, pathologists, and cytologists [24]. Cancer sites and the corresponding first three digits of the ICD revision seven codes included in this study are as follows: breast (170), endometrium (172), ovary (175), cervix (171), prostate (177), testis (178), upper aerodigestive track (140, 141, 143–148, and 161), esophagus (150), stomach (151), small intestine (152), colon (153), rectum (154 except 1541), liver (155 and 156), pancreas (157), lung (162 and 163), kidney (180), urinary bladder (181), melanoma (190), nervous system (193), thyroid gland (194), connective tissue (197), non-Hodgkin's lymphoma (200 and 202), Hodgkin's disease (201), myeloma (203), and leukemia (204–209). These cancer sites represent the most common sites in the Database with a sufficient number of fatalities for analysis. Squamous cell skin cancer, for example, was excluded on account of low associated mortality.
cancer fatalities
Information regarding the date and cause of a cancer fatality was obtained from the Swedish cause of death register which contains the date and cause of deaths for all deceased individuals residing in Sweden at the time of death, reported by the treating physician. Cancer site-specific mortality was defined by the ICD revision six, seven, eight, nine, and 10 codes for a main or contributory cause of death.
statistical analyses
Survival time was defined as the interval between the month and year of a first invasive cancer diagnosis occurring on or after 1 July 1990 and the month and year of the first occurrence of death, emigration, or the study end date, 31 December 2004. The hazard ratio (HR) and 95% confidence interval (CI) for the adjusted association between each education level and cancer mortality was calculated using Cox's proportional hazards regression methods [25]. Separate regression models were used for each cancer site separate by sex, including only individuals who were diagnosed with each site-specific cancer during the study follow-up (1 July 1990 to 31 December 2004). Women and men completing <9 years of education comprised the largest category of incident cancers for men (47 675) and the second largest for women (38 586), and these strata were used as the reference category for the HR calculations. Several time-varying and time-fixed covariates were considered in the regression models to control for potential confounding, including family history of site-specific cancer (zero, one, or two first-degree relatives), age (by 5-year categories), time period (1990–1999 or 2000–2005), and geographic residence (Stockholm area, Göteborg-Malmö, Götaland, Svealand, or Norrland). Additionally, age at first birth (<21, 21–29, 30–34, or 
35 years) and parity (zero, one to two, or three or more) were considered in the breast, cervical, endometrial, and ovarian cancer regression models, to control for aspects of hormonal influence. The final models were stratified by the only two covariates that consistently and meaningfully altered the HRs, which were age and time period at diagnosis. Linear trend tests and corresponding P values were computed for each model.
Subanalyses stratified by birth cohort, 1926–1939 versus 1940–1960, were also conducted. On account of the large change associated with education over time (i.e. a small percentage of university educated individuals in the early birth cohorts and a small percentage of <9 years educated individuals in the later birth cohorts) education categories were collapsed to form a binary variable, 
11 years (reference category) versus 12 years of education.
Several checks were conducted to determine whether the proportional hazards assumption was reasonable for each multivariate model. Interaction terms for survival time and education were modeled and tested for a statistically significant coefficient at an alpha value of 0.05. Scaled Schoenfeld's residuals were calculated for education in each model [26], regressed against survival time, and tested for a nonzero slope at an alpha value of 0.05. The plots of the Schoenfeld's residuals against survival time were also visually inspected with a smooth line. These checks indicated that the proportional assumption was reasonable for all models.
|
results |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
In total, 1 580 570 women and 1 667 460 men were at risk for a first incident cancer diagnosis and subsequent fatality in 1990 (Table 1). In all, 108 586 women and 106 455 men were diagnosed with an invasive cancer during the study period and 23% and 29% experienced a cancer fatality, respectively. The mean length of follow-up time for each cancer case differed by cancer site and sex. Follow-up extended for the longest period of time for thyroid cancer (women, 80 months) and testicular cancer (men, 97 months), and for the shortest period of time for liver cancer (women, 10 months) and pancreatic cancer (men, 9 months).
