Associations between DSM-IV mental disorders and subsequent self-reported diagnosis of cancer
- ️Sat Jun 02 0046
. Author manuscript; available in PMC: 2016 Nov 30.
Abstract
Objective
The associations between mental disorders and cancer remain unclear. It is also unknown whether any associations vary according to life stage or gender. This paper examines these research questions using data from the World Mental Health Survey Initiative.
Methods
The World Health Organization Composite International Diagnostic Interview retrospectively assessed the lifetime prevalence of 16 DSM-IV mental disorders in face-to-face household population surveys in nineteen countries (n = 52,095). Cancer was indicated by self-report of diagnosis. Smoking was assessed in questions about current and past tobacco use. Survival analyses estimated associations between first onset of mental disorders and subsequently reported cancer.
Results
After adjustment for comorbidity, panic disorder, specific phobia and alcohol abuse were associated with a subsequently self-reported diagnosis of cancer. There was an association between number of mental disorders and the likelihood of reporting a cancer diagnosis following the onset of the mental disorder. This suggests that the associations between mental disorders and cancer risk may be generalised, rather than specific to a particular disorder. Depression is more strongly associated with self-reported cancers diagnosed early in life and in women. PTSD is also associated with cancers diagnosed early in life.
Conclusion
This study reports the magnitude of the associations between mental disorders and a self-reported diagnosis of cancer and provides information about the relevance of comorbidity, gender and the impact at different stages of life. The findings point to a link between the two conditions and lend support to arguments for early identification and treatment of mental disorders.
Keywords: Cancer, Psychiatry, Mental disorder, Epidemiology
Introduction
The high prevalence of treatable mental disorders and their relatively early age of onset [1] means that any possible associations between mental disorders and cancer may merit investigation. There are several mechanisms through which mental disorder may facilitate the occurrence of cancer. The physical sequelae of stress and the symptoms of mental disorder are associated with physiological changes which can promote cancer [2]. Mental disorders are associated with poor health behaviours which increase cancer risk [3]. In addition, mental disorders may impact upon screening uptake and subsequent intervention [4].
Population studies of stress, mental disorders and subsequent cancer have yielded conflicting results. Some studies have demonstrated that stress and anxiety disorders are related [5–7]. Other studies suggest that people with diagnosed mental disorders are no more likely than the general population to have a diagnosis of cancer [5,8,9]. However, these studies focus on clinical populations and exclude people who have not received a formal psychiatric diagnosis. Additionally, in these studies mental disorders are treated as a single category, despite varying characteristics and levels of severity.
The aims of this study were to examine the association of first onset of a range of mental disorders with subsequent onset of cancer, with and without adjustment for mental disorder comorbidity using the World Mental Health (WMH) surveys dataset. Second, to assess whether an increasing number of mental disorders are associated with an increased likelihood of reporting cancer. Third, to assess whether associations vary by gender, or across the life course. These variables were examined in relation to the time of the onset of the mental disorder and the reported time of the cancer diagnosis.
Method
Samples and procedures
This study uses data from 19 World Mental Health (WMH) surveys (see Table 1). A stratified multi-stage clustered area probability sampling strategy was used to select adult respondents (18 years+) in most countries. Most of the surveys were based on nationally representative household samples whilst Colombia, Mexico and Shenzhen were based on nationally representative household samples in urbanised areas. Interviews were undertaken face-to-face by trained lay interviewers.
Table 1.
Characteristics of WMH samples and percent (and number) with a self-reported history of cancer
| Country | Field dates | Age range | Sample size
|
Response rate (%) | History of self-reported cancer diagnosis
|
||
|---|---|---|---|---|---|---|---|
| Part 1 | Part 2 sub-sample | Number unweighted (N) | Weighted (%) | ||||
| Americas | |||||||
| Colombia | 2003 | 18–65 | 4426 | 2381 | 87.7 | 27 | 1.0 |
| Mexico | 2001–2 | 18–65 | 5782 | 2362 | 76.6 | 7 | 0.2 |
| United States | 2002–3 | 18+ | 9282 | 5692 | 70.9 | 361 | 6.9 |
| Peru | 2005–6 | 18–65 | 3930 | 1801 | 90.2 | 9 | 0.7 |
| Asia and South Pacific | |||||||
| Japan | 2002–6 | 20+ | 4129 | 1682 | 55.1 | 57 | 2.9 |
| PRC Shenzhen | 2006–7 | 18+ | 7132 | 2475 | 80.0 | 6 | 0.2 |
| New Zealand | 2003–4 | 18+ | 12,790 | 7312 | 73.3 | 367 | 6.0 |
| Europe | |||||||
| Belgium | 2001–2 | 18+ | 2419 | 1043 | 50.6 | 38 | 3.4 |
| France | 2001–2 | 18+ | 2894 | 1436 | 45.9 | 50 | 4.2 |
| Germany | 2002–3 | 18+ | 3555 | 1323 | 57.8 | 54 | 3.6 |
| Italy | 2001–2 | 18+ | 4712 | 1779 | 71.3 | 45 | 2.4 |
| The Netherlands | 2002–3 | 18+ | 2372 | 1094 | 56.4 | 47 | 3.6 |
| Spain | 2001–2 | 18+ | 5473 | 2121 | 78.6 | 40 | 1.0 |
| Northern Ireland | 2004–7 | 18+ | 4340 | 1986 | 68.4 | 43 | 2.5 |
| Portugal | 2008–9 | 18+ | 3849 | 2060 | 57.3 | 72 | 2.7 |
| Romania | 2005–6 | 18+ | 2357 | 2357 | 70.9 | 26 | 0.9 |
| Poland | 2010–11 | 18–64 | 10,081 | 4000 | 50.4 | 68 | 1.4 |
| Middle East | |||||||
| Israel | 2002–4 | 21+ | 4859 | 4859 | 72.6 | 172 | 3.3 |
| Iraq | 2006–7 | 18+ | 4332 | 4332 | 95.2 | 10 | 0.3 |
| Weighted average response rate (%) | 78.0 | ||||||
| Total sample size | 98,714 | 52,095 | 1499 |
Internal subsampling was used to reduce burden and interview time by dividing the interview into two parts. All respondents completed Part 1 which included the core diagnostic assessment of common mental disorders. All Part 1 respondents who met the lifetime criteria for any mental disorder and a probability sample of other respondents were administered Part 2 which assessed physical conditions and other risk factors. Part 2 respondents were weighted by the inverse of their probability of selection for Part 2 adjusted for differential sampling. These analyses are based on the weighted Part 2 subsample (n = 52,095). Additional weights adjusted for differential probabilities of selection within households, non-response, and to match population sociodemographic distributions. Quality control procedures are described in detail elsewhere [10,11]. All respondents provided informed consent and protocols were approved by the Institutional Review Boards in each country [11].
