Abstract
Background
The association between stress and cancer incidence has been studied for more than seven decades. Despite plausible biological mechanisms and evidence from laboratory studies, findings from clinical research are conflicting. The objective of this study was to examine the association between PTSD and various cancer outcomes.
Methods
This nation-wide cohort study included all Danish-born residents of Denmark from 1995 – 2011. The exposure was PTSD diagnoses (n = 4,131). The main outcomes were cancer diagnoses including: 1) all malignant neoplasms; 2) hematologic malignancies; 3) immune-related cancers; 4) smoking- and alcohol-related cancers; 5) cancers at all other sites. Standardized incidence ratios (SIR) were calculated.
Results
Null associations were found between PTSD and nearly all cancer diagnoses examined, both overall (SIR for all cancers = 1.0, 95% confidence interval (CI) = 0.88, 1.2) and in analyses stratified by gender, age, substance abuse history and time since PTSD diagnosis.
Conclusions
This study is the most comprehensive examination to date of PTSD as a predictor of many cancer types. Our data show no evidence of an association between PTSD and cancer in this nationwide cohort.
Keywords: stress disorders, posttraumatic, neoplasms, cohort studies
INTRODUCTION
Posttraumatic stress disorder (PTSD) is a prevalent public health problem associated with significant costs to individuals and society.(1) Cancer, as a potential sequelae of chronic or severe stress, has been discussed in scientific publications during the past seven decades and even longer in historical literature.(2–4) Biological mechanisms proposed for this association include impairment of immune system function resulting from stress, which could impact the body’s ability to fight tumor growth; alterations in DNA repair, which could promote cancer-related mutations; and the possibility that stress-related hormones play a role in tumor growth.(3, 4) Behavioral mechanisms have been proposed as well, most notably that stress precipitates behaviors that are risk factors for cancer (e.g., smoking, alcohol misuse).(4)
Given the plausible biological mechanisms for this association, stressful events and cancer have been the subjects of much clinical research. Early small case-control studies reported an association of stressful life events with breast cancer and gastric cancer. (5–8) In the Finnish Twin Study (one of the largest studies to date), which combined self-reported data and medical records, specific acute stress events (e.g., death of a husband) were associated with incidence of breast cancer.(9) A population-based study in Denmark found smoking-related malignancies among mothers to be weakly associated with loss of a child.(10)
Despite the few studies that have found associations between stress and cancer and despite the plausible mechanisms proposed, the majority of studies have found no evidence supporting an association.(4) Population-based studies of the association between stress and breast cancer have had null findings.(11) Job stress was not associated with breast cancer risk in the Nurse’s Health Study. (12) A Danish population-based study found that cancer in a child (as the stressor) was not associated with a variety of forms of cancer among parents.(13) Most recently, a population-based study using registry data from Western Australia found no increased incidence of any type of cancer among persons diagnosed with stress or anxiety disorders, compared with the general population.(14)
Research to date is limited by a focus on specific stressful events or selected cancer types, which may explain the inconsistent findings. A recent review of studies of various forms of stress and breast cancer, encompassing epidemiology and molecular biology, called for well-designed cohort studies to elucidate these potentially important associations.(15) We know of no population-based study that has examined posttraumatic stress disorder (PTSD)--the diagnosis given following exposure to a traumatic life event1 --as a risk factor for diagnoses of many types of cancers. The current study fills this important gap in the literature by examining the incidence of various forms of cancer in a nationwide cohort of patients with a prior diagnosis of PTSD. Given the ubiquity of PTSD and cancer and their costs to individuals and society,(1, 16) any observed associations could have meaningful public health implications.
METHODS
We conducted a cohort study using Danish national administrative and healthcare registries to compare the incidence of various forms of cancer among patients with a recorded PTSD diagnosis with that expected in the general population during the same time period. The cumulative base population consisted of Danish-born residents of Denmark from January 1, 1995 through December 31, 2011 (80,607,865 person-years of follow-up).
Data Sources
The Danish Civil Registration System (CRS) contains a unique personal identifier (the central personal registry (CPR) number), date of birth, gender, and additional demographic data for all persons residing in Denmark since 1968.(17) The CRS contains data on vital status and address changes for each resident and is updated daily. The CPR number can be used to link data across all Danish administrative and medical registries.
