Tobacco, Alcohol Underwriting in North America by Hank George, FALU 

We recently completed a survey of US and Canadian life insurers asking about underwriting practices and perceptions related to assessing tobacco/nicotine, alcohol and drug use. Over 100 carriers participated.

This essay covers the survey findings most likely to be of interest to underwriters and other insurance professionals in the Australasian market.


Tobacco Use

The survey began with a question asking respondents how often their applicants prevaricate when they say they do not use tobacco. This is a critical matter in the U.S. with the advent of so-called accelerated underwriting programs wherein we forego routine cotinine screening on some portion of applications.

Before discussing survey results further, let us define what we mean by accelerated underwriting.

By now, the majority of American insurers are either offering and or planning to roll out programs wherein individuals within stipulated age and sum insured limits can be assessed without having routine paramedicals and laboratory tests. In order to qualify, the applicant must be deemed a preferred risk based on his medical history and the results of several immediate-access underwriting resources.

At this time, some carriers are offering as much as $2.5 million of cover, through ages as old as 65, on an accelerated basis. This said, the portion of qualifying applicants is far lower at ages 55-65 as compared to between ages 18 and 45.

Because there are no laboratory tests done, we do not have the results of our staple marker for tobacco use: cotinine, a prominent metabolite of nicotine.

With this background, you see why tobacco use nondisclosure could have a major impact, doubly so because the same premium rates are used for accelerated policies as those that are more fully underwritten.

Back to the survey results…

One-third of respondents said less than 5% of coverage seekers do not disclose their tobacco use, followed by 44% reporting between 5% and 10% nondisclosure and the remainder saying that over 10% of applicants fail to acknowledge current tobacco consumption.

These are data based on full underwriting where alert applicants know that we are doing “nicotine” testing. It is likely that the rate of “smoker’s amnesia” will be higher in the absence of testing on accelerated cases.

Given the contestability clause in our policies, if these non-disclosers manage to avoid post-issue detection for 24 months, they will have beaten the system, so to speak. This is because American insurers have just 2 years from date of issue to reform the policy (change the higher premium rate), rescind coverage altogether or, for that matter, to deny a claim on the basis of incident nondisclosure.

Roughly 60% of carriers ask applicants how long they have been smoking and 47% inquire as to the average amount consumed per day. There is no point in asking either question unless they ask both, using responses to calculate pack-years of cigarette smoking.

One “pack-year” is defined as smoking one pack (20 cigarettes) per day for 12 months. If the applicant smokes 2 packs daily, he acquires 2 pack-years in 1 calendar year; conversely, if you smoke a half pack per day it takes 24 months to acquire 1 pack year of exposure.

Only 13% of carriers calculate pack-years, 8% are considering this practice and additional 8% either stopped or decided not to do this. The rest have not given this matter any consideration.

The driver of mortality in cigarette smoking is pack-years, not current use. Many studies have shown that a threshold in the range of 30-40 pack-years is a viable cut point beyond which the effects of smoking never wear off and thus have lifelong excess mortality implications.

Most insurers underwrite cigar smokers on the same basis as cigarette users. Unfortunately, because of pressure brought to bear on certain quarters, we came to embrace (goofy) “occasional cigar use” rules. These scientifically absurd practices countenance deeming those who (allegedly) smoke anywhere from 1 to 8 cigars monthly to be akin to nonusers for underwriting purposes!

Over 80% of insurers confessed to having “occasional user” guidelines for cigars only. No equivalent guidelines were are used for pipes, hookah or smokeless tobacco.

Aside from oral cavity neoplasms, mainly later in life, there is little mortality concern with smokeless tobacco (ST) in the manner it is mainly consumed (snus, moist snuff). This explains why 24% of insurers offer non-tobacco user rates to ST consumers; that is, provided they do not also smoke tobacco.

The only concern here is that we cannot distinguish ST consumption from smoking with a cotinine test. Therefore, an applicant with an affinity for “coffin nails” (cigarettes) could tell us he uses snuff, test cotinine-positive and get coverage on a nonuser basis.

There is a blood test for a substance called thiocyanate that distinguishes ST use from smoking with about 80% accuracy. Nevertheless, the substantial added cost makes its use impractical. Moreover, thiocyanate is probably unfamiliar to 99.9% of clinical doctors and would thus complicate matters if the insured asked us (as they often do now) to send test results to his physician!

After some interval of abstinence, we allow those who quit smoking to qualify for nonuser premium rates. Most carriers (86%) require just 12 months away from tobacco overall, with longer intervals (24 to as much as 60 months) for those seeking coverage at preferred rates.

However, because we base our assessment of tobacco use on cotinine test results, an ex-smoker will only be eligible for lower nonuser premiums by also abstaining from the use of nicotine patches, gum and spray, as well as those “electronic cigarette” contraptions.

