Acoustic neuroma risk in relation to mobile telephone use: results of the INTERPHONE international case-control study

Background: The rapid increase in mobile telephone use has generated concern about possible health risks of radiofrequency electromagnetic fields from these devices. Methods: A case–control study of 1105 patients with newly diagnosed acoustic neuroma (vestibular schwannoma) and 2145 controls was conducted in 13 countries using a common protocol. Past mobile phone use was assessed by personal interview. In the primary analysis, exposure time was censored at one year before the reference date (date of diagnosis for cases and date of diagnosis of the matched case for controls); analyses censoring exposure at five years before the reference date were also done to allow for a possible longer latent period. Results: The odds ratio (OR) of acoustic neuroma with ever having been a regular mobile phone user was 0.85 (95% confidence interval 0.69–1.04). The OR for ≥10 years after first regular mobile phone use was 0.76 (0.52–1.11). There was no trend of increasing ORs with increasing cumulative call time or cumulative number of calls, with the lowest OR (0.48 (0.30–0.78)) observed in the 9th decile of cumulative call time. In the 10th decile (≥1640 h) of cumulative call time, the OR was 1.32 (0.88–1.97); there were, however, implausible values of reported use in those with ≥1640 h of accumulated mobile phone use. With censoring at 5 years before the reference date the OR for ≥10 years after first regular mobile phone use was 0.83 (0.58–1.19) and for ≥1640 h of cumulative call time it was 2.79 (1.51–5.16), but again with no trend in the lower nine deciles and with the lowest OR in the 9th decile. In general, ORs were not greater in subjects who reported usual phone use on the same side of the head as their tumour than in those who reported it on the opposite side, but it was greater in those in the 10th decile of cumulative hours of use. Conclusions: There was no increase in risk of acoustic neuroma with ever regular use of a mobile phone or for users who began regular use 10 years or more before the reference date. Elevated odds ratios observed at the highest level of cumulative call time could be due to chance, reporting bias or a causal effect. As acoustic neuroma is usually a slowly growing tumour, the interval between introduction of mobile phones and occurrence of the tumour might have been too short to observe an effect, if there is one.     

Impact of random and systematic recall errors and selection bias in case--control studies on mobile phone use and brain tumors in adolescents (CEFALO study)

Whether the use of mobile phones is a risk factor for brain tumors in adolescents is currently being studied. Case--control studies investigating this possible relationship are prone to recall error and selection bias. We assessed the potential impact of random and systematic recall error and selection bias on odds ratios (ORs) by performing simulations based on real data from an ongoing case--control study of mobile phones and brain tumor risk in children and adolescents (CEFALO study). Simulations were conducted for two mobile phone exposure categories: regular and heavy use. Our choice of levels of recall error was guided by a validation study that compared objective network operator data with the self-reported amount of mobile phone use in CEFALO. In our validation study, cases overestimated their number of calls by 9% on average and controls by 34%. Cases also overestimated their duration of calls by 52% on average and controls by 163%. The participation rates in CEFALO were 83% for cases and 71% for controls. In a variety of scenarios, the combined impact of recall error and selection bias on the estimated ORs was complex. These simulations are useful for the interpretation of previous case-control studies on brain tumor and mobile phone use in adults as well as for the interpretation of future studies on adolescents.     

Mobile phone use and location of glioma: A case–case analysis

We assessed a new approach for evaluating the glioma risk among users of mobile phones to focus on the part of the brain most heavily exposed to radiofrequency electromagnetic fields from mobile phones. The tumor midpoint was defined from radiological imaging. A case–case analysis with 99 gliomas was performed using logistic regression. The exposed cases were those with the tumor mid‐point within 4.6 cm from the line between the mouth and the external meatus of the ear, representing the most likely location of the mobile phone (the source of exposure). Alternative analyses based on various indicators of mobile phone use as the outcome were also carried out. The majority of cases were regular mobile phone users. A slightly higher proportion of gliomas among mobile phone users than non‐users occurred within 4.6 cm from the presumed location of the mobile phone (28% vs. 14%). Modestly elevated odds ratios were observed for several indicators of mobile phone use, but without an exposure gradient. The highest odds ratios were found for contralateral and short‐term use. Our results, though limited by the small sample size, demonstrate that detailed information on tumor location allows evaluation of the risk related to the most heavily exposed part of the brain, representing direct evaluation of the possible local carcinogenic effects of the radiofrequency fields. However, field strength varies between users and over time also within a given anatomic site, due to the output power of the phone. Collaborative analysis of a larger sample is planned. Bioelectromagnetics 30:176–182, 2009. © 2009 Wiley‐Liss, Inc.     