The Danish Psychiatric Central Research Registry has collected data on inpatient hospitalizations since 1969 and outpatient psychiatric treatment since 1995.(18, 19) It contains treatment dates and up to 20 diagnoses per treatment episode for patients who are admitted to a psychiatric inpatient hospital, receive outpatient care, or are treated in a psychiatric emergency unit.(18, 19) We used the Psychiatric Central Research Registry to create a cohort of all Danish residents with at least one incident ICD-10 diagnosis of severe stress or adjustment disorder from January 1, 1995 to December 31, 2011.(20) Patients with PTSD diagnoses (see Appendix 1 for ICD-10 codes) within that cohort were used for the current study (n = 3143). Validation studies of diagnoses in the Registry (e.g., schizophrenia and affective disorders) have shown high validity measured against computer-generated diagnoses or independent re-interviews.(18, 19) Our own validation study showed that validity is high for PTSD diagnoses recorded in the Psychiatric Registry.(21)
The Danish National Registry of Patients (DNRP) covers all inpatient non-psychiatric hospital treatment in Denmark since 1977 and outpatient and emergency room visits since 1995.(22) Its data were coded according to ICD-8 from 1977 through 1993 and ICD-10 thereafter. We used data from the DNRP to compute a Charlson Comorbidity Index (CCI) score for each patient in our study as a measure of overall physical health status.(23) As well, diagnoses of substance abuse were obtained from the DNRP and used to stratify our analyses (see Appendix 1 for corresponding ICD codes). Patients with PTSD diagnoses in the DNRP only were also included in the current study (n = 988).
The Danish Cancer Registry (DCR) has recorded all incident cancer diagnoses in Denmark since 1943. Diagnoses have been coded according to the ICD-10 since January 1, 2004 and all cancer diagnoses which were coded according to ICD-8 prior to 2004 have been recoded using ICD-10 codes.(24) Validation studies have found that 95% to 98% of records contained in the DCR are valid.(24) We obtained the following cancer diagnoses from the DCR: 1) all malignant neoplasms; 2) hematologic malignancies (e.g., Hodgkin’s lymphoma); 3) immune-related cancers (e.g., liver cancer including cancer of the intrahepatic bile ducts, cervical cancer); 4) smoking- and-alcohol related cancers (e.g., lung, colon, bladder); 5) cancers at all other sites (e.g., ovary, fallopian tube, breast). Appendix 1 lists individual incident cancers and corresponding ICD-10 codes.
Analyses
We calculated the expected number of incident cancer cases after PTSD diagnoses using national rates of incident cancer diagnoses (restricted also to the Danish-born population of Denmark) according to sex, 5-year age groups, and 5-year calendar periods. Multiplying person-years of follow-up by incidence rates yielded the number of cancer cases that would be expected if persons with PTSD diagnoses had the same risk of cancer as the general population.(25) We calculated standardized incidence ratios (SIRs) to measure the association between PTSD and cancer as the ratio of observed to expected cancer cases. Confidence intervals (CIs) were calculated assuming that the observed number of cases for a specific cancer follows a Poisson distribution. Exact confidence limits were calculated when there were fewer than 10 observed cancer cases; otherwise Byar’s approximation was used. Our presentation of results is limited to cancers for which there were 5 or more observed incident cases among persons diagnosed with PTSD during the study period. Presentation of stratified results is also limited to cancer diagnoses with at least 5 incident cases among people with PTSD within each subgroup.
We restricted our analyses to cancers diagnosed one or more years after the PTSD diagnosis date to ensure that the PTSD diagnosis was not linked to the cancer diagnosis. We further restricted our analyses to adults, defined as being age 16 or older. Analyses were stratified by time interval between first PTSD diagnosis and incident cancer diagnosis (1 to <5 years and 5+ years), sex, age at PTSD diagnosis (16 –39 years, 40 – 59 years, and 60+ years), substance abuse diagnoses, and CCI score. All statistical analyses were conducted using SAS version 9.2. The study was approved by the Danish Data Protection Agency (record no. 2012-41-0841) and by the Institutional Review Board at Boston University.
RESULTS
We identified 4,131 adults with a diagnosis of PTSD who had not had a cancer diagnosis one year after their PTSD diagnosis (60% female). PTSD patients were followed for an average of 7.6 years (median follow-up: 6.7 years; range: 1 year through 17 years). Age at PTSD diagnosis ranged from 16 years through 94 years (mean age: 39.4 years; median age: 39 years). Among PTSD patients, 159 cases of cancer were diagnosed one or more years after the PTSD diagnosis.