In our analysis of the survey findings, we cited 5 major studies showing that the excess mortality associated with cigarette smoking does not “wear off” over the modest abstinence intervals of just 1 to 5 years. Indeed, these studies reveal that even those with merely 10-20 pack-years of cigarette use have substantial greater mortality than never-smokers for at least 2 decades after stopping their habit.

When confronted with these facts, those who favor current medically-illogical practices argue that formerly smoking applicants expect lower premium rates and therefore “if we do not make this concession, our competitors will.”

And they are likely correct!

Historically, cotinine screening has commenced at age 18. Now 36% of companies report screening at younger ages.

Given the association between teenage smoking and other risk-taking behaviors, this practice can easily be justified. On the other hand, we could in effect “blow the whistle” on clandestine smoker (much to the dismay of their parents!).

Over 90% of chief underwriters believe that cotinine tests done on current smokers will rarely if ever be negative.

This is a fallacy, mainly because of the rising prevalence of individuals who either smoke just a few cigarettes a day or, in many cases, do not smoke every day.

Given the cotinine level cutoffs we use for a positive test, a sizeable portion of such cases will be false-negatives (e.g., smokers with just tiny quantities of cotinine in their urine).

Another longstanding myth holds that heavy exposure to environmental tobacco smoke (passive smoking) can cause a positive result when cotinine testing is performed for insurance screening.

Despite being told many times that this is pure bollocks, 17% of chief underwriters continue to let applicants conjure up the passive smoking myth as an excuse for their positive cotinine test!

Betel (areca) nut use is common in many SE Asian countries and the habit is typically retained when they immigrate to Western countries.

When we asked survey takers how they underwrite current betel consumption 39% said they do so as if its insurability implications were equivalent to that of tobacco use. An additional 20% consider this practice compatible with nonuser status, whereas 12% wisely base their decision on how betel is consumed (making a distinction between those who use it like smokeless tobacco versus as a foodstuff).

Interestingly, the remainder (29%) did not have a policy in this regard, with many saying they have never encountered this situation!

Betel use does not cause a positive cotinine test unless consumed with tobacco (which it often is).

Habitual use is linked to increased risk of oral cancer and a 17-study literature review found that betel aficionados had 21% greater all-cause mortality than nonusers.


Alcohol Use

We asked survey takers how many drinks per day they consider the threshold for excessive drinking such that they might take adverse action on a middle age male.

The most prevalent cutoff was 4 drinks (42%) with 35% saying less and 23% countenancing 5 or more.

From a strictly medical perspective, the risk of alcohol-mediated disease is said to become significant at 60 to 80 grams of ethanol intake daily. Given that the average drink here contains roughly 12 grams of alcohol, the risk threshold is 6 or more drinks.

This is supported by a 2014 study of over 380,000 adults followed 13 years, wherein the all-cause mortality ratio did not increase significantly until at least 60 grams were imbibed daily. At that threshold the multivariate hazard ratio was 1.53 as compared to the lowest intake level (< 5 g/day).

The issue here, in terms of underwriting practices, is more about trauma-related vs. disease-mediated death, most notably due to impaired driving. Under age 40, trauma is the leading cause of death in North America and higher levels of ethanol use are associated with excess trauma deaths at all ages.

The other factor is that applicants habitually understate their true level of alcohol intake as we have seen by matching application responses to medical records from their GPs.

Most of the accident mortality related to drinking clusters in a subset of consumers qualifying as binge drinkers. We define this as consuming, on average, 5 or more drinks at one sitting (or falling, as the case may be!).

We addressed this issue in a 2013 paper, which may be downloaded, for free:

For the record, just 43% of survey respondents said they ask about bingeing on their applications. The ideal question in this context is: “on average, how many times per month do you consume 5 or more drinks on 1 drinking occasion.”

In the latest iteration of the “bible” of psychiatric diagnosis, DSM-5, alcohol abuse and alcohol dependency (alcoholism) are now combined into one entity dubbed “alcohol use disorder” (AUD). This is further subdivided as mild, moderate or severe based on the number of diagnostic criteria present.

Two-thirds of survey takers continue to distinguish between abuse vs. dependency in their guidelines. This does not pose a major problem as yet but it will as we see more cases diagnosed on the basis of DSM-5 criteria.

When asked how many years following AUD treatment and recovery they will offer cover with a loading, half of respondents reported 5 years. Most of the rest mandated 6 years or longer with just 16% wiling to consider issuing a loaded policy after a shorter interval.

Half of insurers would not offer coverage to a recovered alcohol use disorder patient if that individual consumed alcohol on any basis. The reason we asked is because there are some experts who maintain that so-called “controlled drinking” is not a significant risk factor for relapse to abuse or dependency.

Most of those who did not rule out coverage in this setting would consider each case on its merits (which is a slippery slope to say the least!).

We have access to low cost carbohydrate deficient transferrin (CDT) testing and this should be mandatory if anyone considered insuring such individuals. For the record, seventy percent (70%) of insurers use the CDT on some basis.

