Is consuming wine really as unhealthy as tobacco?

The answer to that question that is currently being pushed by several lobby groups is: “yes”. I am not at all convinced by this answer, for several reasons that I will discuss in this post. (I am apparently not alone in this view: Why is moderate wine consumption being villainized in 2023?).

The main issue for me is that we are dealing with public health — that is, groups of people, rather than individuals. The claim is that wine is unhealthy for people as a whole. By contrast, one of the medical centers that I occasionally deal with, has as part of its stated Vision: we improve public health — one individual at a time (Swedish: förbättrar vi folkhälsan — en individ i taget). The current official medical pressure seems to be the opposite approach, for alcohol in general (WHO shifts its alcohol narratives, and the wine industry faces new challenges).

Put simply, human health is a standard cost-benefit analysis. That is, for any given situation we explicitly lay out both the benefits and the costs, for everyone to be able to see and evaluate. We then compare those costs and benefits, to see how they balance each other. If the costs out-weigh the benefits then we would be best to change the situation; and if it is the other way around then we might leave things as they are. (As a practical example, see: Ag groups cheer court’s defense in science-based regulation of chlorpyrifos.)

Here is a simple cost-benefit analysis of the personal consumption of wine and tobacco.

Health	Benefits	Costs
Smoking tobacco	• few, except stress relief	• lung cancer causing mortality • costs for people within breathing distance
Drinking wine	• J-curve of personal health: low consumption reduces risks • social interactions (since we don't drink alone)	• J-curve of personal health: high consumption increases risks • social costs of drunkenness and alcoholism

The J-curve of personal health with regard to wine drinking, referred to in this table, is discussed in my recent post: A scientist looks at alcohol and health, and is concerned. Basically, low consumption seems to have somewhat of a health benefit for wine drinking, but increased consumption definitely increases the personal health risks. That is, it is better for your body to drink small amounts regularly rather than to binge drink (‘Wine math’ is trending on TikTok). In particular, people who tend to drink little and often have a longer life expectancy than do those who guzzle or those who abstain (Why do moderate drinkers live longer than abstainers?).

There seems to be very little basis here for directly comparing smoking and drinking in terms of health costs and benefits. Mind you, George F. Koob (director of the US National Institute on Alcohol Abuse and Alcoholism) has suggested that: “recent research finds the health risks of alcohol outweigh any benefits.” For example, one reason for the J-curve is that alcohol seems to be protective against cardiovascular disease, while at the same time increasing the risk of death by other means, such as cancers and accidents (WHO lobbies EU lawmakers against watering down alcohol cancer risk).

Nevertheless, smoking was banned in public because of a direct effect on humans even at low levels, and also on people nearby the smoker (chewing tobacco was excepted); * no such effect is know for low levels of alcohol consumption. Furthermore, given the social nature of wine drinking, the idea that banning it would have a good cost-benefit analysis is unlikely.

This has not stopped an increasing level of discussion, both ways (summarized in: Is alcohol the new tobacco?). Indeed, there have even been suggestions that the anti-alcohol lobby is taking us back in time (How much of a threat is the New Prohibitionist Movement?). These discussions are definitely worth reading. One outcome of interest is that a move towards healthier lifestyles by younger people is reducing alcohol consumption (Is Generation Z drinking less?), although apparently not in India (How do you tackle the Indian beverage alcohol market?).

Even wine label warnings should not be as dire as those for smoking. Nevertheless, some sort of label warning is probably inevitable for all alcoholic drinks, so wine-makers may need to get used to it (Concern among winemakers about potential new requirements for alcohol health warnings on bottles), just as they are now doing for other label requirements (New EU regulations for wine ingredients and nutritional labeling come into effect December 8). There are, however, potential health-warning issues for all of us (6 ways the new public health anti-alcohol warnings may affect consumers).

Associating wine and tobacco has a long history, even in the medical profession. We might, for example, remember When cigarette companies used doctors to push smoking. Even the Bollywood movies of India [Hindi cinema, based in Mumbai] associated the two: “a woman drinking wine [in Bollywood movies] was usually a vamp with a glass in one hand and a cigarette in the other, up to devious things.” So, I will end this post with two apt quotes.