|
Statistically significant improved survival (decreased HRs) for more highly educated women, compared with women completing <9 years of education, was observed for breast cancer, endometrial cancer, cervical cancer, upper aerodigestive track cancer, colon cancer, pancreatic cancer, lung cancer, kidney cancer, urinary bladder cancer, melanoma, nervous system cancer, non-Hodgkin's lymphoma, and leukemia (Table 2). Compared with men completing <9 years of education, more highly educated men were associated with improved survival from prostate cancer, upper aerodigestive track cancer, colon cancer, rectum cancer, liver cancer, pancreatic cancer, lung cancer, kidney cancer, urinary bladder cancer, melanoma, nervous system cancer, non-Hodgkin's lymphoma, myeloma, and leukemia (Table 3). Education level was not significantly associated with survival from stomach cancer, small intestine cancer, thyroid cancer, connective tissue cancer, and Hodgkin's disease for either women or men.
|
|
Stratified analyses did not reveal any significant survival differences by birth cohort, with one exception. Among women in the early birth cohort, a higher level of education (
12 years versus 11 years) was associated with an improved survival from endometrial cancer, HR = 0.47 (95% CI 0.32–0.67), which was not observed among women in the later birth cohort, HR = 1.03 (95% CI 0.67–1.61). |
discussion |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
We observed statistically significant positive associations between education level and survival from several indolent and aggressive malignancies, in both women and men in the Swedish population. The survival differences by education level were the most pronounced for urinary bladder cancer, melanoma, and non-Hodgkin's lymphoma. Furthermore, several possible screen-detected cancers (i.e. breast cancer, cervical cancer, prostate cancer, and colon cancer) showed pronounced survival differences by education level. Despite the fact that the percentage of more highly educated individuals has increased over the years, the site-specific age-adjusted HRs across birth cohorts remained largely the same.
Our results corroborate previous reports from countries with universal health systems that collect nominal charges from its users, similar to Sweden, for invasive breast cancer [6, 8, 14], melanoma [8, 14], digestive/gastrointestinal cancers [14, 27], prostate cancer [15], cervical cancer [8], head and neck cancers [28], hematologic cancers [14], and respiratory cancers [10, 14, 28]. Several personal and community-level characteristics that are associated with education, or socioeconomic position, may help to explain some of these observed associations.
Population-based cancer screening leads to early detection and improved survival from several cancers [20, 29]. In Sweden, population-based screening has been ongoing for breast cancer since the mid-1980s and cervical cancer since the early 1970s. Despite the effort to cover all at-risk women, 
20% of eligible invited women were not participating in mammography screening in 1992 [17, 30]. A small proportion, 30%, of Pap smears for cervical cancer screening are conducted by the Swedish national screening program, with the remainder conducted by private practice [31]. Coverage by either public or private cervical screening is estimated to be 86% in certain urban setting [32]. Furthermore, opportunistic screening for prostate and colon cancers has been ongoing since the early 1990s. Screening attendance in Sweden has been associated with indicators of high socioeconomic position [33–36], which may provide one explanation for the observed survival differences by education level for possible screen-detected cancers.
Alternatively, lead time bias may also be a plausible explanation if more highly educated individuals, on account of screening and early detection, simply have a longer interval between diagnosis and death. However, we calculated case fatality rates by education level for breast, cervix, colon, and prostate cancers, separately for women and men, and observed significant inverse associations (P for trend < 0.001) at all sites which suggests real survival differences not attributable to a lead time bias.
Tumor stage at clinical presentation is an important prognostic indicator. There is evidence to indicate that an early stage at diagnosis is associated with increased socioeconomic position [37–43] and improved survival [37–39, 42]. However, several studies that have adjusted for tumor stage still report socioeconomic gradients in cancer survival from potential screen-detected and nonscreen-detected cancers, suggesting other influences [7]. Treatment type and treatment quality may also differ by markers of socioeconomic position and alter cancer prognosis, which has been observed for esophageal cancer [1], colon cancer [44], and breast cancer [2]. Furthermore, previous studies on head and neck cancer [45] and colon cancer [46] indicate that a low prevalence of comorbidities may also be associated with an improved survival. In Sweden, like many developed countries, there is some indication that individuals of higher socioeconomic position have a lower burden of chronic disease such as cardiovascular disease and diabetes [47, 48] which may ultimately lead to a greater cancer treatment success and improved survival.