Measures
The WMH survey version of the WHO Composite International Diagnostic Interview (now CIDI 3.0) assessed lifetime history of mental disorders using the criteria of the DSM-IV. The mental disorders adjusted for in this paper include anxiety disorders (panic disorder, agoraphobia without panic, specific phobia, social phobia, post-traumatic stress disorder, generalised anxiety disorder, obsessive compulsive disorder); mood disorders (major depressive disorder/dysthymia, bipolar disorders I, II and broad); substance use disorders (alcohol abuse and dependence, drug abuse and dependence); and impulse control disorders (intermittent explosive disorder, bulimia nervosa and binge eating disorder). CIDI organic exclusion rules were applied. Clinical reappraisal studies indicate that lifetime diagnoses based on the CIDI have good concordance with diagnoses based on blinded clinical interviews [12].
Cancer status: In questions adapted from the U.S. Health Interview Survey, respondents were asked about the lifetime presence of selected chronic conditions. Respondents were asked: “Did a doctor or other health professional ever tell you that you had any of the following illnesses….Cancer?” If respondents endorsed this item they were classified as having a history of cancer. Respondents were also asked how old they were when they were first diagnosed with cancer. This is referred to as the age of onset of cancer, although it is recognised that malignancy may develop over a period of time.
Statistical analysis
Discrete-time survival analyses [13] were used to test sequential associations between first onset of mental disorders and self-reported subsequent cancer. For these analyses a data set was created in which each year in the life of each respondent (up to and including the age of onset of cancer or their age at interview, whichever came first) was treated as a separate observational record. Year of reported cancer onset was coded 1 and earlier years coded 0. People who reported cancer onset before age 21 were excluded from analysis. Mental disorders were coded 1 from the year after first onset of each individual mental disorder. This time lag of 1 year in the coding of the mental disorders ensured that in cases where the first onset of a mental disorder and cancer occurred in the same year, the mental disorder would not count as an independent variable predicting the reported cancer. Only person-years up to cancer diagnosis were analysed so that only mental disorder episodes occurring before the onset of cancer were included. Logistic regression was used, with the survival coefficients presented as odds ratios. The odds ratios indicated the relative odds of cancer onset in a given year for a person with a prior history of mental disorder compared to people without mental disorders.
A series of bivariate and multivariate models was developed including the independent variable, mental disorder, plus control variables. Models control for person-years, countries, gender, current age, and in the multivariate models, other mental disorders. Bivariate models investigated associations of specific mental disorders with self-reported subsequent cancer. The next model, a multivariate model, estimated the associations of each mental disorder with reported cancer onset adjusting for mental disorder comorbidity (other mental disorders occurring at any stage prior to cancer onset). A second multivariate model included a series of independent variables for number of mental disorders (e.g., one such variable for respondents who experienced exactly one mental disorder and another for respondents who experienced two), as well as the control variables. Other more complex non-additive multivariate models were also run, for example including both type and number of mental disorders, but these did not provide a better fit for the data, so the simpler models are reported here (model fitting statistics available on request).
Our general approach was to not control for covariates that may be associated with both mental disorders and cancer. However, these variables may also confound associations so we re-estimated the multivariate model with adjustment for history of smoking (ever/never) and educational attainment. This made virtually no difference to associations (all previously significant associations remained significant and none reduced in magnitude, data available on request) so we report the results from the model unadjusted for smoking and education here.
We examined life course variation in two ways. First, we examined whether early versus late onset mental disorders differed significantly in their associations with self-reported cancer through creation of mental disorder-specific dummy variables for early onset mental disorder (< =21 years) and late onset disorder (>21 years) (see table footnotes for model specification). Second, we assessed whether associations varied by the time in the life course when cancer was diagnosed by including cross-product terms between person-years (coded as a continuous variable) and each type of mental disorder in the multivariate type model. Gender differences were examined by including cross-product terms between gender and each mental disorder in the multivariate model.
Our earlier studies of concurrent mental–physical comorbidity in the WMH surveys found that these associations are generally consistent cross-nationally, despite varying prevalence of mental disorder and physical conditions. All analyses for this paper were therefore run on the pooled cross-national dataset. As the WMH data are clustered and weighted, the design-based Taylor series linearisation [13] implemented in version 10 of SUDAAN [14] was used to estimate standard errors and evaluate statistical significance.
