SEARCH

SEARCH BY CITATION

Keywords:

  • circadian clock;
  • night-time shift work;
  • non-Hodgkin lymphoma

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Our aim was to find out whether non-Hodgkin lymphoma (NHL) was more common than expected among night-time shift workers. The Finnish job-exposure matrix (FINJEM) provided estimates of the proportion of exposed persons and the mean level of exposure among the exposed in each occupation. The probability of night-time work in each occupation was assessed, the observed and expected numbers of cancer cases in a cohort of persons born in 1906–1945 during the years of 1971–1995 were calculated, and the cumulative index of night-time work was scored. The cohort compromised of 1,669,272 persons of whom 6,307 (3,813 men and 2,494 women) had NHL during the follow-up. Night-time work increased significantly (p = 0.01) the risk of NHL in men, the overall relative risk being 1.10 (95% confidence interval of 1.03–1.19). Using the lag period of 10 years, the risk ratio was 1.28 (1.03–1.59) for men who worked in night-time shifts to a high degree as compared with those who had not been exposed to night-time work. Night-time workers are cancer prone and have a greater risk of NHL than population on average. © 2008 Wiley-Liss, Inc.

The incidence rates of many cancers have been decreasing in recent years but few have been increasing.1 One of the latter is non-Hodgkin lymphoma (NHL) which is a heterogeneous group of diseases caused by malignant growth in lymphoid tissue in any part of the human body.2–6 The risk for NHL increases with age similar to most cancers, and NHL is more common in men than women (for the Finnish statistics, please see www.cancerregistry.fi). Concerning NHL in Finland, the incidence per 1 million as corrected for age and adjusted to the World Standard Population in five-year periods from 1972 to 2001 increased in men from 58 to 119 and in women from 35 to 90. In recent years, its incidence has been increasing in men but decreasing in women.

The etiology of NHL is largely unknown. Disruption of the circadian clockwork is one of the factors suggested to predispose to NHL.7 The circadian clock shares common features with cell-division cycle and there seems to be common regulatory elements beyond these systems. Disturbed circadian rhythms can lead to failures in control of the cell-division cycle and thereby may cause not only sleep disorders or depressive disorders but also cancers. To understand the pathogenesis of cancer more in detail it will be important to identify mechanisms of action which contribute to the loss of control of the cell-division cycle.

Cells of the suprachiasmatic nuclei of the anterior hypothalamus in the brain are the principal regulator of biological rhythms including the circadian and seasonal cycles. The circadian clock genes are involved in the cell division cycle, metabolic cycles and reproductive actions in addition to their roles in the generation of the circadian rhythms and sleep-wake cycles.8 The circadian clock genes expressed in the central nervous system are able to guide directly the function of a stem cell niche in peripheral tissues,9 and the regenerative programs of tissues by the extramedullary restoration of specialized hematopoietic cells may be lost due to abnormalities in the circadian clockwork. ARNTL (alias BMAL1 or MOP3) is one of the key proteins encoded by the circadian clock genes and heterodimerizes with either CLOCK or NPAS2, and each pair of these proteins can activate transcription of more than 100 target genes among which are additional circadian clock genes such as the Per1, Per2 and Per3 genes. The Arntl gene has been implicated as critical in the development of B cells, the differentiation of pre-B cells to mature B cells in specific, but not in the development of other immune cells.10

Abnormalities in the circadian clockwork produced abruptly by jet lag or gradually by shift work may have a range of public health implications. These genes and their encoded proteins are relevant for malignant growth being sensitive to the circadian clockwork abnormalities such as NHL or breast cancer. Mice deficient in the mPer2 gene are cancer prone, and after gamma radiation, these mice show a marked increase in tumor development of malignant lymphomas in multiple organs including liver, lung, spleen, heart, ovary, salivary gland, muscle, pancreas, stomach, intestines and bone.11

The system-driven circadian expression of the Per2 gene is of particular interest, since it provides a direct link between the signals from the suprachiasmatic nuclei and the phase of peripheral oscillators12 and is thereby a key to desynchronized oscillations. Genes and their encoded proteins which make up the core molecular clock for generation of the circadian rhythms have been linked to cancer, to NHL in particular but also to breast cancer.7, 13 These recent findings gave a rationale for our study herein.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Study population

The study population consisted of all Finns born in 1906–1945 who participated in the national population census on December 31, 1970 (2,186,029 persons). The census files are maintained by Statistics Finland, and they are updated for death and emigration to allow exact person-year calculation. Data on the occupation and industry held for the longest time in 1970 were obtained from the 1970 population census records.

The social class for each person was determined based on his or her occupation, education and industrial status in 1970.14 In our analyses, farmers were included as a separate social class, and the rest of the population was classified as higher or white-collar, clerical, skilled or blue-collar and unskilled workers. The socioeconomic status was taken as an indicator of lifestyle, having an effect on and potentially confounding the health status, and therefore controlled for in the analysis.