First, we have George Bernard Shaw:

“A cigarette is a pinch of tobacco rolled in paper, with fire at one end and a fool at the other.”

Second, we have Robert Mondavi:

“We believe wine is the temperate, civilized, sacred, romantic mealtime beverage recommended in the Bible. It is a liquid food that has been part of civilization for 8,000 years. Wine has been praised for centuries by statesmen, scholars, poets, and philosophers. It has been used as a religious sacrament, as the primary beverage of choice for food, and as a source of pleasure and diversion. Wine is the natural beverage for every celebration:  the birth of a child, graduations, engagements, weddings, anniversaries, promotions, family gatherings, toasts between governments, and other festivities.”

---

* Note that vaping is an improvement, but people like me can still smell the tobacco.

The origin and diversification of cultivated grape-vines

I spent my professional life as a biological scientist in universities. I specialized in three areas:

• fire ecology — the effects of fire (intensity, frequency, season) on the native plants of eastern Australia
• phylogenetics — the long-term evolutionary history of plant species
• phylogenetic data-analysis methods [see my previous blog: The genealogical world of phylogenetic networks].

It may therefore seem a bit odd that these are topics that I have rarely written about in this current blog, with respect to the wine industry, although they all apply to some extent.

This is not from lack of interest in either the phylogenetics or fire ecology of grape-vines, as I do read all of the literature when it appears. It has more to do with the fact that these are complex topics, from my point of view (e.g. the wildfire effects on the wine industry in California in 2020), and they would need to be simplified considerably if I am going to write about them for the general public (in this blog).

Today, however, I am going to try to explain a recent scientific publication about the domestication of grape-vines from their wild progenitors.

The grape cultivars that we currently use in agriculture are (almost) all considered to be a single species (called Vitis vinifera; one of c. 80 species in the genus Vitis). However, their wild ancestors were grouped into a species that we call Vitis sylvestris. Sounds simple doesn’t it? However, even this is actually complicated, because some people think that there is only one species, with two subspecies (Vitis vinifera ssp. vinifera and Vitis vinifera ssp. sylvestris), rather than two separate species (Vitis vinifera and Vitis sylvestris). Don't get involved in this (technical) argument!

Concerning this topic, there was a scientific paper published earlier this year called: Dual domestications and origin of traits in grapevine evolution, published in the journal Science vol. 379 pp. 892–901. Just to give you an idea of the complexity of this particular study, there are 89 authors, from around the world. [Can you imagine trying to co-ordinate this number of people, to agree on a single publication?] The reason for this large number is that they collected 3,525 samples of wild and cultivated grape-vines (2,503 V. vinifera and 1,022 V. sylvestris), and looked in detail at the genetic makeup of each one. *

From this mass of information, a core collection of 2,448 distinctly different grape-vines (1,604 V. vinifera and 844 V. sylvestris) were studied in more detail. They did this by throwing a whole heap of complicated mathematical analyses at the data (many of which I do actually understand!), and came up with some conclusions.

They decided that there are four distinctive genetic groups within V. sylvestris, from distinct geographic regions: Western Asia, the Caucasus, Central Europe, and the Iberian Peninsula. The first pair of these groups was designated as the eastern ecotype (Syl-E) and the second pair as the western ecotype (Syl-W).

They also decided that there are six distinctive genetic groups within cultivated grape-vines (V. vinifera), generally from: Western Asian table grape-vines (CG1), Caucasian wine grape-vines (CG2), Muscat grape-vines (CG3), Balkan wine grape-vines (CG4), Iberian wine grape-vines (CG5), and Western European wine grape-vines (CG6). Note the distinction they make between table grapes and wine grapes.

The idea, then, is to look at the evolutionary history in terms the splitting and merging of these various genetic groups (4+6 = 10 groups) through time, and the time at which these events occurred — this is called a phylogeny. Their phylogeny is shown in the figure immediately above.