There is ongoing debate about the impact of psychosocial support on cancer survival [49–53]; however, evidence at the population level [5, 6, 54, 55] and molecular level [49, 56–58] is accumulating. For example, marriage, which is one indicator for positive social support, was associated with improved survival for bladder cancer [59], breast cancer [5, 60], and prostate cancer [15, 61]. Studies also show that marriage dissolution is associated with low levels of education in women [62, 63]. The positive psychosocial support associated with being married may be biased towards women of higher education levels, and thus to some degree explain the observed improved survival associated with these education strata.
Although women and men with <9 years of education comprised a large proportion of our study population, it is clear that there is a trend towards increasing education with time in Sweden. However, a large proportion of individuals with a low education level are currently at risk for cancer diagnosis and mortality. For example, individuals born in the 1940s, 25%–30% of whom have <9 years of education, are currently (in 2007) of or below peak age for most cancer diagnoses, including colon cancer, lung cancer, prostate cancer, and urinary bladder cancer [64]. This illustrates the point that the cancer survival differences by education level which were observed in this study are a current issue for the Swedish population. Furthermore, our analyses stratified by birth cohort indicate that the observed survival differences are similar for nearly all cancer sites for cohorts that differ by education composition.
Education is one, but not the only marker of socioeconomic position that may influence cancer survival. For example, a recent published cohort study from Sweden showed that above and beyond the influence of one's lifetime highest education level, parental occupation during key childhood years may influence future adult mortality from smoking-related cancers and stomach cancer [65]. Furthermore, mental ability (IQ) in early life may be inversely associated with subsequent adult mortality [66]. Although cancer incidence could not explain this mortality difference [67], cancer survival might.
Several limitations must be considered when interpreting these study results. Importantly, a lack of data on tumor stage at diagnosis and type of treatment, known to influence cancer survival, precludes quantification of the effect of these factors in cancer survival. Time period adjustment helped to account for gross changes in cancer treatment over time; however it does not provide insight into the equality or inequality of treatment according to education level. Another limitation is that causes of death related to a cancer, but not attributed to the tumor itself, such as death due to treatment side-effects, may have resulted in an underestimation of cancer deaths in this study. However, as long as this underestimation was similarly distributed between women and men of different educational backgrounds, the HRs should not be biased. Furthermore, it is possible that women and men may have advanced their education after 1990, which would cause a misclassification of their education level. However, the fact that we restricted this study population to individuals aged 30–64 at the start of the study offers some assurance that the majority of individuals would have completed their lifetime highest education level, minimizing the possibility of this misclassification bias.
This study is strengthened by the near complete (97%) nationwide coverage of women and men born from 1926 to 1960 and precision of the HR estimates due to the large number of cancer diagnoses and length of follow-up time. The results demonstrate that overall and site-specific cancer survival was positively associated with education level. Interestingly, survival differences were observed for highly fatal cancers (i.e. pancreatic cancer and lung cancer), as well as generally treatable tumors with relatively low associated fatalities (i.e. urinary bladder cancer and kidney cancer). Further research is warranted to elucidate the pathways and mechanisms underlying these observed survival differences. Such information may help to improve cancer survival for women and men with low levels of education, thereby decreasing years of life loss due to premature death.
|
Acknowledgements |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
The Family-Cancer Database was created by linking registers maintained at Statistics Sweden and the Swedish Cancer Registry and is supported by Deutsche Krebshilfe (107406), the Swedish Cancer Society (050494), the Swedish Council for Working Life (2005-0039) and Social Research and European Union (LSHC-CT-2004-503465).
Received for publication April 19, 2007. Revision received July 15, 2007. Accepted for publication July 16, 2007.
|
References |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
1. van Vliet EPM, Eijkemans MJC, Steyerberg EW, et al. The role of socio-economic status in the decision making on diagnosis and treatment of oesophageal cancer in the Netherlands. Br J Cancer (2006) 95:1180–1185.[CrossRef][Web of Science][Medline]
2. Griggs JJ, Culakova E, Sorbero MES, et al. Effect of patient socioeconomic status and body mass index on the quality of breast cancer adjuvant chemotherapy. J Clin Oncol (2007) 25:277–284.