Results
Descriptive characteristics
Characteristics of the contributing surveys and prevalence of cancer are shown in Table 1. A total of 1499 respondents reported a diagnosis of cancer.
Type and number of mental disorders as predictors of cancer diagnosis
The first column of Table 2 shows the results from bivariate models in which each mental disorder was modelled as a separate predictor of subsequent cancer, without taking mental disorder comorbidity into account. In these models all mood disorders, panic disorder, specific phobia, PTSD, OCD, IED, binge eating disorder, alcohol abuse, alcohol dependence and drug dependence were associated with self-reported cancer. Odds ratios ranged from 1.3 to 2.2. The next column shows the results from a multivariate model adjusting for mental disorder comorbidity. This reduced the magnitude of associations but panic disorder, specific phobia and alcohol abuse remained significantly associated with self-reported cancer with ORs ranging from 1.3 to 1.5. This demonstrates the importance of comorbidity in increasing cancer risk. The global chi square value for the joint effect of all mental disorders was significant (X2 = 61.7, P < =.05). The test for variation in the multivariate ORs approaches significance (X2 = 21.6) this means that we cannot exclude the possibility that the associations between mental disorders and a self-reported diagnosis of cancer are generalised rather than specific. The ORs for PTSD and OCD are also close to significance.
Table 2.
Bivariate and multivariate associations (odds ratios) between DSM-IV mental disorders and the subsequent self-reported diagnosis of cancer
| Bivariate modelsa
|
Multivariate type modelb
|
Multivariate number modelc
|
||||
|---|---|---|---|---|---|---|
| OR | (95% CI) | OR | (95% CI) | OR | (95% CI) | |
| I. Mood disorders | ||||||
| Major depressive episode/dysthymia | 1.3* | (1.1–1.6) | 1.2 | (1.0–1.4) | – | – |
| Bipolar disorder (broad) | 1.6* | (1.1–2.4) | 1.1 | (0.7–1.8) | – | – |
| II. Anxiety disorders | ||||||
| Panic disorder | 1.8* | (1.4–2.5) | 1.5* | (1.1–2.0) | – | – |
| Generalised anxiety disorder | 1.2 | (0.9–1.5) | 0.9 | (0.7–1.2) | – | – |
| Social phobia | 1.2 | (1.0–1.5) | 0.9 | (0.7–1.2) | – | – |
| Specific phobia | 1.4* | (1.2–1.7) | 1.3* | (1.1–1.5) | – | – |
| Agoraphobia without panic | 1.4 | (0.9–2.3) | 1.1 | (0.7–1.8) | – | – |
| Post-traumatic stress disorder | 1.5* | (1.2–1.9) | 1.3 | (1.0–1.6) | – | – |
| Obsessive compulsive disorder | 2.2* | (1.2–3.8) | 1.7 | (1.0–3.0) | – | – |
| III. Impulse-control disorders | ||||||
| Intermittent explosive disorder | 1.6* | (1.1–2.3) | 1.3 | (0.9–1.9) | – | – |
| Binge eating disorder | 1.9* | (1.0–3.4) | 1.5 | (0.8–2.7) | – | – |
| Bulimia nervosa | 1.8 | (1.0–3.4) | 1.2 | (0.6–2.4) | – | – |
| IV. Substance disorders | ||||||
| Alcohol abuse | 1.6* | (1.3–2.1) | 1.5* | (1.1–2.1) | – | – |
| Alcohol dependence with abuse | 1.5* | (1.1–2.1) | 0.8 | (0.5–1.3) | – | – |
| Drug abuse | 1.5 | (1.0–2.3) | 0.9 | (0.5–1.6) | – | – |
| Drug dependence with abuse | 2.0* | (1.1–3.6) | 1.3 | (0.7–2.7) | – | – |
| Joint effect of all types of disorders, c162 | 61.7* | |||||
| Difference between types of disorders, c152 | 21.6 | |||||
| V. Number of disorder | ||||||
| Exactly 1 disorder | – | – | – | – | 1.3* | (1.1–1.6) |
| Exactly 2 disorders | – | – | – | – | 1.4* | (1.1–1.8) |
| Exactly 3 disorders | – | – | – | – | 1.6* | (1.2–2.2) |
| Exactly 4 disorders | – | – | – | – | 1.3 | (0.9–1.8) |
| 5+ disorders | – | – | – | – | 2.3* | (1.6–3.3) |
| Joint effect of number of disorders, c52 | 36.6* |
The final columns of Table 2 show the association between number of mental disorders and self-reported cancer diagnosis, regardless of type of mental disorder. The odds of self-reported cancer increase from 1.3 to 1.6 from one to three disorders. The OR of 1.3 for four disorders is not significant. Finally, having five or more lifetime disorders is associated with the highest OR for self-reported cancer (2.3). The global chi square value for the joint effect of all mental disorders was significant (X2 = 36.6, P < =.05). This again demonstrates that self reported cancer may be associated with having more than one mental disorder. This model provided a better explanation for the data than either the multivariate model or a more complex model including information about number and type (model fitting results available on request).