I stared at this picture for quite some time, as an expert, and eventually worked out all of what it is saying. [As we scientists like to say: I extracted all of the pertinent information.] However, I challenge you to do the same thing, yourselves; any time you have a few spare hours, have a go!

From your perspective, in trying to understand this, it seems like it will be easier to look at one of their other figures, to work out what the authors concluded about long-term grape-vine history. This is the map shown immediately above, which illustrates the geographic dispersal of the cultivated grape-vine groups.

Their actual description of the pattern [with my extra notes] is:

“In the Pleistocene [epoch, 200,000 to 400,000 years ago], harsh climate drove the separation of wild grape ecotypes caused by continuous habitat fragmentation. Then, domestication occurred concurrently about 11,000 years ago in Western Asia and the Caucasus to yield table [labeled CG1] and wine [labeled CG2] grapevines. The Western Asia domesticates dispersed into Europe with early farmers, introgressed [introgression = the transfer of genetic material from one species into the genome of another] with ancient wild western ecotypes, and subsequently diversified along human migration trails into muscat [CG3] and unique western wine grape [CG4—CG6] ancestries by the late Neolithic [7,000 to 8,000 years ago].”

So, there you have it — you will find that this description does actually match the pattern of arrows shown in the map, showing the origin and dispersal of the newly domesticated grape types. **

There is, of course, a lot more information in the paper than this. However, this obvious complexity is why I rarely write about such things in this blog.

---

* They didn’t collect new samples, but instead got them “from a dozen Eurasian germplasm and private collections”, plus previously obtained genomic data.

** Note the irony of the geographical location where the diversification starts: the Middle East. This is now the location of the biggest alcohol-related conflict that we have; see Tom Wark’s recent review of Wine and the Clash of Civilizations.

A scientist looks at alcohol and health, and is concerned

As a scientist (in my professional life), I am becoming increasingly concerned about the nature of the current attacks by health authorities on alcohol consumption. Indeed, I am personally getting to be thoroughly sick of it.

There have been a number of sensible commentaries in the wine web universe, and I will summarize (and quote)  some of them here. Much of the rest of it leaves a lot to be desired, to say the least. Here, I offer a perspective on this.

As a professional scientist, I realized during my working life that the most important thing is to protect yourself from those scientists (fortunately few) who do not themselves have a high standard. After that, things should flow smoothly regarding the accumulation of useful information. The same thing seems to be true in the wine industry.

I noted in last week’s post (It is difficult to study the effect on children of parental drinking) that many, if not most, medical experiments are simply “correlative surveys”, where we quantify two (or more) things in order to assess the degree to which they are correlated. This is not scientifically the best approach, as it does not demonstrate any causal relationships between the things measured.

I have also discussed this topic previously: Wine and health — why is there so much argument, pro and con? I have also noted that: The alleged health benefits of wine depend on who funded the research. In particular, it is important to note that one needs a big study (lots of people studied) for correlations to work, and yet the bigger is the study then the harder it is to look at the many confounding factors that will be present, which can obscure what we wish to be studying (discussed in Debate over alcohol's health effects grows).

Anyway, many of these medical claims are influencing public policy regarding wine consumption. There was a recent meeting in the wine industry to discuss this: Wine and health: Challenging the ‘No safe level’ claims. This topic was summarized as:

When wine executives gathered in Toledo, Spain, their mood could be described as “worried”. They had come to the Lifestyle, Diet, Wine & Health Congress to hear what science and medicine have to say about wine and health.
   The impetus for the gathering, organised by the Wine Information Council and FINVIN, is a growing concern about public health messaging around alcohol, which has gone from “drink in moderation” to “there is no safe level of alcohol”.
   This messaging is already having an impact. A recent Gallup study revealed that 39% of Americans now see consuming alcohol in moderation as unhealthy.
   Things are about to get even tougher for the wine sector, as the World Health Organisation (WHO) wants to see more alcohol taxes, more advertising restrictions, and even tighter restrictions on the availability of alcohol — recommendations that are already informing the EU’s policy positions.

