Wednesday March 9 2011
Nail samples were taken up to 13 years before cancer developed
“Toenails can show if you are likely to get lung cancer,” reported The Sun. It said that a study measuring nicotine levels in toenail clippings from both smokers and non-smokers found that people with the highest nicotine readings were over three times more likely to get cancer than those with the lowest levels.
This research compared nicotine levels in toenail samples from 210 men who developed lung cancer between 1988 and 2000, and 630 men who did not. The samples were taken in 1987, up to 13 years before cancer developed.
The results – that toenail nicotine levels roughly equated to men’s smoking habits when samples were taken and that more men with cancer were smokers – are to be expected. Men with cancer also had higher nicotine levels in their nails than those without cancer.
As an objective measure of nicotine exposure, this technique may have academic or other specialist uses. However, in the absence of genetic or physiological tests, a person’s smoking habits remain the most effective way of predicting their risk of lung cancer. The small sample of men that was analysed means that this study cannot accurately determine how efficiently nail nicotine levels can predict smoking exposure or cancer. Studies involving a much larger sample are needed.
Where did the story come from?
The study was carried out by researchers from the School of Medicine at the University of California. Funding was provided by grants from the Flight Attendant Medical Research Institute and the National Cancer Institute. The study was published in the peer-reviewed American Journal of Epidemiology.
Though the newspapers reported the main findings of the study, they did not consider the limitations of this research. For example, when the participants were grouped according to their nicotine levels, there were small numbers in each group. This increases the likelihood that the results may be inaccurate. There are also wider issues and implications to consider, such as the questionable value of telling someone their nicotine levels and, from this, implying that they are "safe" or "at risk" if they continue smoking.
What kind of research was this?
This study investigated whether there is an association between nicotine levels in people’s toenails and their risk of lung cancer. This study was a case control study nested within a cohort study. This means that the researchers took people with lung cancer (the cases) and without the disease (the controls) from a large cohort study, and compared them in a smaller, separate study. The case and control groups were matched for age, and toenail clippings were taken before they were diagnosed with cancer.
The researchers’ theory was that toenail clippings would strongly predict the risk of lung cancer. The use of nail clippings to determine nicotine levels in a person’s body, rather than relying on people to report how much they smoke, could be considered a more “objective marker” of smoking, the main cause of lung cancer.
The study’s design was appropriate for investigating the researchers' theory. A cohort study, where researchers look at an exposure and then see whether this is associated with the later development of a disease, is an accepted method of identifying cause-and-effect associations. The case-control design is usually used for rarer disease outcomes, where the cohort size would have to be very large and people followed over an extensive period to allow for the development of a reasonable number of outcomes. Lung cancer is quite common, as is smoking, so a cohort design is appropriate. This study incorporated elements of both study designs. The results would have been more robust had the participants initially been recruited for the specific purpose of examining how nicotine levels in toenail samples taken at the start of the study related to later development of cancer (i.e. analysing a full cohort, rather than a case control within a cohort).
What did the research involve?
In 1986, 51,529 men were recruited in the US to take part in the Health Professionals Follow-up Study. The men were mostly white, and were all health professionals aged 40-75 years old (average age 62). At the start of the study, the men completed a questionnaire by mail on their medical history and lifestyle. The men were contacted every two years and were asked questions, including whether they had cancer.
In 1987, 65% of the cohort (33,737 men) provided toenail samples. Only these men were included in the current study. In these men, 221 lung cancers were diagnosed between 1988 and 2000. Nail samples were analysed for nicotine levels at a specialist laboratory in New Zealand, and nicotine levels split into five categories, from lowest to highest. Cancers were identified by questionnaire responses and then verified through medical records and pathology reports.
Men with cancer were randomly matched to three men from the study who did not have cancer, based on age and when their nail clipping was returned (666 men). The researchers reported that as most of the men with cancer were active smokers, they matched each of them to three controls to ensure that about one in three controls was a smoker. This would give a comparable number of smokers with cancer and smokers without cancer.
Smoking had been assessed in a questionnaire in 1986 and people were classified as “never”, “past” (had previously smoked at least 20 packs) or “current” smokers (with intake classified as 1-24 or 25-plus cigarettes a day). After excluding any cancers diagnosed before the nail samples had been taken, as well as those in men with no information on smoking status, the researchers were left with 210 cases and 630 controls (a sample of 840 men overall).