Cancer data

The Finnish Cancer Registry has collected data on all incident cancer cases diagnosed in Finland since 1953. Physicians, hospitals, and pathological, cytological and hematological laboratories in the country are obligated to send notification of all cancer cases that come to their attention to the Registry. In addition, Statistics Finland annually sends the Registry a computerized file on death certificates in which a malignant disease is mentioned. The coverage of the Registry is virtually complete, and data accuracy is high.15

For our cohort study, all the cancers diagnosed in 1971–1995 among persons born in 1906–1945 were extracted from the Finnish Cancer Registry database and sent to Statistics Finland for linkage with the 1970 census files. All persons residing in Finland since 1967 have been assigned a unique 11-digit personal identification code, which was used in the linkage. For 2.2% of the cancer cases, there was no record in the 1970 census file, and they were excluded from the cohort. Economically inactive persons were also excluded. The cohort of economically active persons in ages 25–64 during population census 1970 comprised of 1,669,272 persons; 6,307 of them (3,813 men and 2,494 women) had NHL during the follow-up.

Job-exposure matrix

The occupational exposure estimates used in this study were based on the Finnish job-exposure matrix FINJEM.16 The basic dimensions of exposure assessment in FINJEM include occupations, agents and calendar time or period. FINJEM provides estimates of the proportion (P) of exposed persons, the mean level (L) of exposure among the exposed in each occupation, and the cumulative index (P × L) as a function of time for each occupational exposure.

Herein, the cumulative index of night-time work was calculated. The prevalence of night-time work in all occupations was assessed on the basis of responses to the question “How is your working time arranged?” as it was presented in the Quality of Work Life Survey 1990 (http://www.stat.fi/meta/til/tyoolot_en.html). The 3 reply options were: regular daytime work (06.00 to 18.00 hr); two-shift work, regular evening work, or weekend work or other irregular working hours that do not include night work; regular or irregular three-shift work, or regular night-time work. Only in a few occupations all workers did regular daytime work without any irregular working hours. Concerning the cumulative index of night-time work, P (proportion) is the probability (no unit) of night-time work for a given occupation, L (level) is 1 (no unit) for all occupations, since the number of hours per week or of days per year of night-time work was not scored in the survey, and t is time (in years) spent in a given occupation. For example, the index of 10 means that a person was estimated to have worked in an occupation whose probability of night-time work is 1 for 10 years, or in an occupation whose probability of night-time work is 0.5 for 20 years, etc. Low, moderate and high exposures to night-time work refer to 1–10, 11–20, and 21 or more years of night-time work respectively. It was not possible to use regular daytime workers as the unexposed reference group, because the exposure assessment was not based on individual data but on the prevalence of night-time work in occupational groups.

Statistical analyses

For every occupation, the observed and expected numbers of cases were calculated for each gender, 5-year birth cohort, and 5-year calendar period. The expected number in each stratum was calculated by multiplying the number of person-years in that occupation by the cancer incidence rate of the entire Finnish population. The standardized incidence ratio used for this external comparison was defined as the ratio of observed to expected numbers of cases. These ratios were also calculated with adjustment for social class by splitting each stratum into 5 social classes and using the national reference rates specific for social classes.

A Poisson regression analysis of the stratum-specific observed numbers of NHL cases and person-years at risk was used to study the exposure-response patterns. The 95% confidence interval (95% CI) was calculated for the risk ratio (RR). In this internal comparison, the amount of night-time work was classified as high, moderate, low and none as the reference category. In addition to night-time work, we analyzed the effect of three-shift work on the occurrence of NHL but found no significant association with NHL. The lag times used for analysis were primarily 10 years and secondarily 2, 5, 15 and 20 years. The rationale for using a range of lag times was the fact that NHL is a heterogeneous group of malignant diseases and its clinical course may vary from months to years. Age and cohort period were taken as standard covariates for the models, whereas social class (socioeconomic status) was taken because it was a unique piece of information and reflects general health and life style. Other occupational exposures than working hours, smoking, etc. were not taken because none of them is known to have a robust association with NHL.

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Lag time of 10 years

The most consistent associations of night-time work with NHL were with a lag of 10 years. In the basic screening analysis, night-time work increased significantly the risk of NHL in men, the overall relative risk being 1.10 (95% CI of 1.03–1.19; see Table I), but not in women whose overall relative risk was 1.02 (95% CI of 0.94–1.12; see Table II). In the subsequent detailed analysis, the risk ratio was 1.28 (95% CI = 1.03–1.59) for those men who had a high exposure to night-time work as compared with those who had not been exposed to night-time work (P value for trend was 0.008).