Timing of mental disorder onset (early versus late onset)
Table 3 presents the models examining whether mental disorders with first onset prior to age 21 were more strongly associated with a self-reported cancer diagnosis than later onset disorders. For depression, phobia, PTSD, IED, binge eating disorder, alcohol abuse, alcohol dependence, drug abuse and drug dependence with abuse early onset disorders had quantitatively larger associations with a self-reported cancer diagnosis compared to the late onset mental disorders. For panic disorder, OCD, bulimia nervosa and alcohol abuse late onset had a larger association with self-reported cancer. The magnitude of difference was significant only for depression; the OR of subsequently reporting cancer diagnosis following early onset depression was 1.7 compared with 1.2 for late onset depression. This finding may reveal an association between depression and self-reported cancer that was not apparent in Table 2, that is: depression is associated with cancer earlier in life. In the multivariate models that took other mental disorders into account, there was no association between the time of onset of disorder and self-reported cancer. The apparent effect of timing of mental disorder onset in the bivariate models is therefore due to early onset disorders being markers of comorbidity.
Table 3.
Associations (odds ratios) between early vs. late mental disorder onset and the subsequent self-reported diagnosis of cancer
| Bivariate modelsa
|
Multivariate modelb
|
|||||||
|---|---|---|---|---|---|---|---|---|
| Early | Late | Test of the difference between early and late | Early | Late | Test of the difference between early and late | |||
|
|
|
|
|
|
|
|||
| OR (95% CI) | OR (95% CI) |
c12 |
[P] | OR (95% CI) | OR (95% CI) |
c12 |
[P] | |
| I. Mood disorders | ||||||||
| Major depressive episode/dysthymia | 1.7* (1.3–2.2) | 1.2 (1.0–1.5) | 4.2* | [.041] | 1.4* (1.1–1.9) | 1.1 (0.9–1.4) | 2.2 | [.139] |
| Bipolar disorder (broad) | 1.7 (0.9–3.0) | 1.6 (0.9–2.7) | 0.0 | [.890] | 1.1 (0.6–2.1) | 1.2 (0.7–2.1) | 0.0 | [.903] |
| II. Anxiety disorders | ||||||||
| Panic disorder | 1.5 (0.9–2.6) | 2.0* (1.4–3.0) | 0.8 | [.379] | 1.1 (0.7–1.9) | 1.7* (1.2–2.5) | 1.6 | [.210] |
| Generalised anxiety disorder | 1.4 (1.0–2.0) | 1.1 (0.8–1.5) | 0.8 | [.381] | 0.9 (0.6–1.4) | 0.9 (0.6–1.3) | 0.0 | [.852] |
| Social phobia | 1.1 (0.9–1.4) | 1.6 (0.8–3.3) | 0.9 | [.344] | 0.9 (0.7–1.1) | 1.3 (0.6–2.7) | 1.2 | [.280] |
| Specific phobia | 1.5* (1.2–1.7) | 1.2 (0.6–2.6) | 0.2 | [.633] | 1.3* (1.1–1.5) | 1.1 (0.5–2.3) | 0.2 | [.634] |
| Agoraphobia without panic disorder | 1.6 (0.9–2.8) | 1.1 (0.5–2.6) | 0.4 | [.523] | 1.2 (0.7–2.3) | 0.9 (0.4–2.2) | 0.3 | [.590] |
| Post-traumatic stress disorder | 1.8* (1.3–2.6) | 1.3 (0.9–1.9) | 2.3 | [.133] | 1.4* (1.0–2.0) | 1.1 (0.8–1.6) | 1.3 | [.253] |
| Obsessive compulsive disorder | 1.8 (0.9–3.7) | 3.0* (1.4–6.2) | 1.1 | [.291] | 1.4 (0.7–3.0) | 2.1 (1.0–4.4) | 0.6 | [.423] |
| III. Impulse-control disorders | ||||||||
| Intermittent explosive disorder | 1.7* (1.1–2.5) | 1.3 (0.5–2.9) | 0.4 | [.550] | 1.3 (0.9–2.0) | 1.0 (0.5–2.4) | 0.2 | [.622] |
| Binge eating disorder | 2.3* (1.0–5.3) | 1.5 (0.6–3.7) | 0.5 | [.471] | 2.0 (0.8–4.7) | 1.2 (0.5–2.8) | 0.6 | [.422] |
| Bulimia nervosa | 1.3 (0.5–3.3) | 2.6* (1.1–6.3) | 1.0 | [.310] | 0.9 (0.4–2.4) | 2.0 (0.8–5.0) | 1.3 | [.254] |
| IV. Substance disorders | ||||||||
| Alcohol abuse | 1.5* (1.1–2.0) | 1.7* (1.3–2.4) | 0.7 | [.411] | 1.3 (0.9–1.9) | 1.7* (1.2–2.4) | 1.1 | [.301] |
| Alcohol dependence with abuse | 1.5 (0.9–2.4) | 1.5 (0.9–2.5) | 0.0 | [.947] | 0.8 (0.4–1.6) | 0.9 (0.5–1.5) | 0.0 | [.916] |
| Drug abuse | 1.7* (1.1–2.7) | 1.3 (0.6–2.5) | 0.6 | [.452] | 0.9 (0.5–1.6) | 0.9 (0.4–2.1) | 0.0 | [.888] |
| Drug dependence with abuse | 3.0* (1.6–5.7) | 0.8 (0.3–2.9) | 3.5 | [.060] | 2.1* (1.0–4.4) | 0.6 (0.2–2.2) | 3.3 | [.069] |
| V. Joint effect of all early onset disorders, c162 | 50.5* | |||||||
| VI. Joint effect of all late onset disorders, c162 | 39.3* | |||||||
| VII. Joint effect of early onset disorders independent of joint effect of any disorders, c162 | 20.3 |
Variation over the life-course (timing of cancer diagnosis)
We next investigated whether associations between mental disorders and self-reported subsequent cancer diagnosis varied depending on whether the cancer was diagnosed at an early age or later in life. Table 4 shows the predictor mental disorders which demonstrated statistically significant interactions with person-year in multivariate models. These were: depression/dysthymia, specific phobia, PTSD and drug dependence with abuse. We then stratified the person year dataset by quartiles of the timing of the reported cancer diagnosis distribution and examined the multivariate associations between each mental disorder and self-reported cancer in each of these person year subsets (Table 4). The significant interactions indicate that the effect of mental disorders varied by time of life and the stratified analyses illustrate the pattern that variation takes. For depression/dysthymia, specific phobia, PTSD and drug dependence, associations were stronger with cancer diagnosed earlier in life. This explains why associations between PTSD and self-reported cancer were not shown in the multivariate model in Table 2 as this analysis included cancer diagnosed throughout the lifetime.