The argument is, of course, that the health risks of alcohol outweigh any benefits. There are clearly two edges to this sword: the benefits, and the risks. Life, for all of us on a personal level, is a balance between these two things. However, whether there are social benefits and risks is actually a quite separate issue. The latter is an accumulation of individual benefits and risks.

How pervasive are the social risks? It has recently been noted (The heffalump and the alcohol czar):

In 2018, George Koob’s colleagues at the Center for Disease Control and Prevention (CDC) found that a little over 5 percent of American adults engaged in heavy drinking in the past year, 15.5 percent engaged in moderate drinking and 45.7 percent in light drinking; 33.7 percent did not consume alcohol.

That does not seem to be too bad, to me.

The above picture is from: Health benefits of moderate alcohol consumption and how much you should drink: What the experts say. It shows the standard relationship that has been repeatedly shown for the effects of alcohol on individual human health, usually called a J-curve: zero consumption is associated with low health risk (there is always some risk in life!), low alcohol consumption is associated with even lower health risk, and then increased consumption (to moderate and high levels) strongly increases the risks.

In other words, a little bit of wine is actually better for you than no wine at all. This has been known for a long time, although it is repeatedly disputed. The disputes come from the fact that it is based on correlative experiments, which, as I discussed above, have limitations.

It is also important to emphasize here that there is not a single J-curve that fits the entire human species. The following graph is reproduced from Alcohol awareness week: The J-Curve, which summarizes the situation very well — many situations in life have different J-curves. Not only do different people have different J-curves, different groups of people do, as well.

Therefore, The J-curve isn’t one-size-fits-all: “The J-curve is everywhere, but French and Italian people (for example] can drink more safely because the beneficial effect remains longer than in Germany or the UK or in Sweden [and Finland, Norway].”

There is one other important thing that seems to often be overlooked, when thinking about the J-curve — people usually do not drink alone, unless they have some sort of alcohol problem. This point was emphasized in a recent formal study, discussed here: Wine’s biggest health benefit might be drinking with friends. To quote:

A new study suggests that drinking wine with friends offers more health benefits than drinking alone ... According to their study, published in the journal The Gerontologist, they questioned whether published studies on the benefits of moderate alcohol consumption for the elderly population could be attributed to the lifestyle adopted by these moderate drinkers rather than alcohol itself as a substance. Their theory was that moderate drinking correlated to how often respondents socialized and that this increase in social activities is what produced the positive health outcomes.

I cannot emphasize this social point enough. When I was young, I went to the pub on Friday nights with my friends. We played snooker or pool, and had a few beers. The beers were not the important part — the socializing was the important part. I now share a bottle of wine with my wife, and often with our shared friends. You get the idea: I do not drink alone. If I did, then I would probably consider myself to have a drinking problem.

Indeed, the multiple J-curves, mentioned above, may actually be different social effects, as well. After all, the effect of drinking 1 glass of wine per day is probably different from that of drinking 7 glasses (1 bottle) on the weekend only. These two scenarios would produce different J-curves for the people involved.

I should end with something positive. Jeff Siegel recently noted (Down to zero: Ignoring neo-Prohibitionists could prove dangerous to the wine industry):

For much of the past decade, the wine business has watched — with seemingly little interest — as health groups and national regulators took strong measures to reduce alcohol consumption around the world. Their assertion, backed up by a variety of studies, was that any drinking, even in moderation, was deadly. Those studies included a report earlier this year by the World Health Organization, as well as Irish regulators who linked drinking with cancer, and a Canadian proposal to cut safe alcohol limits from two drinks per day to two per week.

He then collated some pertinent suggestions for useful actions by people in the wine industry, to address the current attacks from some parts of the medical and political world:

• Call out the bad science
• Reinforce the good science
• Offer the mainstream media another point of view
• Take neo-Prohibitionism seriously
• Sharpen wine’s marketing focus
• Support transparency in labeling and ingredients
• Position wine as the choice of moderation

It is difficult to study the effect on children of parental drinking

Everyone in the wine industry should care about the effects on children of parents’ wine consumption. For example, over-indulgence by parents should not lead to any current or future consumption problems for their children. Alternatively, wine enjoyment by parents might usefully be passed on to their children.