Photodynamic therapy and tumor imaging of hypericin-treated squamous cell carcinoma

Background

The use of cellular and cordless telephones has increased dramatically during the last decade. There is concern of health problems such as malignant diseases due to microwave exposure during the use of these devices. The brain is the main target organ.

Methods

Since the second part of the 1990's we have performed six case-control studies on this topic encompassing use of both cellular and cordless phones as well as other exposures. Three of the studies concerned brain tumours, one salivary gland tumours, one non-Hodgkin lymphoma (NHL) and one testicular cancer. Exposure was assessed by self-administered questionnaires.

Results

Regarding acoustic neuroma analogue cellular phones yielded odds ratio (OR) = 2.9, 95 % confidence interval (CI) = 2.0–4.3, digital cellular phones OR = 1.5, 95 % CI = 1.1–2.1 and cordless phones OR = 1.5, 95 % CI = 1.04–2.0. The corresponding results were for astrocytoma grade III-IV OR = 1.7, 95 % CI = 1.3–2.3; OR = 1.5, 95 % CI = 1.2–1.9 and OR = 1.5, 95 % CI = 1.1–1.9, respectively. The ORs increased with latency period with highest estimates using > 10 years time period from first use of these phone types. Lower ORs were calculated for astrocytoma grade I-II. No association was found with salivary gland tumours, NHL or testicular cancer although an association with NHL of T-cell type could not be ruled out.

Conclusion

We found for all studied phone types an increased risk for brain tumours, mainly acoustic neuroma and malignant brain tumours. OR increased with latency period, especially for astrocytoma grade III-IV. No consistent pattern of an increased risk was found for salivary gland tumours, NHL, or testicular cancer.     

Mobile phones, radiofrequency fields, and health effects in children--epidemiological studies

In 2004, when WHO organized a workshop on children's sensitivity to electromagnetic fields, very few studies on radiofrequency fields were available. With the recent increase in mobile phone use among children and adolescents, WHO has identified studies on health effects in this age-group as a high priority research area. There are no empirical data supporting the notion that children and adolescents are more susceptible to RF exposure, but the number of studies is still relatively small. There are a few cross-sectional studies on well-being, cognitive effects and behavioral problems, and some cohort studies, mainly of maternal use of mobile phones during pregnancy. Cancer outcomes have been studied in relation to environmental RF exposure, e.g. from transmitters, and only one study on mobile phone use in children and adolescents and brain tumor risk has been published. Several methodological limitations need to be taken into consideration when interpreting the findings of the epidemiological studies. The cross-sectional design does not allow determination of the temporal sequence of exposure and outcome, and for several outcomes there is a large potential for reversed causality, i.e. that the outcome causes an increased RF exposure rather than the opposite. Biases such as recall errors in self-reported mobile phone use, lack of confounding control, e.g. of other aspects of mobile phone use than RF fields, trained behaviors, and pubertal development, makes causal interpretations impossible. Future studies need to include prospectively collected exposure information, incident outcomes, and proper confounding control. Monitoring of brain tumor incidence trends is strongly recommended.     

Systematic review of wireless phone use and brain cancer and other head tumors

We conducted a systematic review of scientific studies to evaluate whether the use of wireless phones is linked to an increased incidence of the brain cancer glioma or other tumors of the head (meningioma, acoustic neuroma, and parotid gland), originating in the areas of the head that most absorb radiofrequency (RF) energy from wireless phones. Epidemiology and in vivo studies were evaluated according to an agreed protocol; quality criteria were used to evaluate the studies for narrative synthesis but not for meta-analyses or pooling of results. The epidemiology study results were heterogeneous, with sparse data on long-term use (≥ 10 years). Meta-analyses of the epidemiology studies showed no statistically significant increase in risk (defined as P < 0.05) for adult brain cancer or other head tumors from wireless phone use. Analyses of the in vivo oncogenicity, tumor promotion, and genotoxicity studies also showed no statistically significant relationship between exposure to RF fields and genotoxic damage to brain cells, or the incidence of brain cancers or other tumors of the head. Assessment of the review results using the Hill criteria did not support a causal relationship between wireless phone use and the incidence of adult cancers in the areas of the head that most absorb RF energy from the use of wireless phones. There are insufficient data to make any determinations about longer-term use (≥ 10 years).     