Table 4.
Variations in associations between mental disorders and self-reported cancer by life course timing of cancer onset (diagnosis)
| Type of mental disorders | Mental disorder* person-year interactiona | Stratified modelsb
|
|||||
|---|---|---|---|---|---|---|---|
| Up to age 44 | Age 45–56 | Age 57–66 | Age 67+ | ||||
|
|
|
|
|
|
|||
| OR (95% CI) |
c12 |
[P] | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| Major depressive episode/dysthymia | 0.98* (0.97–0.99) | 22.3* | [.000] | 1.6* (1.2–2.1) | 1.1 (0.8–1.5) | 0.9 (0.6–1.3) | 1.0 (0.6–1.7) |
| Specific phobia | 0.98* (0.98–0.99) | 12.9* | [.000] | 1.6* (1.2–2.1) | 1.1 (0.8–1.6) | 1.1 (0.6–2.0) | 0.7 (0.3–1.3) |
| Post-traumatic stress disorder | 0.97* (0.96–0.99) | 16.4* | [.000] | 1.5* (1.0–2.1) | 1.3 (0.8–2.0) | 1.2 (0.6–2.6) | 0.6 (0.2–1.3) |
| Drug dependence with abuse | 0.95* (0.92–0.99) | 7.4* | [.006] | 1.5 (0.7–3.2) | 0.2 (0.0–1.8) | 0.0* (0.0–0.0) | 0.0* (0.0–0.1) |
Gender differences
The association between depression and self-reported cancer diagnosis was significantly stronger for women (interaction OR of 1.5; 95% CI: 1.0–2.4) (Table 5).
Table 5.
Interactions between type of mental disorder and gender in predicting the subsequent self-reported diagnosis of cancer
| Cancera
|
||||
|---|---|---|---|---|
| OR | (95% CI) | Equ | [P] | |
| I. Mood disorders | ||||
| Major depressive episode/dysthymia | 1.5* | (1.0–2.4) | 3.9* | [.049] |
| Bipolar disorder (broad) | 1.1 | (0.4–2.9) | 0.0 | [.875] |
| II. Anxiety disorders | ||||
| Panic disorder | 1.3 | (0.6–2.7) | 0.6 | [.455] |
| Generalised anxiety disorder | 0.8 | (0.4–1.3) | 1.1 | [.303] |
| Social phobia | 1.3 | (0.8–2.2) | 1.4 | [.236] |
| Specific phobia | 1.0 | (0.6–1.5) | 0.0 | [.880] |
| Agoraphobia without panic | 0.6 | (0.2–1.7) | 1.0 | [.323] |
| Post-traumatic stress disorder | 0.8 | (0.4–1.6) | 0.4 | [.535] |
| Obsessive compulsive disorder | 1.5 | (0.3–6.4) | 0.3 | [.592] |
| III. Impulse-control disorders | ||||
| Intermittent explosive disorder | 0.9 | (0.4–2.0) | 0.0 | [.856] |
| Binge eating disorder | 1.7 | (0.3–9.8) | 0.4 | [.538] |
| Bulimia nervosa | 1.4 | (0.1–16.0) | 0.1 | [.780] |
| IV. Substance disorders | ||||
| Alcohol abuse | 1.5 | (0.8–2.8) | 1.3 | [.261] |
| Alcohol dependence with abuse | 0.6 | (0.3–1.3) | 1.7 | [.193] |
| Drug abuse | 1.6 | (0.6–4.5) | 0.7 | [.393] |
| Drug dependence with abuse | 0.7 | (0.2–2.5) | 0.4 | [.527] |
Again, it is important to note that the non-significant association between self-reported cancer and mental disorder in Table 2 applies only to males.
Discussion
This study has a number of limitations. The data on mental disorders is based on retrospective recall of symptoms and, for mental disorders, this is associated with underestimates and errors in estimating onset [15]. The data on cancer is also based on recall and self report rather than clinical data. The validity of self reported cancer diagnosis and the accuracy of timing of onset data may also be questioned; however data on the accuracy of self reported cancer demonstrate acceptable levels of validity [16,17]. Depression has not been found to bias the self-reporting of physical disorders [18,19]. It remains possible that the associations reported here are affected by the misclassification of either mental disorder or cancer. Due to the low prevalence of individual cancer types we have combined all forms of cancer (except non-melanoma skin cancers). This means that cancers with behavioural risk factors are combined with those with alternative aetiologies and any differences in the associations with mental disorders between these two categories are therefore not detectable. Finally, the associations between mental disorder and subsequent cancer are based on analysis of a population which only includes those who had survived and were well enough to participate. Many of those most affected are therefore missing due to illness and premature mortality. There is evidence that people with a mental disorder have a higher cancer mortality rate [6,20,21]. The associations reported here are therefore likely to be under-estimations.