3. Spiegel D. Effects of psychotherapy on cancer survival. Nat Rev Cancer (2002) 2:383–389.[CrossRef][Web of Science][Medline]
4. Spiegel D, Giese-Davis J. Depression and cancer: mechanisms and disease progression. Biol Psychiatry (2003) 54:269–282.[CrossRef][Web of Science][Medline]
5. Spiegel D, Kraemer HC, Bloom JR, Gottheil E. Effect of psychosocial treatment on survival of patients with metastatic breast-cancer. Lancet (1989) 2:888–891.[CrossRef][Web of Science][Medline]
6. Lehto US, Ojanen M, Dyba T, et al. Baseline psychosocial predictors of survival in localised breast cancer. Br J Cancer (2006) 94:1245–1252.[CrossRef][Web of Science][Medline]
7. Kogevinas M, Pearce N, Susser M, et al. Social Inequalities and Cancer (1997) Lyon, France: IARC Scientific Publications.
8. Vagero D, Persson G. Risks, survival and trends of malignant-melanoma among white and blue collar workers in Sweden. Soc Sci Med (1984) 19:475–478.[CrossRef][Web of Science][Medline]
9. Vagero D, Persson G. Cancer survival and social-class in Sweden. J Epidemiol Community Health (1987) 41:204–209.
10. Rosengren A, Wilhelmsen L. Cancer incidence, mortality from cancer and survival in men of different occupational classes. Eur J Epidemiol (2004) 19:533–540.[CrossRef][Web of Science][Medline]
11. Geyer S, Hemstrom O, Peter R, Vagero D. Education, income, and occupational class cannot be used interchangeably in social epidemiology. Empirical evidence against a common practice. J Epidemiol Community Health (2006) 60:804–810.
12. Weiderpass E, Pukkala E, Kauppinen T, et al. Breast cancer and occupational exposures in women in Finland. Am J Ind Med (1999) 36:48–53.[CrossRef][Web of Science][Medline]
13. Goldman DP, Smith JP. Can patient self-management help explain the SES health gradient? Proc Natl Acad Sci USA (2002) 99:10929–10934.
14. Kravdal O. Social inequalities in cancer survival. Popul Stud (2000) 54:1–18.[CrossRef]
15. Harvei S, Kravdal O. The importance of marital and socioeconomic status in incidence and survival of prostate cancer—an analysis of complete Norwegian birth cohorts. Prev Med (1997) 26:623–632.[CrossRef][Web of Science][Medline]
16. Lund E, Jacobsen BK. Education and breast-cancer mortality—experience from a large Norwegian cohort study. Cancer Causes Control (1991) 2:235–238.[CrossRef][Web of Science][Medline]
17. Bjurstam N, Bjorneld L, Duffy SW, et al. The Gothenburg breast screening trial—first results on mortality, incidence, and made of detection for women ages 39–49 years at randomization. Cancer (1997) 80:2091–2099.[CrossRef][Web of Science][Medline]
18. Dickman PW, Adami HO. Interpreting trends in cancer patient survival. J Intern Med (2006) 260:103–117.[CrossRef][Web of Science][Medline]
19. Hansson LE, Sparen P, Nyren O. Survival in stomach cancer is improving—results of a nationwide population-based Swedish study. Ann Surg (1999) 230:162–169.[CrossRef][Web of Science][Medline]
20. Laara E, Day NE, Hakama M. Trends in mortality from cervical-cancer in the Nordic countries—association with organized screening programs. Lancet (1987) 1:1247–1249.[Web of Science][Medline]
21. Malmstrom PU, Thorn M, Lindblad P, et al. Increasing survival of patients with urinary-bladder cancer—a nationwide study in Sweden 1960–1986. Eur J Cancer (1993) 29A:1868–1872.[CrossRef]
22. Sundelof M, Ye M, Dickman PW, Lagergren J. Improved survival in both histologic types of oesophageal cancer in Sweden. Int J Cancer (2002) 99:751–754.[CrossRef][Web of Science][Medline]
23. Hemminki K, Li XJ, Plna K, et al. The nation-wide Swedish family-cancer database—updated structure and familial rates. Acta Oncol (2001) 40:772–777.[CrossRef][Web of Science][Medline]
24. Center for Epidemiology Cancer Incidence in Sweden 2004 (2006) The National Board of Health and Welfare, Stockholm. http://www.sos.se/.