Table 1 displays associations between PTSD and incidence of cancers in major groups and for specific cancers. An overall null association was found for PTSD and cancer diagnoses (SIR = 1.0, 95% (confidence interval) CI = 0.88, 1.2). Null associations (shown in Tables 1 and 2) were also found for PTSD and all immune-related cancers (SIR = 1.0, 95% CI = 0.77, 1.4); non-melanoma skin cancer (SIR = 1.2, 95% CI = 0.88, 1.6); all smoking and alcohol-related cancers (SIR = 0.99, 95% CI = 0.72, 1.3); lung, bronchial and tracheal cancers (SIR = 1.3, 95% CI = 0.73, 2.0); colon cancer including cancer of the rectosigmoid junction (SIR = 0.59, 95% CI = 0.19, 1.4); cancer at other common sites (SIR = 1.1, 95% CI = 0.87, 1.5); breast cancer (SIR = 1.2, 95% CI = 0.82, 1.7); and uterine cancer (SIR = 1.4, 95% CI = 0.46, 3.3). No meaningful differences in this association were noted in stratified analyses (Table 2).
Table 1.
SIRs for selected incident cancers among patients with PTSD, Denmark, 1995 – 2011.
Association with PTSD |
|||
---|---|---|---|
Observed | Expected | SIR (95% CI) | |
All malignant neoplasms | 159 | 154.4 | 1.0 (0.88, 1.2) |
Any immune-related cancer | 52 | 50.2 | 1.0 (0.77, 1.4) |
Non-melanoma skin cancer | 45 | 37.2 | 1.2 (0.88, 1.6) |
Any smoking or alcohol-related cancer | 42 | 42.3 | 0.99 (0.72, 1.3) |
Lung, bronchial and tracheal cancer | 17 | 13.6 | 1.3 (0.73, 2.0) |
Colon cancer (incl. rectosigmoid junction) | 5 | 8.4 | 0.59 (0.19, 1.4) |
Other common cancer sites | 61 | 53.8 | 1.1 (0.87, 1.5) |
Breast cancer | 32 | 26.6 | 1.2 (0.82, 1.7) |
Ovarian and fallopian tube cancer | 7 | 2.9 | 2.4 (0.96, 4.9) |
Uterine cancer | 5 | 3.5 | 1.4 (0.46, 3.3) |
Table 2.
SIRs for cancer one or more years after PTSD diagnosis, stratified by sex, age at PTSD diagnosis, substance abuse, and induction time.
All malignant neoplasms |
Any immune- related cancer |
Non- melanoma skin cancer |
Any smoking or alcohol-related cancer |
|
---|---|---|---|---|
Sex | ||||
Female | 1.0 (0.84, 1.3) | 0.83 (0.54, 1.2) | 0.95 (0.60, 1.4) | 1.1 (0.67, 1.6) |
Male | 1.0 (0.79, 1.3) | 1.4 (0.92, 2.1) | 1.6 (1.0, 2.4) | 0.93 (0.57, 1.4) |
Age at first PTSD diagnosis, yrs | ||||
16 – 39 | 1.2 (0.79, 1.6) | 0.86 (0.43, 1.5) | 1.5 (0.77, 2.8) | 1.6 (0.52, 3.7) |
40 – 59 | 1.0 (0.84, 1.3) | 1.2 (0.84, 1.7) | 1.2 (0.81, 1.8) | 0.82 (0.52, 1.2) |
60+ | 0.91 (0.60, 1.3) | 0.72 (0.26, 1.6) | 0.85 (0.31, 1.8) | 1.3 (0.68, 2.1) |
Substance abuse | ||||
Yes | 0.93 (0.56, 1.5) | 1.0 (0.41, 2.1) | 1.4 (0.56, 2.9) | 1.6 (0.73, 3.0) |
No | 1.0 (0.88, 1.2) | 1.0 (0.76, 1.4) | 1.2 (0.83, 1.6) | 0.90 (0.62, 1.3) |
Time from PTSD to cancer diagnosis | ||||
1 to < 5 years | 1.1 (0.89, 1.4) | 1.3 (0.83, 1.9) | 1.5 (0.93, 2.3) | 1.0 (0.60, 1.6) |
5+ years | 0.96 (0.77, 1.2) | 0.88 (0.58, 1.3) | 1.0 (0.66, 1.5) | 0.97 (0.63, 1.4) |
Note: Results not presented when fewer than 5 incident cancer cases were identified in a stratification subgroup
Effect of overall health
PTSD did not confer any additional risk for overall cancer incidence among persons with at least one comorbidity (as indicated by a CCI score of 1 or greater). The SIR among those with 1+ comorbidities was 0.60 (95% CI = 0.36, 0.95), while the SIR for those with no comorbid diagnoses was 1.1 (95% CI = 0.95, 1.3).