This is the first epidemiological examination reporting an association between DSM mental disorders and self-reported subsequent cancer. The results conflict with earlier findings that people with mental disorders are no more likely to develop cancer [5,9]. However these are studies of clinical populations rather than general populations and this is an important distinction. People who have an undiagnosed mental disorder may have a higher risk of reporting cancer and they may have poorer lifestyle behaviours. Panic disorder, specific phobia and alcohol abuse are associated with reporting cancer in this study, however the magnitude of the associations is reduced following adjustment for comorbidity, and the risk of reporting cancer rises if a person has more mental disorders. This leads us to conclude that the associations between mental disorders and cancer risk may be associated with features common to a range of disorders, rather than specific to a particular disorder. This finding supports the hypothesis that the stress associated with having a mental disorder increases cancer risk thus supporting findings that stress is related to cancer [5–7].
Stressful events are associated with mental disorders and having a mental disorder is itself stressful [22,23]. Anxiety disorders are accompanied by hypothalamic–pituitary–adrenal (HPA) activation which can impact upon immunological responses, thereby increasing cancer risk. Alterations in HPA functioning associated with the hyper-arousal in PTSD may explain this disorder's link with early cancers [24,25]. In this study both depression and anxiety were associated with cancer and this is difficult to explain in terms of HPA dysregulation given that these disorders are associated with opposing HPA responses [26] this suggests that anxiety–depression comorbidity is associated with a variant of HPA dysregulation [24]. Additionally, whilst adjustment for smoking and education level had no impact in this study, other behavioural risk factors, such as poor diet and lack of exercise, are associated with both stress and cancer [3]. Heavy consumption of alcohol is associated with certain cancer types [27] and stress [3]; this may explain the association between alcohol abuse and reported cancer. Finally, stress may impact upon screening and interventions; however this is likely to account for a small minority of these cancers.
Depression is more strongly associated with cancer among women than men and depression is also more strongly associated with early self-reported cancer. There is also evidence that tricyclic anti-depressants are associated with a higher risk of breast cancer [28,29]; however the effect may be removed when confounders are considered [30]. Finally, it is important to note that non-causal factors such as environmental exposure, diet, heredity and childhood adversities may also explain the associations between cancers and mental disorders and these were not considered in this study.
This is the first study to examine the associations between mental disorders and a self-reported subsequent diagnosis of cancer worldwide. The study reveals the magnitude of these associations and provides important information about the relevance of comorbidity and associations between mental disorders and cancers at different life stages. The study also reveals an association between depression and self-reported cancer for women. Further research is required to determine whether the associations are causal. Nonetheless, the associations between anxiety disorders, comorbid disorders and increasing risk for people with higher numbers of mental disorders point to the existence of a link between the two types of conditions and add further weight to arguments for the need to identify and treat mental disorders as early as possible.
Acknowledgments
The World Health Organization World Mental Health (WMH) Survey Initiative is supported by the National Institute of Mental Health (NIMH; R01 MH070884), the John D. and Catherine T. MacArthur Foundation, the Pfizer Foundation, the US Public Health Service (R13-MH066849, R01-MH069864, and R01 DA016558), the Fogarty International Center (FIRCA R03-TW006481), the Pan American Health Organization, Eli Lilly and Company, Ortho-McNeil Pharmaceutical, GlaxoSmithKline, and Bristol-Myers Squibb. We thank the staff of the WMH Data Collection and Data Analysis Coordination Centres for assistance with instrumentation, fieldwork, and consultation on data analysis. The Colombian National Study of Mental Health (NSMH) was supported by the Ministry of Social Protection, with supplemental support from the Saldarriaga Concha Foundation. The European surveys were funded by the European Commission (contracts QLG5-1999-01042; SANCO 2004123; EAHC 20081308), the Piedmont Region (Italy), the Fondo de Investigación Sanitaria, Instituto de Salud Carlos III, Spain (FIS 00/0028), the Ministerio de Ciencia y Tecnología, Spain (SAF 2000-158-CE), the Departament de Salut, Generalitat de Catalunya, Spain, Instituto de Salud Carlos III (CIBER CB06/02/0046, RETICS RD06/0011 REM-TAP), and other local agencies and by an unrestricted educational grant from GlaxoSmithKline. The World Mental Health Japan (WMHJ) survey was supported by the Grant for Research on Psychiatric and Neurological Diseases and Mental Health (H13-SHOGAI-023, H14-TOKUBETSU-026, H16-KOKORO-013) from the Japan