However, it is not easy to research this topic in any formal way. After all, we cannot do the sort of standard scientific experiment where we manipulate the world to see what effect our changes have. So, we are stuck with what are called “correlative surveys”, where we quantify both parental and child drinking, to assess the degree to which they are correlated.

Sadly, such experiments are not necessarily easy, as I discuss here.

A specific example that I will discuss here was published as: Associations between parental drinking and alcohol use among their adolescent children: Findings from a national survey of United States parent-child dyads (Michele K. Bohm and Marissa B. Esser. Journal of Adolescent Health 73: 961–964, 2023). This study examined associations between parent and child drinking using recent United States national survey data:

We analyzed responses of 740 parent—child dyads from 2020 SummerStyles and YouthStyles surveys. Parents and their adolescent children answered questions about past 30-day alcohol use ... An adult answered the first survey and, if applicable, one of their children ages 12—17 answered the same questions in the second survey.

The results are straightforward, as presented in the publication:

We estimated prevalence of adolescent drinking and explored differences by socio-demographics ... Overall, 6.6% of adolescents drank alcohol, with no significant differences by socio-demographics. Adolescents whose parents drank frequently (≥5 days/month), or binge drank, had significantly higher odds of drinking than adolescents whose parents did not drink or did not binge drink, respectively.

Unfortunately, the problem with this sort of descriptive experimental approach is that the researchers do not have any control over the sample of people surveyed. If there is any bias in the sampling of the people, then there is nothing that can be done about it. In this case, the characteristics of the sampled people are shown in this table.

A look through these numbers reveals several characteristics that are likely not representative of the overall US population:

• two-thirds of the parents are female
• three-quarters of the parents have some tertiary education
• more than half of the parents earn a lot of money
• half of the parents do not drink.

Any of these is a potential bias in the experiment, meaning that we do not know how the sample relates to the general population of people. That is, I assume that the data as presented are quite correct: but I cannot deduce what this tells me about US people in general. We need unbiased surveys if we are to be able to draw unbiased conclusions.

This does not in any way down-play the potential effects of parental drinking behavior on their children. An earlier publication (Michael Windle. Effect of parental drinking on adolescents. Life­-Stage Issues 20: 181–184, 1996) correctly noted:

It is evident that parental alcohol abuse may have a range of potential adverse effects on adolescents. Problem drinking by parents may influence role­—modeling behaviors, parent­ing skills, and marital and family relations, all of which may contribute to a host of problematic outcomes for adolescents.

Nor, of course, does this preclude any attempt to do better studies. However, this is not easy in this case. For example, a survey paper (Ingeborg Rossow, Patrick Keating, Lambert Felix & Jim McCambridge. Does parental drinking influence children’s drinking? A systematic review of prospective cohort studies. Addiction 111: 204–217, 2015) involving a search for what is often considered to be the best sort of experiment in this situation (called: prospective cohort studies) found that:

Four of the 21 included studies filled several, but not all, criteria and were assessed to have some capacity for causal inference. These four studies found some evidence that parental drinking predicted drinking behaviour in adolescent offspring. The remaining 17 studies had little or no such capacity.

Mind you, children to not necessarily copy their parents’ behavior. For example, it has recently been pointed out that the current generation of young adults do not seem to be all that keen on consuming alcohol (How is Gen Z approaching beverage alcohol?):

Younger, legal-drinking aged Gen Z consumers increasingly enjoy a very different relationship with alcohol versus older age cohorts, exhibiting rising levels of abstention, moderation, experimenting with new categories, and turning away from traditional, high-volume categories, ... although there are wide variations between individual countries.

So, it seems that the youngest generation of adults have not copied their parents’ (Millennials) drinking habits. Indeed, this blog may be coming progressively redundant. I hope not!