Table I. Risk of Non-Hodgkin Lymphoma Relative to 2, 5, 10 and 20-Year Lag Cumulative Night-Time Shift Work Exposure in Men
Amount of exposureNumber of personsRisk ratio195% CIp-value
  • 1

    Risk ratios and their 95% confidence intervals (95% CI) adjusted for the age, social class and cohort period.

Basic analysis    
 10-year lag exposure    
  Unexposed (never)2,286   
  Exposed (ever)1,5271.101.03–1.190.01
Detailed analysis    
 2-year lag exposure    
  Unexposed2,209   
  Low1,3561.081.00–1.170.05
  Moderate1301.030.86–1.240.74
  High1181.200.99–1.450.06
 5-year lag exposure    
  Unexposed2,232   
  Low1,3451.101.02–1.180.02
  Moderate1301.050.88–1.260.59
  High1061.221.00–1.500.05
 10-year lag exposure    
  Unexposed2,286   
  Low1,3201.101.02–1.190.0108
  Moderate1161.020.84–1.230.8405
  High911.281.03–1.590.0254
 15-year lag exposure    
  Unexposed2,311   
  Low1,3221.101.02–1.190.01
  Moderate1141.150.95–1.390.15
  High661.230.96–1.580.11
 20-year lag exposure    
  Unexposed2,353   
  Low1,3121.101.02–1.180.02
  Moderate1011.251.02–1.540.03
  High471.200.89–1.610.23
Table II. Risk of Non-Hodgkin Lymphoma Relative to 2, 5, 10 and 20-Year Lag Cumulative Night-Time Shift Work Exposure in Women
Amount of exposureNumber of personsRisk ratio195% CIp-value
  • 1

    Risk ratios and their 95% confidence intervals (95% CI) adjusted for the age, social class and cohort period.

Basic analysis    
 10-year lag exposure    
  Unexposed (never)1,337   
  Exposed (ever)1,1571.020.94–1.120.58
Detailed analysis    
 2-year lag exposure    
  Unexposed1,294   
  Low9980.980.90–1.080.72
  Moderate1171.231.01–1.490.04
  High851.110.88–1.390.39
 5-year lag exposure    
  Unexposed1,313   
  Low9890.980.90–1.080.71
  Moderate1201.261.04–1.530.02
  High721.040.81–1.330.75
 10-year lag exposure    
  Unexposed1,337   
  Low9881.000.91–1.090.9970
  Moderate1121.241.02–1.510.0342
  High571.040.79–1.360.7936
 15-year lag exposure    
  Unexposed1,349   
  Low1,0041.000.92–1.100.93
  Moderate981.210.98–1.490.07
  High431.070.79–1.470.65
 20-year lag exposure    
  Unexposed1,364   
  Low1,0211.020.93–1.110.74
  Moderate811.180.94–1.480.16
  High281.090.75–1.600.65

Alternative lag times

Using the lag times of 20 and 15 years there were significant associations in men only. For the former the risk ratio was a significant one among those who had low or moderate exposures to night-time work, whereas for the latter the risk ratio was a significant one among those who had a low exposure to night-time work.

Using the lag time of 5 and 2 years there were significant associations in men and women. For men the risk ratios were significant ones among those who had low exposures to night-time work, while for women the risk ratios were significant ones among those who had moderate exposures to night-time work. In addition, the risk ratio was 1.22 (95% CI = 1.00–1.50) for men who had a high exposure to night-time work using the lag time of 5 years.

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Our current findings indicate that night-time work predisposes to NHL. There is a significant increase in the overall risk of NHL in men. The increased risk is rather consistent using different lag times, since men being exposed to night-time work had elevated risk ratios in all the comparisons. The primary analysis using the lag time of 10 years yielded in addition a significant trend, suggesting that there is a linear relationship between the magnitude of exposure and the increase in risk.

There is not a significant increase in the overall risk of NHL in women. This finding suggests that the results from the detailed analysis using a range of lag times needs to be interpreted with caution, since the 3 significant associations of moderate exposures to night-time work with the increased risk may be due to a chance only. In women, the risk ratios are elevated in all the comparisons except for those who had low exposures to night-time work using the lag times of 2–15 years.