Ministry of Health, Labour and Welfare. The Mexican National Comorbidity Survey (MNCS) was supported by the National Institute of Psychiatry Ramon de la Fuente (INPRFMDIES 4280) and by the National Council on Science and Technology (CONACyT-G30544-H), with supplemental support from the PanAmerican Health Organization (PAHO). The Peruvian World Mental Health Study was funded by the National Institute of Health of the Ministry of Health of Peru. The Polish project Epidemiology of Mental Health and Access to Care — EZOP Poland was carried out by the Institute of Psychiatry and Neurology in Warsaw in consortium with the Department of Psychiatry — Medical University in Wroclaw and National Institute of Public Health-National Institute of Hygiene in Warsaw and in partnership with the Psykiatrist Institut Vinderen - Universitet, Oslo. The project was funded by the Norwegian Financial Mechanism and the European Economic Area Mechanism as well as the Polish Ministry of Health. No support from pharmaceutical industry neither other commercial sources was received. The Shenzhen Mental Health Survey is supported by the Shenzhen Bureau of Health and the Shenzhen Bureau of Science, Technology, and Information. Implementation of the Iraq Mental Health Survey (IMHS) and data entry were carried out by the staff of the Iraqi MOH and MOP with direct support from the Iraqi IMHS team with funding from both Japanese and European Funds through the United Nations Development Group Iraq Trust Fund (UNDG ITF). The Israel National Health Survey is funded by the Ministry of Health with support from the Israel National Institute for Health Policy and Health Services Research and the National Insurance Institute of Israel. Te Rau Hinengaro: The New Zealand Mental Health Survey (NZMHS) was supported by the New Zealand Ministry of Health, Alcohol Advisory Council and the Health Research Council. The Portuguese Mental Health Study was carried out by the Department of Mental Health, Faculty of Medical Sciences, NOVA University of Lisbon, with collaboration of the Portuguese Catholic University, and was funded by the Champalimaud Foundation, Gulbenkian Foundation, Foundation for Science and Technology (FCT) and the Ministry of Health. The Romania WMH study projects “Policies in Mental Health Area” and “National Study regarding Mental Health and Services Use” were carried out by National School of Public Health & Health Services Management (former National Institute for Research & Development in Health, present National School of Public Health Management & Professional Development, Bucharest), with technical support of the Metro Media Transilvania, the National Institute of Statistics — National Centre for Training in Statistics, the SC. Cheyenne Services SRL, and the Statistics Netherlands and were funded by the Ministry of Public Health (former Ministry of Health) with supplemental support of Eli Lilly Romania SRL. The US National Comorbidity Survey Replication (NCS-R) is supported by the National Institute of Mental Health (NIMH; U01-MH60220) with supplemental support from the National Institute of Drug Abuse (NIDA), the Substance Abuse and Mental Health Services Administration (SAMHSA), the Robert Wood Johnson Foundation (RWJF; grant 044708), and the John W. Alden Trust. A complete list of all within-country and cross-national WMH publications can be found at http://www.hcp.med.harvard.edu/wmh/.
Additional funding: Work on this paper was funded by a grant from the Health Research Council of New Zealand to Kate M Scott.
Footnotes
Conflict of interest
The authors have no competing interests to report.
References
- 1.Kessler RC, Angermeyer M, Anthony JC, De Graff R, Demyttenaere K, Gasquet I, et al. Lifetime prevalence and age-of-onset distributions of mental disorders in the World Health Organization's World Mental Health Survey Initiative. World Psychiatry. 2007;6:168–76. [PMC free article] [PubMed] [Google Scholar]
- 2.Pariante CM, Lightman SL. The HPA axis in major depression: classical theories and new developments. Trends Neurosci. 2008;31:464–8. doi: 10.1016/j.tins.2008.06.006. [DOI] [PubMed] [Google Scholar]
- 3.Scott D, Happell B. The high prevalence of poor physical health and unhealthy lifestyle behaviours in individuals with severe mental illness. Issues Ment Health Nurs. 2011;32:589–97. doi: 10.3109/01612840.2011.569846. [DOI] [PubMed] [Google Scholar]
- 4.Aggarwal A, Pandurangi A, Smith W. Disparities in breast and cervical cancer screening in women with mental illness: a systematic literature review. Am J Prev Med. 2013;44:392–8. doi: 10.1016/j.amepre.2012.12.006. [DOI] [PubMed] [Google Scholar]
- 5.Liang JA, Sun LM, Su KP, Chang Sn, Sung FC, Muo CH, et al. A nationwide population-based cohort study: will anxiety disorders increase subsequent cancer risk? PLoS One. 2012;7:e36370. doi: 10.1371/journal.pone.0036370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hamer M, Chida Y, Molloy GJ. Psychological distress and cancer mortality. J Psychosom Res. 2009;66:255–8. doi: 10.1016/j.jpsychores.2008.11.002. [DOI] [PubMed] [Google Scholar]
- 7.Chida Y, Hamer M, Wardle J, Steptoe A. Do stress-related psychosocial factors contribute to cancer incidence and survival? Nat Clin Pract Oncol. 2008;5:466–75. doi: 10.1038/ncponc1134. [DOI] [PubMed] [Google Scholar]