Analysis of ear side of mobile phone use in the general population of Japan

This study aimed to clarify the distribution of the ear side of mobile phone use in the general population of Japan and clarify what factors are associated with the ear side of mobile phone use. Children at elementary and junior high schools (n = 2,518) and adults aged ≥20 years (n = 1,529) completed an Internet‐based survey. Data were subjected to a logistic regression analysis. In children, due to the tendency to use the dominant hand, we analyzed the factors associated with the use of right ear in right‐handed people. Statistically significant differences were observed only in talk time per call (odds ratio (OR) = 2.17; 95% confidence interval (CI): 1.22–3.99). In adults, due to the tendency to use the left ear, we analyzed factors associated with the use of left ear in right‐handed people. Significant differences were observed in those aged 30–39 years (OR = 2.55; 95% CI: 1.79–3.68), those aged 40–49 years (OR = 3.08; 95% CI: 2.15–4.43), those aged >50 years (OR = 1.85; 95% CI: 1.20–2.85), and in those with a percentage of total talk time when using mobile phones at work of 51–100% (OR = 1.75; 95% CI: 1.21–2.55). We believe that future epidemiological studies on mobile phone use can be improved by considering the trends in mobile phone use identified in this study. Bioelectromagnetics. 39:53–59, 2018. © 2017 Wiley Periodicals, Inc.     

An evaluation of self‐reported mobile phone use compared to billing records among a group of engineers and scientists

Most epidemiologic studies of potential health impacts of mobile phones rely on self‐reported information, which can lead to exposure misclassification. We compared self‐reported questionnaire data among 60 participants, and phone billing records over a 3‐year period (2002–2004). Phone usage information was compared by the calculation of the mean and median number of calls and duration of use, as well as correlation coefficients and associated P‐values. Average call duration from self‐reports was slightly lower than billing records (2.1 min vs. 2.8 min, P = 0.01). Participants reported a higher number of average daily calls than billing records (7.9 vs. 4.1, P = 0.002). Correlation coefficients for average minutes per day of mobile phone use and average number of calls per day were relatively high (R = 0.71 and 0.69, respectively, P < 0.001). Information reported at the monthly level tended to be more accurate than estimates of weekly or daily use. Our findings of modest correlations between self‐reported mobile phone usage and billing records and substantial variability in recall are consistent with previous studies. However, the direction of over‐ and under‐reporting was not consistent with previous research. We did not observe increased variability over longer periods of recall or a pattern of lower accuracy among older age groups compared with younger groups. Study limitations included a relatively small sample size, low participation rates, and potential limited generalizability. The variability within studies and non‐uniformity across studies indicates that estimation of the frequency and duration of phone use by questionnaires should be supplemented with subscriber records whenever practical. Bioelectromagnetics 32:37–48, 2011. © 2010 Wiley‐Liss, Inc.     

No effects of mobile phone use on cortical auditory change‐detection in children: An ERP study

We investigated the effect of mobile phone use on the auditory sensory memory in children. Auditory event‐related potentials (ERPs), P1, N2, mismatch negativity (MMN), and P3a, were recorded from 17 children, aged 11–12 years, in the recently developed multi‐feature paradigm. This paradigm allows one to determine the neural change‐detection profile consisting of several different types of acoustic changes. During the recording, an ordinary GSM (Global System for Mobile Communications) mobile phone emitting 902 MHz (pulsed at 217 Hz) electromagnetic field (EMF) was placed on the ear, over the left or right temporal area (SAR_1g = 1.14 W/kg, SAR_10g = 0.82 W/kg, peak value = 1.21 W/kg). The EMF was either on or off in a single‐blind manner. We found that a short exposure (two 6 min blocks for each side) to mobile phone EMF has no statistically significant effects on the neural change‐detection profile measured with the MMN. Furthermore, the multi‐feature paradigm was shown to be well suited for studies of perception accuracy and sensory memory in children. However, it should be noted that the present study only had sufficient statistical power to detect a large effect size. Bioelectromagnetics 31:191–199, 2010. © 2009 Wiley‐Liss, Inc.     