DISCUSSION
This nationwide cohort study -- the largest to date examining PTSD as a risk factor for a number of cancer outcomes -- showed no evidence of associations. This is consistent with other population-based studies, which reported that stressful life events are generally not associated with cancer incidence.(11–14) In addition to corroborating results of other studies, our large population sample enabled us to conduct important stratified analyses, which provided no strong evidence of associations among select subsamples.
Study strengths include use of data from a large population-based cohort with complete follow-up and little to no selection bias. As well, the large sample and long study period allowed us to examine associations that have not been studied previously. We were able to look at rare cancer outcomes and associations among important subgroups. In addition, we were able to examine validated PTSD diagnoses as a marker for prolonged clinically detected stress triggered by trauma from any source. Previous research has focused predominantly on specific stressful events.
Despite the size of the current study, it was too small to examine associations with rare cancers, particularly in the stratified analyses. Other limitations included an inability to adjust for behavioral risk factors for cancer, such as smoking. Future population-based studies are needed to replicate our results while incorporating these potentially important confounders.
In conclusion, our data shows no evidence of an association between PTSD diagnoses and various forms of cancer in a nationwide cohort overall or among particular population subgroups.
Acknowledgements
This work was supported by a National Institute of Mental Health (1R21MH094551-01A1).
Appendix 1. ICD Codes for Analytic Variables
ICD Code | |
---|---|
Posttraumatic stress disorder (PTSD) | F43.1 |
Substance abuse | ICD-8: 291, 303, 979, 980, 571.09, 571.10, 577.10 T36-T65, F10, G31.2, G62.1, G72.0, G72.1, I42.6, K29.2, K86.0, K70, R78.0-78.9, T51, Z72.1, Z71.4 |
Cancer Type | |
Hematological cancers | |
Hodgkin’s lymphoma | C81 |
Non-Hodgkin’s lymphoma excl. leukemia and myelomatosis | C82-85, C88 |
Multiple myeloma and malignant plasma cell neoplasms | C90 |
Leukemia | C91-C95 |
Malignant neoplasm of lymphoid, hematopoietic and related tissues, unspecified | C96 |
Immune-related cancers | |
Liver including intrahepatic bile ducts* | C22 |
Malignant melanoma including those located in anus and anal canal (morphological code 872-879) | C43 |
Non-melanoma skin cancers | C44 |
Kaposi’s sarcoma | C46, B210 |
Cervix | C53 |
Anus and anal canal excl. malignant melanomas (morphologic code 872-879) and basal cell cancers (morphologic code 809) | C21 |
External female genitalia excluding basal cell carcinomas (morphological code 809) | C51 |
Smoking- and alcohol-related cancers | |
Lip | C00 |
Tongue | C01-02 |
Mouth | C03-06 |
Tonsil and pharynx | C09-C13 |
Other and poorly specified location in lip, oral cavity, and pharynx | C14 |
Larynx | C32 |
Other and poorly specified location in airways and respiratory organs | C39 |
Esophagus | C15 |
Stomach | C16 |
Colon incl. rectosigmoid junction | C18-C19 |
Rectum | C20 |
Pancreas | C25 |
Lung, bronchus and trachea | C33-C34 |
Kidney | C64 |
Renal pelvis | C65 |
Ureter | C66 |
Urinary bladder | C67 |
All other sites | |
Salivary gland | C07-C08 |
Small intestine | C17 |
Gallbladder and bile ducts | C23-C24 |
Other and ill-defined cancers of digestive organs | C26 |
Nasal cavity, middle ear and accessory sinuses | C30-C31 |
Thymus | C37 |
Heart and mediastinum | C381-383, C388 |
Pleura incl. mesothelioma pleura | C384, C450 |
Bone and articular cartilage | C40-C41 |
Mesothelioma | C45.1-C45.9 |
Peripheral nerves and autonomic nervous system | C47 |