- 8.Lillberg K, Verkasalo PK, Kaprio J, Teppo L, Helenius H, Koskenvuo M. Stress of daily activities and risk of breast cancer: a prospective cohort study in Finland. Int J Cancer. 2001;91:888–93. doi: 10.1002/1097-0215(200002)9999:9999<::aid-ijc1138>3.0.co;2-d. [DOI] [PubMed] [Google Scholar]
- 9.Lawrence D, Holman CD, Jablensky AV, Threlfall TJ, Fuller SA. Excess cancer mortality in Western Australian psychiatric patients due to higher case fatality rates. Acta Psychiatr Scand. 2000;101:382–8. doi: 10.1034/j.1600-0447.2000.101005382.x. [DOI] [PubMed] [Google Scholar]
- 10.Kessler RC, Ustun B. The World Mental Health (WMH) Survey Initiative version of the World Health Organization (WHO) Composite International Diagnostic Interview (CIDI) Int J Methods Psychiatr Res. 2004;13:93–121. doi: 10.1002/mpr.168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kessler RC, Ustun TB, editors. The WHO World Mental Health Surveys: global perspectives on the epidemiology of mental disorders. New York: Cambridge University Press; 2008. [Google Scholar]
- 12.Haro JM, Arbabzadeh-Bouchez S, Brugha TS, de Girolamo G, Guyer ME, Lepine JP, et al. Concordance of the Composite International Diagnostic Interview Version 3.0 (CIDI 3. 0) with standardized clinical assessments in the WHO World Mental Health Surveys. Int J Methods Psychiatr Res. 2006;15:167–80. doi: 10.1002/mpr.196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Singer JD, Willett JB. It's about time: using discrete-time survival analysis to study duration and the timing of events. J Educ Behav Stat. 1993;18:155–95. [Google Scholar]
- 14.Institute RT. Software for the statistical analysis of correlated data. North Carolina, USA: Research Triangle Park; 1999. SUDAAN. [Google Scholar]
- 15.Wells JE, Horwood HJ. How accurate is recall of key symptoms of depression?: a comparison of recall and longitudinal reports. Psychol Med. 2004;34:1001–11. doi: 10.1017/s0033291703001843. [DOI] [PubMed] [Google Scholar]
- 16.Kriegsman DM, Penninx BW, van Eijk JT, Boeke AJ, Deeg DJ. Self-reports and general practitioner information on the presence of chronic diseases in community dwelling elderly. A study on the accuracy of patients' self-reports and on determinants of inaccuracy. J Clin Epidemiol. 1996;49:1407–17. doi: 10.1016/s0895-4356(96)00274-0. [DOI] [PubMed] [Google Scholar]
- 17.Zajacova A, Dowd JB, Schoeni RF, Wallace RB. Consistency and precision of cancer reporting in a multiwave national panel survey. Popul Health Metrics. 2010;8:20. doi: 10.1186/1478-7954-8-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kolk AM, Hanewald GJ, Schagen S, Gijsbers van Wijk CM. Predicting medically unexplained physical symptoms and health care utilization. A symptom-perception approach. J Psychosom Res. 2002;52:35–44. doi: 10.1016/s0022-3999(01)00307-5. [DOI] [PubMed] [Google Scholar]
- 19.Vassend O, Skrondal A. The role of negative affectivity in self-assessment of health. J Health Psychol. 1999;4:465–82. doi: 10.1177/135910539900400402. [DOI] [PubMed] [Google Scholar]
- 20.Kisely S, Sadek J, MacKenzie A, Lawrence D, Campbell LA. Excess cancer mortality in psychiatric patients. Can J Pychiatry. 2008;53:753–61. doi: 10.1177/070674370805301107. [DOI] [PubMed] [Google Scholar]
- 21.Kisely S, Crowe E, Lawrence D. Cancer-related mortality in people with mental illness. JAMA Psychiatry. 2013;70:209–17. doi: 10.1001/jamapsychiatry.2013.278. [DOI] [PubMed] [Google Scholar]
- 22.McEwen BS. Mood disorders and allostatic load. Biol Psychiatry. 2003;54:200–7. doi: 10.1016/s0006-3223(03)00177-x. [DOI] [PubMed] [Google Scholar]
- 23.Stetler C, Miller GE. Depression and hypothalamic–pituitary–adrenal activation: a quantitative summary of four decades of research. Psychosom Med. 2011;73:114–26. doi: 10.1097/PSY.0b013e31820ad12b. [DOI] [PubMed] [Google Scholar]
- 24.Yehuda R. Current status of cortisol findings in post-traumatic stress disorder. Psychiatry Clin N Am. 2002;25:341–68. doi: 10.1016/s0193-953x(02)00002-3. [DOI] [PubMed] [Google Scholar]
- 25.Wessa M, Rohleder N, Kirschbaum C, Flor H. Altered cortisol awakening response in posttraumatic stress disorder. Psychoneuroendocrinology. 2006;31:209–15. doi: 10.1016/j.psyneuen.2005.06.010. [DOI] [PubMed] [Google Scholar]
- 26.Young EA, Abelson JL, Cameron OG. Effect of comorbid anxiety disorders on the hypothalamic–pituitary–adrenal axis response to a social stressor in major depression. Biol Psychiatry. 2004;56:113–20. doi: 10.1016/j.biopsych.2004.03.017. [DOI] [PubMed] [Google Scholar]
- 27.Giacosaa A, Adam-Blondon AF, Baer-Sinnott S, Barale R, Bavaresco L, Di Gaspero G, et al. Alcohol and wine in relation to cancer and other diseases. Eur J Cancer Prev. 2012;21:103–8. doi: 10.1097/CEJ.0b013e32834761d3. [DOI] [PubMed] [Google Scholar]
- 28.Cotterchio M, Kreiger N, Darlington G, Steingart A. Antidepressant medication use and breast cancer risk. J Clin Epidemiol. 2000;56:155–63. doi: 10.1093/oxfordjournals.aje.a010138. [DOI] [PubMed] [Google Scholar]
- 29.Sharpe DR, Collet JP, Belzile E, Hanley JA, Boivin JF. The effects of tricyclic antidepressants on breast cancer risk. Br J Cancer. 2002;86:92–7. doi: 10.1038/sj.bjc.6600013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Lawlor DA, Ebrahim S, Egger M. Systematic review of the epidemiologic and trial evidence of an association between antidepressant medication and breast cancer. J Clin Epidemiol. 2003;56:155–63. doi: 10.1016/s0895-4356(02)00568-1. [DOI] [PubMed] [Google Scholar]