The association between socioeconomic status and exposure to mobile telecommunication networks in children and adolescents

A potential association between socioeconomic status (SES) and self‐reported use of mobile phones has been investigated in a few studies. If measured exposure to mobile phone networks differs by SES in children, it has not yet been studied. Interview data of 1,481 children and 1,505 adolescents on participants' mobile phone use, socio‐demographic characteristics and potential confounders were taken from the German MobilEe‐study. Sociodemographic data was used to stratify participants into three “status groups” (low, middle, high). Using a personal dosimeter, we obtained an exposure profile over 24 h for each of the participants. Exposure levels during waking hours were expressed as mean percentage of the reference level. Children with a low SES were more likely to own a mobile phone (OR 2.1; 95% CI: 1.1–3.9) and also reported to use their mobile phone longer per day (OR 2.4; 95% CI: 1.1–5.4) than children with a high SES. For adolescents, self‐reported duration of mobile phone use per day was also higher with a low SES (OR: 3.4; 95% CI: 1.4–8.4) compared with a high SES. No association between SES and measured exposure to mobile telecommunication networks was seen for children or adolescents. Mobile phone use may differ between status groups with higher use among disadvantaged groups. However, this does not result in higher overall exposure to mobile telecommunication networks. Whether short duration of own mobile phone use or the small numbers of participants with a low SES are causal, have to be investigated in further studies. Bioelectromagnetics 31:20–27, 2010. © 2009 Wiley‐Liss, Inc.     

Mobile phone ownership and use among school children in three Hungarian cities

In spite of rapid worldwide increase in mobile phone use and public concerns about associated potential health effects, little is known about patterns of mobile phone ownership and use in the general population and among children. In April 2005, we conducted a survey of mobile phone ownership and use among fourth grade school children in three Hungarian cities. From 24 schools, 1301 student filled out a short, self-administered questionnaire on basic demographics, mobile phone ownership, pattern of mobile phone use, and after-school activities. Overall, 989 students (76%) owned a mobile phone. Three hundred thirteen students (24%) used a mobile phone daily to make phone calls, and an additional 427 students (33%) used mobile phones for phone calls at least several times per week. Sixty-six students (5%) sent text messages daily and an additional 308 students (24%) sent text messages at least several times per week. Girls, children with no siblings, children who were members of a sport club, and children who played computer games daily were more likely to own and use mobile phones regularly. A higher number of socially disadvantaged children in a class predicted lower likelihood of regular mobile phone use among children. Our results suggest that mobile phone ownership and regular use is highly prevalent among school children in Hungary. Due to rapid changes in ownership patterns follow up surveys will be required to obtain information on temporal trends and changes in mobile phone ownership and pattern of use among school children.     

Effect of chronotype on emotional processing and risk taking

There is increasing evidence to suggest that late chronotypes are at increased risk for depression. The putative psychological mechanisms underpinning this risk, however, have not been fully explored. The aim of the present study was to examine whether, similar to acutely depressed patients and other “at risk” groups, late chronotype individuals display biases in tasks assaying emotional face recognition, emotional categorisation, recognition and recall and attention. Late chronotype was associated with increased recognition of sad facial expressions, greater recall and reduced latency to correctly recognise previously presented negative personality trait words and reduced allocation of attentional resources to happy faces. The current results indicate that certain negative biases in emotional processing are present in late chronotypes and may, in part, mediate the vulnerability of these individuals to depression. Prospective studies are needed to establish whether the cognitive vulnerabilities reported here predict subsequent depression.     

Electromagnetic fields and epidemiology: An overview inspired by the fourth course at the International School of Bioelectromagnetics

The fourth course at the International School of Bioelectromagnetics addressed various aspects of the epidemiology of exposure to electromagnetic fields (EMF). In this overview, inspired by the lectures and the discussions among participants, we summarize current knowledge on exposure to EMF and disease risk, with emphasis on studies of use of mobile phones and brain tumours and exposure to power lines and childhood leukaemia. Sources of bias and error hamper straightforward conclusions in some areas and, in order to move forward, improvements in study design and exposure assessment are necessary. The scientific evidence available to date on possible long‐term effects from exposure to ELF and RF fields is not strong enough to revise current protection limits based on the known acute effects of such exposures. Precautionary measures may be considered to reduce ELF exposure of children or exposure to RF during mobile phone use, keeping in mind that it is unclear whether they involve any preventive benefit. Possible health effects from mobile phone use in adults and in children should be investigated further by prospective epidemiological studies with improved exposure assessment and brain tumour incidence rates should be monitored. Further studies on the relation between childhood leukaemia and ELF magnetic fields would be worthwhile if they focus on heavily exposed groups and attempt to minimize possible selection bias. In conclusion, epidemiological studies conducted with appropriate diligence can play a key role in finding the answers. Bioelectromagnetics 30:511–524, 2009. © 2009 Wiley‐Liss, Inc.     