Retroperitoneum and peritoneum, and malignant neoplasm of other connective and soft tissue | C48-C49 |
Breast | C50 |
Vagina excluding basal cell carcinomas (morphological code 809) | C52 |
Uterus | C54-C55 |
Ovary and fallopian tube | C56, C570-574 |
Placenta | C58 |
Other and unspecified female genital organs | C577-579 |
Penis excluding basal cell carcinomas (morphological code 809) | C60 |
Prostate | C61 |
Testis | C62 |
Other and unspecified cancers in male genital organs excluding basal cell carcinomas (morphological code 809) | C63 |
Other and unspecified urinary organs | C68 |
Eye and adnexa | C69 |
Meninges | C70 |
Brain including hypophysis, corpus pineale, and ductus craniopharyngealis | C71 |
Spinal cord, cranial nerves and other parts of central nervous system | C72 |
Endocrine glands and related structure | C73-C75 |
Metastasis and unspecified cancer in lymph nodes (only if there is no primary tumor coded) | C77-79 |
Malignant neoplasm of other, ill-defined, or unspecified sites | C76, C80 |
Malignant neoplasms of independent (primary) multiple sites | C97 |
All codes ICD-10 unless otherwise noted.
Footnotes
Additional ICD diagnostic criteria for PTSD include the traumatic event being exceptionally threatening or catastrophic in nature, which is likely to cause pervasive distress in almost anyone; persistent re-experience of the event or distress when exposed to reminders of the event; avoidance of things associated with the event; and symptoms of hyperarousal, all within six months of the event.
Contributor Information
Jaimie L. Gradus, National Center for PTSD, VA Boston Healthcare System, Departments of Psychiatry and Epidemiology, Boston University, Department of Clinical Epidemiology, Aarhus University Hospital.
Dóra Körmendiné Farkas, Department of Clinical Epidemiology, Aarhus University Hospital.
Elisabeth Svensson, Department of Clinical Epidemiology, Aarhus University Hospital.
Vera Ehrenstein, Department of Clinical Epidemiology, Aarhus University Hospital.
Timothy L. Lash, Department of Epidemiology, Emory University, Department of Clinical Epidemiology, Aarhus University Hospital.
Arnold Milstein, Clinical Excellence Research Center, Stanford University.
Nancy Adler, Department of Psychiatry, University of California San Francisco.
Henrik Toft Sørensen, Department of Clinical Epidemiology, Aarhus University Hospital.
References
- 1.Kessler RC. Posttraumatic stress disorder: the burden to the individual and to society. Journal of Clinical Psychiatry. 2000;61(Suppl 5):4–12. [PubMed] [Google Scholar]
- 2.LeShan L. Psychological states as factors in the development of malignant disease: a critical review. Journal of the National Cancer Institute. 1959;22:1–18. [PubMed] [Google Scholar]
- 3.Reiche EMV, Nunes SOV, Morimoto HK. Stress, depression, the immune system and cancer. Lancet Oncology. 2004;5:617–625. doi: 10.1016/S1470-2045(04)01597-9. [DOI] [PubMed] [Google Scholar]
- 4.Chida Y, Hamer M, Wardle J, Steptoe A. Do stress-related psychosocial factors contribute to cancer incidence and survival? Nature Clinical Practice Oncology. 2008;5(8):466–475. doi: 10.1038/ncponc1134. [DOI] [PubMed] [Google Scholar]
- 5.Forsen A. Psychosocial stress as a risk factor for breast cancer. Psychotherapy and Psychosomatics. 1991;55(2–4):176–185. doi: 10.1159/000288427. [DOI] [PubMed] [Google Scholar]
- 6.Cheang A, Cooper CL. Psychosocial factors in breast cancer. Stress Medicine. 1985;1:61–66. [Google Scholar]
- 7.Cooper CL, Cooper R, Faragher B. Incidence and perception of psychological stress: the relationship with breast cancer. Psychological Medicine. 1989;19(2):415–422. doi: 10.1017/s0033291700012459. [DOI] [PubMed] [Google Scholar]