25. Cox DR. Regression models and life-tables. J R Stat Soc Ser B Stat Methodol (1972) 34:187–202.
26. Schoenfeld D. Partial residuals for the proportional hazards regression-model. Biometrika (1982) 69:239–241.
27. Fontana V, Decensi A, Orengo MA, et al. Socioeconomic status and survival of gastric cancer patients. Eur J Cancer (1998) 34:537–542.[CrossRef][Web of Science][Medline]
28. Mackillop WJ, ZhangSalomons J, Groome PA, et al. Socioeconomic status and cancer survival in Ontario. J Clin Oncol (1997) 15:1680–1689.[Abstract]
29. Vainio H, Bianchini F. IARC Handbook of Cancer Screening: Breast Cancer Screening (2002) Lyon, France: IARCPress.
30. Olsson S, Andersson I, Karlberg I, et al. Implementation of service screening with mammography in Sweden: from pilot study to nationwide programme. J Med Screen (2000) 7:14–18.
31. Dillner J. Cervical cancer screening in Sweden. Eur J Cancer (2000) 36:2255–2259.[CrossRef][Web of Science][Medline]
32. Rodvall Y, Kemetli L, Tishelman C, Tornberg S. Factors related to participation in a cervical cancer screening programme in urban Sweden. Eur J Cancer Prev (2005) 14:459–466.[CrossRef][Web of Science][Medline]
33. Lagerlund M, Hedin A, Sparen P, et al. Attitudes, beliefs, and knowledge as predictors of nonattendance in a Swedish population-based mammography screening program. Prev Med (2000) 31:417–428.[CrossRef][Web of Science][Medline]
34. Banks E, Beral V, Cameron R, et al. Comparison of various characteristics of women who do and do not attend for breast cancer screening. Breast Cancer Res (2002) 4:6.[CrossRef]
35. Cockburn J, Sutherland M, Cappiello M, Hevern M. Predictors of attendance at a relocatable mammography service for rural women. Aust N Z J Public Health (1997) 21:739–742.[Web of Science][Medline]
36. Sutton S, Bickler G, Sanchoaldridge J, Saidi G. Prospective-study of predictors of attendance for breast screening in Inner London. J Epidemiol Community Health (1994) 48:65–73.
37. Reyes-Ortiz CA, Goodwin JS, Freeman JL. The effect of socioeconomic factors on incidence, stage at diagnosis and survival of cutaneous melanoma. Med Sci Monit (2005) 11:RA163–RA172.[Web of Science][Medline]
38. Brenner H, Mielck A, Klein R, Ziegler H. The role of socioeconomic-factors in the survival of patients with colorectal-cancer in Saarland Germany. J Clin Epidemiol (1991) 44:807–815.[CrossRef][Web of Science][Medline]
39. Stavraky KM, Kincade JE, Stewart MA, Donner AP. The effect of socioeconomic-factors on the early prognosis of cancer. J Chronic Dis (1987) 40:237–244.[CrossRef][Web of Science][Medline]
40. Schwartz KL, Crossley-May H, Vigneau FD, et al. Race, socioeconomic status and stage at diagnosis for five common malignancies. Cancer Causes Control (2003) 14:761–766.[CrossRef][Web of Science][Medline]
41. Hoffman RM, Gilliland FD, Eley JW, et al. Racial and ethnic differences in advanced-stage prostate cancer: the prostate cancer outcomes study. J Natl Cancer Inst (2001) 93:388–395.
42. Chang AE, Karnell LH, Menck HR. The National Cancer Data Base report on cutaneous and noncutaneous melanoma—a summary of 84,836 cases from the past decade. Cancer (1998) 83:1664–1678.[CrossRef][Web of Science][Medline]
43. Bennett CL, Ferreira MR, Davis TC, et al. Relation between literacy, race, and stage of presentation among low-income patients with prostate cancer. J Clin Oncol (1998) 16:3101–3104.
44. Auvinen A. Social-class and colon cancer survival in Finland. Cancer (1992) 70:402–409.[CrossRef][Web of Science][Medline]
45. Piccirillo JF. Importance of comorbidity in head and neck cancer. Laryngoscope (2000) 110:593–602.[CrossRef][Web of Science][Medline]
46. Yancik R, Wesley MN, Ries LAG, et al. Comorbidity and age as predictors of risk for early mortality of male and female colon carcinoma patients—a population-based study. Cancer (1998) 82:2123–2134.[CrossRef][Web of Science][Medline]
47. Manhem K, Dotevall A, Wilhelmsen L, Rosengren A. Social gradients in cardiovascular risk factors and symptoms of Swedish men and women: the Goteborg MONICA Study 1995. J Cardiovasc Risk (2000) 7:359–368.[Web of Science][Medline]
48. Westerling R, Gullberg A, Rosen M. Socioeconomic differences in ‘avoidable’ mortality in Sweden 1986–1990. Int J Epidemiol (1996) 25:560–567.