The researchers found that smoking and physical activity were both independently associated with cancer (the cases smoked a significantly higher number of cigarettes for a longer time than controls, and cases did less exercise than controls). Therefore, when examining the relationship between nicotine levels and cancer, they adjusted their analyses for these potential confounding factors.
What were the basic results?
The 840 men were roughly evenly distributed across the five categories of nail nicotine levels. As would be expected, levels of nicotine in the nail generally reflected smoking history: 85% of men whose nail nicotine level was in the highest fifth (quintile) were smokers at the time the sample was taken, compared to 0% of men in the lowest quintile. By comparison, 29% of those in the lowest quintile had never smoked, while 6% of those in the highest quintile had never smoked. Of the men who reported being past smokers, 21% were in each of the lowest three quintiles, 23% had levels in the second-highest quintile, and 13% had levels in the highest quintile.
As expected, cases were more likely to have smoked than controls. At the beginning of the study when the nail samples were taken, 37% of cases reported being current smokers, 54% reported being past smokers, and 9% reported never smoking. Compared to this, 6% of controls reported they were current smokers, 48% reported being past smokers and 45% reported never smoking. The average nail nicotine level among cases was 0.95ng/mg and among controls was 0.25ng/mg.
When the authors adjusted their analyses for how much and for how long the men smoked and their physical activity levels, men with nicotine levels in the highest quintile had a more than trebled risk of lung cancer (relative risk 3.57, 95% confidence interval 1.73 to 7.37). Associations between lung cancer and any lower levels of nicotine in the nails were non-significant.
How did the researchers interpret the results?
The researchers concluded that toenail nicotine levels can strongly predict the risk of lung cancer, independent of smoking history.
This research, involving 210 men with lung cancer and 630 men without, looked at levels of nicotine measured in toenail samples provided 1-13 years earlier (depending on when cancer was diagnosed). The results are somewhat as expected, with nicotine levels in nail samples roughly equating to the men’s reported smoking history. Nicotine levels in current smokers were generally the highest, men who had never smoked had the lowest nicotine levels, and past smokers were more evenly spread across the five categories of nicotine level. There were also, as expected, more smokers among the men with lung cancer than those without the disease, and men with cancer had on average higher nail nicotine levels.
In general, the research supports what is already well known – that smoking or smoke exposure is the biggest risk factor for lung cancer. How accurately nicotine levels in nails can predict cancer risk is slightly less clear. In the current study, nicotine levels in the highest quintile (average 1.28ng/mg) were predictive of cancer, but the confidence intervals surrounding the 3.57 relative risk were quite wide (95% confidence interval 1.73 to 7.37). This suggests that the trebling in lung cancer risk in men with the highest nicotine levels may not be an accurate estimate.
In the other nicotine level categories, while there was a correlation between increasing cancer risk and higher nicotine levels, these associations were not significant. In addition, there were only 210 cases of cancer. When they were subdivided into the five categories according to their nicotine levels, the number of cases in each category is small, which results in these unclear risk associations. A much larger sample of cases and controls would give clearer results. Research in women is also needed.
A further point is that nicotine levels in nail samples were not compared to other objective measures of nicotine exposure, such as hair samples. Further research into this would be needed to better establish the accuracy of nicotine in nail samples for predicting lung cancer risk.
It is well known that smoking is the main cause of lung cancer. An objective measure of a smoker’s tobacco exposure, such as nail nicotine levels, may simply confirm what smokers already know, that they are increasing their risk of cancer by smoking. However, there may be a practical use for these findings in research. The researchers say that studies comparing smokers to people who have never smoked often fail to take into account exposure to second-hand smoke, an established risk factor for lung cancer. They say that this, in addition to imperfect measures of smoking among active smokers (usually by questionnaire), may lead to them being wrongly classified. There may be other uses for an objective reading of a non-smoker’s exposure to nicotine.
The general health message remains the same: the best way to reduce your risk of lung cancer, as well as numerous other respiratory and cardiovascular diseases, cancers and disease complications, is to stop smoking.