- 8.Lehrer S. Life change and gastric cancer. Psychosomatic Medicine. 1980;2:499–502. doi: 10.1097/00006842-198009000-00004. [DOI] [PubMed] [Google Scholar]
- 9.Lillberg K, Verkasalo PK, Kaprio J, Teppo L, Helenius H, Koskenvuo M. Stressful life events and risk of breast cancer in 10,808 women: a cohort study. American Journal of Epidemiology. 2003;157:415–423. doi: 10.1093/aje/kwg002. [DOI] [PubMed] [Google Scholar]
- 10.Li J, Johansen C, Hansen D, Olsen J. Cancer incidence in parents who lost a child: a nationwide study in Denmark. Cancer. 2002;95(10):2237–2242. doi: 10.1002/cncr.10943. [DOI] [PubMed] [Google Scholar]
- 11.Roberts FD, Newcomb PA, Trentham-Dietz A, Storer BE. Self-reported stress and risk of breast cancer. Cancer. 1995;77:1089–1093. doi: 10.1002/(sici)1097-0142(19960315)77:6<1089::aid-cncr13>3.0.co;2-y. [DOI] [PubMed] [Google Scholar]
- 12.Schernhammer ES, Hankinson SE, Rosner B, Kroenke CH, Willett WC, Colditz GA, et al. Job stress and breast cancer risk. American Journal of Epidemiology. 2004;160(11):1079–1086. doi: 10.1093/aje/kwh327. [DOI] [PubMed] [Google Scholar]
- 13.Johansen C, Olsen JH. Psychological stress, cancer incidence and mortality from non-malignent disease. British Journal of Cancer. 1997;75(1):144–148. doi: 10.1038/bjc.1997.24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kisely S, Crowe E, Lawrence D. Cancer-related mortality in people with mental illness. JAMA Psychiatry. 2013;70(2):209–217. doi: 10.1001/jamapsychiatry.2013.278. [DOI] [PubMed] [Google Scholar]
- 15.Antonova L, Aronson K, Mueller CR. Stress and breast cancer: from epidemiology to molecular biology. Breast Cancer Research. 2011;13(2):208. doi: 10.1186/bcr2836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Soerjomataram I, Lortet-Tieulent J, Parkin DM, Ferlay J, Mathers C, Forman D, et al. Global burden of cancer in 2008: a systematic analysis of disability-adjuted life-years in 12 world regions. The Lancet. 2012;380:1840–1845. doi: 10.1016/S0140-6736(12)60919-2. [DOI] [PubMed] [Google Scholar]
- 17.Pedersen CB, Gotzsche H, Moller J. The Danish civil registration system. A cohort of eight million persons. Danish Medical Bulletin. 2006;53(4):441–449. [PubMed] [Google Scholar]
- 18.Munk-Jorgensen P, Mortensen PB. The Danish Psychiatric Central Register. Danish Medical Bulletin. 1997;44(1):82–84. [PubMed] [Google Scholar]
- 19.Mors O, Perto GP, Mortensen PB. The Danish Psychiatric Central Research Register. Scandinavian Journal of Public Health. 2011;39(7):54–57. doi: 10.1177/1403494810395825. [DOI] [PubMed] [Google Scholar]
- 20.Gradus JL, Bozi I, Antonsen S, Svensson E, Lash TL, Resick PA, et al. Severe stress and adjustment disorder diagnoses in the population of Denmark. Journal of Traumatic Stress. 2014;27(3):370–374. doi: 10.1002/jts.21926. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Svensson E, Antonsen S, Lash TL, Resick PA, Hansen JG, Gradus JL. Validity of reaction to severe stress and adjustment disorders diagnoses in the Danish Psychiatric Central Research Registry. doi: 10.2147/CLEP.S80514. under review. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lynge E, Sandegaard JL, Rebolj M. The Danish National Patient Register. Scandinavian Journal of Public Health. 2011;39:30–33. doi: 10.1177/1403494811401482. [DOI] [PubMed] [Google Scholar]
- 23.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. Journal of Chronic Diseases. 1987;40:373–383. doi: 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
- 24.Strom HH, Michelsen EV, Clemmensen IH, Pihl J. The Danish Cancer Registry - history, content, quality and use. Danish Medical Bulletin. 1997;44(5):535–539. [PubMed] [Google Scholar]
- 25.Rothman KJ, Boice JD. Epidemiologic analyses with a programmable calculator. Washington DC: Government Printing Office; 1979. Contract No.: DHHS publication no. 79-1649. [Google Scholar]