49. Antoni MH, Lutgendorf SK, Cole SW, et al. Opinion—the influence of bio-behavioural factors on tumour biology: pathways and mechanisms. Nat Rev Cancer (2006) 6:240–248.[CrossRef][Web of Science][Medline]
50. Chow E, Tsao MN, Harth T. Does psychosocial intervention improve survival in cancer? A meta-analysis. Palliat Med (2004) 18:25–31.
51. Linn MW, Linn BS, Harris R. Effects of counseling for late stage cancer-patients. Cancer (1982) 49:1048–1055.[CrossRef][Web of Science][Medline]
52. Edelman S, Lemon J, Bell DR, Kidman AD. Effects of group CBT on the survival time of patients with metastatic breast cancer. Psychooncology (1999) 8:474–481.[CrossRef][Medline]
53. Goodwin PJ, Leszcz M, Ennis M, et al. The effect of group psychosocial support on survival in metastatic breast cancer. N Engl J Med (2001) 345:1719–1726.
54. Fawzy FI, Fawzy NW, Hyun CS, et al. Malignant-melanoma—effects of an early structured psychiatric intervention, coping, and affective state on recurrence and survival 6 years later. Arch Gen Psychiatry (1993) 50:681–689.
55. Kuchler T, Henne-Bruns D, Rappat S, et al. Impact of psychotherapeutic support on gastrointestinal cancer patients undergoing surgery: survival results of a trial. Hepatogastroenterology (1999) 46:322–335.[Medline]
56. Reiche EMV, Morimoto HK, Nunes SOV. Stress and depression-induced immune dysfunction: implications for the development and progression of cancer. Int Rev Psychiatry (2006) 17:515–527.[Web of Science]
57. Reiche EMV, Nunes SOV, Morimoto HK. Stress, depression, the immune system, and cancer. Lancet Oncol (2004) 5:617–625.[CrossRef][Web of Science][Medline]
58. Andersen BL, Farrar WB, Golden-Kreutz DM, et al. Psychological, behavioral, and immune changes after a psychological intervention: a clinical trial. J Clin Oncol (2004) 22:3570–3580.
59. Gore JL, Kwan L, Saigal CS, Litwin MS. Marriage and mortality in bladder carcinoma. Cancer (2005) 104:1188–1194.[CrossRef][Web of Science][Medline]
60. Osborne C, Ostir GV, Du XL, et al. The influence of marital status on the stage at diagnosis, treatment, and survival of older women with breast cancer. Breast Cancer Res Treat (2005) 93:41–47.[CrossRef][Web of Science][Medline]
61. Krongrad A, Lai H, Burke MA, et al. Marriage and mortality in prostate cancer. J Urol (1996) 156:1696–1700.[CrossRef][Web of Science][Medline]
62. Martin SP. Trends in marital dissolution by women's education in the United States. Demogr Res (2006) 15:537–559.[CrossRef][Web of Science]
63. White L, Rogers SJ. Economic circumstances and family outcomes: a review of the 1990s. J Marriage Fam (2000) 62:1035–1051.[CrossRef][Web of Science]
64. The National Board of Health and Welfare CfE. Cancer Incidence in Sweden (2004) http://www.socialstyrelsen.se/Statistik/statistik_amne/Cancer(26 January 2006, date last accessed).
65. Lawlor DA, Sterne JAC, Tynelius P, et al. Association of childhood socioeconomic position with cause-specific mortality in a prospective record linkage study of 1,839,384 individuals. Am J Epidemiol (2006) 164:907–915.
66. Batty GD, Deary IJ, Gottfredson LS. Premorbid (early life) IQ later mortality risk: systematic review. Ann of Epidemiol (2007) 17:278–288.[CrossRef]
67. Batty GD, Wennerstad KM, Smith GD, et al. IQ in early adulthood and later cancer risk: cohort study of one million Swedish men. Ann Oncol (2007) 18:21–28.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||