Mobile Phone Use,Blood Lead Levels,and Attention Deficit Hyperactivity Symptoms in Children: A Longitudinal Study

Background

Concerns have developed for the possible negative health effects of radiofrequency electromagnetic field (RF-EMF) exposure to children’s brains. The purpose of this longitudinal study was to investigate the association between mobile phone use and symptoms of Attention Deficit Hyperactivity Disorder (ADHD) considering the modifying effect of lead exposure.

Methods

A total of 2,422 children at 27 elementary schools in 10 Korean cities were examined and followed up 2 years later. Parents or guardians were administered a questionnaire including the Korean version of the ADHD rating scale and questions about mobile phone use, as well as socio-demographic factors. The ADHD symptom risk for mobile phone use was estimated at two time points using logistic regression and combined over 2 years using the generalized estimating equation model with repeatedly measured variables of mobile phone use, blood lead, and ADHD symptoms, adjusted for covariates.

Results

The ADHD symptom risk associated with mobile phone use for voice calls but the association was limited to children exposed to relatively high lead.

Conclusions

The results suggest that simultaneous exposure to lead and RF from mobile phone use was associated with increased ADHD symptom risk, although possible reverse causality could not be ruled out.     

Long‐term digital mobile phone use and cognitive decline in the elderly

Research on long‐term exposure to electromagnetic fields on cognition is lacking. We investigated the associations between frequent digital mobile phone use and global and domain‐specific cognitive changes in older persons, a vulnerable group experiencing age‐associated cognitive decline. We assessed 871 non‐demented Chinese participants in the Singapore Longitudinal Ageing Studies cohort on the frequency of digital mobile phone use, neurocognitive performance and confounding variables at baseline, and neurocognitive performance at the 4‐year follow‐up. Findings showed that digital mobile phone users were typically self‐selected to possess characteristics favoring better cognitive functioning and concomitantly demonstrate better performance on cognitive tasks. There was evidently no significant deleterious effect of digital mobile phone use on cognitive functioning in older people. Findings suggest, however, that digital mobile phone use may have an independent facilitating effect on global and executive functioning. Bioelectromagnetics 33:176–185, 2012. © 2011 Wiley Periodicals, Inc.     

Radiation from mobile phone systems: Is it perceived as a threat to people's health?

The aim of this study was to explore the prevalence, nature and determinants of concerns about mobile phone radiation. We used data from a 2006 telephone survey of 1004 people aged 15+ years in Denmark. Twenty‐eight percent of the respondents were concerned about exposure to mobile phone radiation; radiation from masts was of concern to about 15%. In contrast, 82% were concerned about pollution. Nearly half of the respondents considered the mortality risk of 3G phones and masts to be of the same order of magnitude as being struck by lightning (0.1 fatalities per million people per year) while 7% thought it was equivalent to tobacco‐induced lung cancer (～500 fatalities per million per year). Among women, concerns about mobile phone radiation were positively associated with educational attainment, perceived mobile phone mortality risk and concerns about unknown consequences of new technologies. More than two thirds of the respondents felt that they had received inadequate public information about the 3G system. The results of the study indicate that the majority of the population has little concern about mobile phone radiation while a small minority is very concerned. Bioelectromagnetics 30:393–401, 2009. © 2009 Wiley‐Liss, Inc.     