The sex-specific findings may be due to sample sizes and the limited power to detect a significant association among women. Alternatively, they might be due to biological differences in some hormonal factor for example. One explanation can be based on the fact that cells of the human suprachiasmatic nuclei contain estrogen receptor-alpha, estrogen receptor-beta and progesterone receptors.17 The Per2 gene which provides a direct link between the suprachiasmatic nuclei and the peripheral oscillators12 also links the circadian cycle to the estrogen receptor-alpha signaling network,18 so that binding of PER2 enhances estrogen receptor-alpha degradation whereas suppression of PER2 levels leads to estrogen receptor-alpha stabilization. In addition, there appears to be a second link to estrogen signaling where estrogen receptor-beta expression fluctuates in peripheral tissues following a robust circadian pattern, with a peak at the light-dark transition, and is modulated by both the positive and negative limbs of the circadian clock.19 In tissues whom the 2 estrogen receptors occupy at similar levels a complex action is expected, since the effect of estrogen on these tissues may differ from day to night due to the expression ratio of the 2 estrogen receptors. Particular tissue-specific mechanisms of action related to estrogen for instance may modify the circadian clockwork and make a difference in the resilience of the circadian clock to night-time work and subsequently in the risk of NHL. However, these mechanisms of action, if any, await elucidation.

Earlier, it has been demonstrated that distraction of the circadian clock by night-time shift work increases the risk of cancer.20 Moreover, the efficacy of chemotherapies appears to be better at particular time of the day21 and the prognosis of cancer poorer in those patients having disrupted circadian rhythms.22 These observations give support to the assumption that there is a circadian effect on cancer. Subsequently, abnormalities in the circadian clockwork due to night-time work may predispose to cancers whose progression is influenced by the circadian effect.

Molecules of particular interest for the pathogenesis of NHL include the circadian clock genes such as PER2 and NPAS2, since mPER2 is a positive regulator of NPAS2 and stimulates the ARNTL-NPAS2 transcription complex.23 Modulation of NPAS2 activity by mPER2 involves heme interaction, and heme influences the expression of c-Myc and other cell-cycle regulators.23 In Npas2 mutant mice, c-Myc levels are deregulated similar to mPer2 mutant mice. Cobalamin (vitamin B12) competes with heme for binding in both mPER2 and NPAS2. It seems that the mPer2 gene contributes to tumor suppression by regulating DNA damage-responsive pathways.10 These actions may take place in the human body similar to these animal models.

In the USA, the analysis of 77,173 patients with the diagnosis of NHL demonstrated that NHL emerged more often during spring in the years of 1973–1999.24 Excess was observed during the period of March to April, and the B-cell origin subtype was diagnosed more frequently in March. Incidence during spring was high for patients aged 20 to 64 years. Whether this spring peak was due to diagnostic bias or other confounding factors, it needs further elucidation. One explanation may be a disruption of the circadian synchronization that takes place more easily during spring, because the increasing durations of exposure to sunlight is likely to accelerate phase delays of the circadian clockwork.25 Night-time shift work may compromise the circadian clockwork even further and subsequently add on the observed spring peak in the incidence of NHL.

It seems that the circadian clock controls not only the circadian and seasonal cycles but also modulate the immune response. Mice carrying a loss-of-function mutation in the Per2 gene lack the physiologic daily rhythm of interferon-gamma protein expression in the spleen and serum levels, the latter being lowered.26 Moreover, advances in the light-dark transitions hasten the death of aged mice.27 In humans, such advances can take place after flights crossing time zones or shift work schedules. The precise mechanism of action underlying this hazardous effect is not stress-related but links more likely to sleep deprivation or disruption of the immune system.

A growing proportion of people of working age are exposed to irregular shift work schedules. Irregular sleep-wake cycles and disrupted circadian rhythms cause abnormalities in the function and coordination of the principal and peripheral circadian clock and might predispose individuals to cancer, to NHL in particular. Because abnormalities in the circadian clockwork have a potentially big and growing impact on public health, their mechanisms of action on malignant growth are definitely worth of further studies.

As a limitation of the study design several possibilities for confounding exist including industrial exposures. Since individuals who work at night may also have lower exposures to sunlight, confounding by this mechanism is a definite possibility that needs to be explored, since the current evidence is conflicting.2830 However, recent findings derived from the analysis applying the Finnish job-exposure matrix for the assessment of exposure to ultraviolet radiation, concerning the risk of NHL, were consistent with a weak positive association only.31 The night is unnatural time to work from the physiological point of view, and therefore abnormalities in the circadian clockwork are most frequent and potentially most severe after exposure to night-time work in particular. It may explain our findings herein that NHL was associated with night-time work only and not with shift work in general. Another limitation is that the exposure assessment was not based on individual data but on the prevalence of night-time work in occupational groups. A third limitation is that subtypes of NHL cannot be separated in the Finnish Cancer Registry coding system. The whole-population cancer registration systems such as the Finnish one that need to keep up comparable classifications over decades cannot follow every change in diagnostic classifications.

In conclusion, our findings indicate that night-time work predisposes to NHL and suggest that the pathogenesis involves abnormalities in the circadian clockwork. Night-time work challenges the circadian clockwork to a limit and may thereby cause failures in control of the cell division cycle and predispose to NHL in particular.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The study was supported in part by grants from Academy of Finland (to TP).

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References