A positive CDT would confirm that the applicant’s intake was more than “controlled” use. In the same vein, we would insist upon a current normal GGT and further than the mean corpuscular volume (MCV) red blood cell test be consistently normal in the applicant’s medical records (which we would surely review carefully).

Actually, all of the above would usually be required on any potentially insurable applicant with a history of alcohol use disorder even if he claimed to be 100% alcohol-abstinent.

In addition to CDT we also have access, via our two insurance testing laboratories, to blood alcohol screening. If this test is positive, it means the proposed insured consumed alcohol within a relatively short time prior to the blood specimen collection.

Almost 20% of insurers now use this test, realizing as they do so that it is useless if done after midday…because of alcohol consumption with lunch, etc.

At our study group meeting, a member reported on a case of a surgeon who tested alcohol-positive on a blood specimen collected during a paramedical done at 8 in the morning. Suffice to say this is not a bloke one would be keen to encounter, as a patient that is, in a surgical theatre!

We asked if participation in Alcoholics Anonymous (AA) or a similar 12-step program would be given favorable consideration when underwriting a treated and ostensibly recovered alcoholic. Half of respondents do take this into account on all cases and 35% consider it on a case-by-case basis.

Our research reveals that the weight of evidence supports the presumption that AA participation increases the likelihood of sustained alcohol abstinence.

Nearly all U.S. life insurers offer coverage on a preferred risk basis. Indeed, virtually all that do so offer at least 3 distinct preferred classes: regular preferred, super preferred and smoker-preferred.

We asked if carriers use GGT as one of their preferred criteria, in effect meaning that at least super preferred status would be denied if the GGT were elevated.

Two out of 3 insurers that screen with GGT and offer preferred coverage said they did not use the test in this context.

Hats off to the minority!

GGT is a potent all-cause mortality marker after adjusting for alcohol intake, diabetes, metabolic syndrome and liver disease.

In a 2014 literature review at Cambridge, those in the top GGT tertile had 60% higher mortality as compared to those with the lowest third of GGT readings and mortality increased 7% for each additional 5 IU incremental rise in GGT.

Insurers notoriously ignore or understate the mortality implications of GGT. This is due in part to the lack of respect for GGT on the part of clinical physicians as well as the mistaken assumption that elevated GGT is only associated with heavy drinking.


Drug Use

Over the last few years, many insurers have changed their underwriting approach to marijuana use. Today, 53% will issue cover to recreational pot users at non-tobacco user rates. That is, of course, if they do not also smoke tobacco.

Most of the other carriers are considering this change whereas 10% are steadfast in this mistaken belief that these applicants have excess mortality risk.

For those who doubt the wisdom of liberalizing our approach to Cannabis sativa use we recommend you take 15 minutes and read this review paper:

Bottom line: there is no convincing evidence that adult marijuana use correlates with significant excess mortality risk.

Yes, there are some fatal car crashes attributed to marijuana but their incidence pales by comparison with those linked to use of a legal substance (alcohol)…and we do not charge higher premiums for alcohol use per se.

Most carriers limit the amount of marijuana use they will accept at non-tobacco user rates, whereas 24% have no such stipulations. Limits range from less than once per week to at least 3 times weekly.

None of this really matters, however, because any applicant confessing (sic) to pot use can understate his intake and we will have no clue as to whether or not he is telling the truth!

Given that in excess of 40 medical conditions respond favorably to therapeutic marijuana, it is high time (no pun) for us to accept recreational use at standard rates.

The last survey question of relevance to ALUCA members asked survey takers how they would underwrite the admitted use of various “drugs of abuse.”

The disconcerting results showed that insurers considered any use of psychedelics as equivalent in significance to methamphetamine and in some cases even heroin.

After doing considerable research on LSD, psilocybin (“magic mushrooms”), mescaline (peyote) and dimethyltriptamine (DMT), we found no evidence of excess mortality risk.

In studies ranking a wide range of psychoactive drugs in terms of adverse consequences, these substances finish near the bottom along with, in every investigation, marijuana. In fact, benzodiazepines and the so-called “Z drugs” for insomnia (e.g., zolpidem) are said to confer more risk than the aforementioned hallucinogens.

In the wake of distributing this survey to eligible recipients we did a followup review and critical analysis of the survey responses. Given that it was supported by in excess of 100 citations from the medical literature, perhaps it will have some positive impact by motivating carriers to change the obsolete and erroneous practices identified by the survey.

Hopefully you found something of interest in this review.

If you have any questions or comments, please contact me.

And if you are not as yet a subscriber to our free monthly e-magazine Hot Notes, you can receive an email link to each new issue by signing up at Hot Notes has over 5000 readers in 52 countries.


Hank George FALU

Hank George, FALU, is a longtime ALUCA member. He is self-employed as an educator and consultant, based out of his office in Milwaukee, Wisconsin. His websites are and Hank may be reached at


linked in group
Comment and discuss this article with other industry professionals in the ALUCA Linked-In group