Birth sex ratio in captive mammals: Patterns,biases, and the implications for management and conservation

Sex allocation theory predicts that a female should produce the offspring of the sex that most increases her own fitness. For polygynous species, this means that females in superior condition should bias offspring production toward the sex with greater variation in lifetime reproductive success, which is typically males. Captive mammal populations are generally kept in good nutritional condition with low levels of stress, and thus populations of polygynous species might be expected to have birth sex ratios biased toward males. Sex allocation theory also predicts that when competition reduces reproductive success of the mother, she should bias offspring toward whichever sex disperses. These predicted biases would have a large impact on captive breeding programs because unbalanced sex ratios may compromise use of limited space in zoos. We examined 66 species of mammals from three taxonomic orders (primates, ungulates, and carnivores) maintained in North American zoos for evidence of birth sex ratio bias. Contrary to our expectations, we found no evidence of bias toward male births in polygynous populations. We did find evidence that birth sex ratios of primates are male biased and that, within primates, offspring sex was biased toward the naturally dispersing sex. We also found that most species experienced long contiguous periods of at least 7 years with either male‐ or female‐biased sex ratios, owing in part to patterns of dispersal (for primates) and/or to stochastic causes. Population managers must be ready to compensate for significant biases in birth sex ratio based on dispersal and stochasticity. Zoo Biol 19:11–25, 2000. © 2000 Wiley‐Liss, Inc.     

Accessibility maps as a tool to predict sampling bias in historical biodiversity occurrence records

Historical biodiversity occurrence records are often discarded in spatial modeling analyses because of a lack of a method to quantify their sampling bias. Here we propose a new approach for predicting sampling bias in historical written records of occurrence, using a South African example as proof of concept. We modelled and mapped accessibility of the study area as the mean of proximity to freshwater and European settlements. We tested the model's ability to predict the location of historical biodiversity records from a dataset of 2612 large mammal occurrence records collected from historical written sources in South Africa in the period 1497–1920. We investigated temporal, spatial and environmental biases in these historical records and examined if the model prediction and occurrence dataset share similar environmental bias. We find a good agreement between the accessibility map and the distribution of sampling effort in the early historical period in South Africa. Environmental biases in the empirical data are identified, showing a preference for lower maximum temperature of the warmest month, higher mean monthly precipitation, higher net primary productivity and less arid biomes than expected by a uniform use of the study area. We find that the model prediction shares similar environmental bias as the empirical data. Accessibility maps, built with very simple statistical rules and in the absence of empirical data, can thus predict the spatial and environmental biases observed in historical biodiversity occurrence records. We recommend that this approach be used as a tool to estimate sampling bias in small datasets of occurrence and to improve the use of these data in spatial analyses in ecological and conservation studies.     

Factors that influence the radiofrequency power output of GSM mobile phones

Epidemiological studies of mobile phone use and risk of brain cancer have relied on self-reported use, years as a subscriber, and billing records as exposure surrogates without addressing the level of radiofrequency (RF) power output. The objective of this study was to measure environmental, behavioral and engineering factors affecting the RF power output of GSM mobile phones during operation. We estimated the RF-field exposure of volunteer subjects who made mobile phone calls using software-modified phones (SMPs) that recorded output power settings. Subjects recruited from three geographic areas in the U.S. were instructed to log information (place, time, etc.) for each call made and received during a 5-day period. The largest factor affecting energy output was study area, followed by user movement and location (inside or outside), use of a hands-free device, and urbanicity, although the two latter factors accounted for trivial parts of overall variance. Although some highly statistically significant differences were identified, the effects on average energy output rate were usually less than 50% and were generally comparable to the standard deviation. These results provide information applicable to improving the precision of exposure metrics for epidemiological studies of GSM mobile phones and may have broader application for other mobile phone systems and geographic locations.     

Coffee and Pancreatic Cancer in a Rural California County

In a study of the risk of fatal pancreatic cancer according to intake of regular and decaffeinated coffee, cases (N = 30) and controls (N = 47) were identified from death certificates and matched for age (± 5 years), sex, ethnicity, and date of death (± 5 years). Telephone interviews were completed with survivors of about 80% of both groups. In smokers, the relative risk for high (3 + cups) versus low (<3 cups) intake of regular coffee was 4.3 (P < .05), and high verus low decaffeinated coffee, 5.5 (P < .05). In nonsmokers, neither type of coffee influenced the risk. Mean daily intakes of alcohol and cigarettes were virtually identical in cases and controls, although cases had accumulated nonsignificantly more pack-years. Intakes of regular and decaffeinated coffee were uncorrelated, and the smoking-coffee interaction could not be readily explained by recall bias. If coffee intake increases the risk of pancreatic cancer, the mechanism could depend heavily on smoking.