Should you drink your coffee filtered?

Coffee is one of the most widely consumed beverages in the world. Arguably a key reason for this is that coffee has psychoactive properties that we may be hardwired to value, even if subconsciously. For example, it increases alertness; possibly a fitness-enhancing effect in our evolutionary past. Here the term “fitness” in “fitness-enhancing effect” means “reproductive success”, and does not mean having great athletic ability or having shredded abs.

The two most common sources of coffee beans, which are roasted and ground prior to brewing, are the widely favored Coffea arabica, and the "robusta" form Coffea canephora. The arabica form accounts for 80 percent or so of world consumption. The graph below, from a study by Bonita and colleagues (), shows the per capita consumption of coffee in various countries. As you can see, Scandinavian countries are big consumers.



Most people probably drink filtered coffee. However, there are many unfiltered coffee preparation methods that are also widely used. Greek coffee, Turkish coffee, coffee prepared with a French press, and “cowboy coffee” are all unfiltered.

In the photo below (from: Goldenstate.wordpress.com), illustrating cowboy coffee, note that the coffee pot is placed near but not over the fire.



What is “cowboy coffee”? This method of preparation has many variations. A simple one involves mixing ground coffee with hot water, and then keeping the coffee simmering on very low fire for a while. It is called cowboy coffee due to its association with coffee drank by cowboys around a campfire.

After brewed, coffee tends to rise and spill out of the pot if heated at a high temperature. To avoid this, one should turn off the fire just prior to the coffee boiling, heat the coffee in a pot on very low fire, or heat the coffee by placing the pot near but not too close to a campfire. The same is generally true for tea.

With cowboy coffee you need significantly less coffee per measure of water, and the coffee ends up with a stronger flavor – if prepared properly. You also keep two key oily components of the coffee, namely the diterpenes known as kahweol and cafestol; its polyphenols, most notably chlorogenic acid; and some of the coffee particles.

Both kahweol and cafestol seem to be associated with reduction in certain types of cancer in humans, and show strong anti-cancer effects in rats (). The same seems to be generally true for chlorogenic acid (). The coffee particles, if ingested, would probably be treated as indigestible fiber and promote colon health. This is usually the fate of indigestible and partially digestible plant matter.

Why is filtered coffee often recommended? Well, unfiltered coffee is believed to promote heart disease. But that is not primarily due to any strong association having been found between unfiltered coffee consumption and heart disease. In fact, the absence of evidence in favor of this hypothesis in long-term studies is rather conspicuous ().

The belief that unfiltered coffee can promote heart disease is due to evidence showing that consumption of 4 cups per day of unfiltered coffee raises total cholesterol by up to 10 mg/dl ().

Only diehard proponents of the lipid hypothesis would look at total cholesterol increase as a marker of heart disease, in part because total cholesterol may increase due to an increase in HDL cholesterol – a much more reliable marker, but of protection against heart disease, particularly within certain ranges. And yes, unfiltered coffee consumption is associated with an increase in HDL cholesterol ().

Moreover, some of the metabolites of caffeine, 1-methyxanthine and 1-methyluric acid, appear to help prevent LDL oxidation; caffeine metabolites also seem to have potent anti-inflammatory properties ().

Some research provides evidence of the importance of moderation in coffee consumption as an important factor in its relationship with health. In this respect, coffee is like almost anything that can be ingested, including water – the dose makes the poison. In a study of 40,000 post-menopausal women in the US reviewed by Bonita and colleagues (), the hazard ratio of death attributed to heart disease was 0.76 for consumption of 1–3 cups/day, 0.81 for 4–5 cups/day, and 0.87 for ≥6 cups/day. Interestingly, the same study reported that the hazard ratio for death from other inflammatory diseases was 0.72 for consumption of 1–3 cups/day, 0.67 for 4–5 cups/day, and 0.68 for ≥6 cups/day.

Frequently you hear about the possible connection between coffee consumption and gastritis. The most widely cited study I could find that looked into this link found no association between coffee consumption and reflux-associated gastritis ().

By the way, if you have gastritis, you should consider getting tested for Helicobacter pylori (), especially if you like eating raw fish.

Stress and coffee consumption may have similar effects in those who test positive for Helicobacter pylori (see, e.g., ). In those individuals, past research has found a link between: (a) stress, coffee consumption, and other purported “stomach irritants”; and (b) exacerbation of gastritis symptoms, stomach ulcers, and stomach cancer.

This discussion on gastritis is largely unrelated to the issue of drinking unfiltered coffee. It is unclear based on the past studies that I reviewed whether coffee filtration has anything to do with any possible connection between coffee consumption and exacerbation of gastritis symptoms caused by other factors.

As a side note, it is important to keep in mind that the acidity of coffee is nowhere near the acidic of gastric acid, which the stomach is uniquely designed to handle.

I may be wrong, but from what I can see, if you drink coffee regularly and it causes no problems for you, drinking unfiltered coffee is not a bad idea at all.

Sudden cholesterol increase? It may be psychological

There are many published studies with evidence that cholesterol levels are positively associated with heart disease. In multivariate analyses the effects are usually small, but they are still there. On the other hand, there is also plenty of evidence that cholesterol is beneficial in terms of health. Here of course I am referring to the health of humans, not of the many parasites that benefit from disease.

For example, there is evidence () that cholesterol levels are negatively associated with mortality (i.e., higher cholesterol leading to lower mortality), and are positively associated with vitamin D production from skin exposure to sunlight ().

Most of the debris accumulated in atheromas are made up of macrophages, which are specialized cells that “eat” cell debris (ironically) and some pathogens. The drug market is still hot for cholesterol-lowering drugs, often presented in TV and Internet ads as effective tools to prevent formation of atheromas.

But what about macrophages? What about calcium, another big component of atheromas? If drugs were to target macrophages for atheroma prevention, drug users may experience major muscle wasting and problems with adaptive immunity, as macrophages play a key role in muscle repair and antibody formation. If drugs were to target calcium, users may experience osteoporosis.

So cholesterol is the target, because there is a “link” between cholesterol and atheroma formation. There is also a link between the number of house fires in a city and the amount of firefighting activity in the city, but we don’t see mayors announcing initiatives to reduce the number of firefighters in their cities to prevent house fires.

When we talk about variations in cholesterol, we usually mean variations in cholesterol carried by LDL particles. That is because LDL cholesterol seems to be very “sensitive” to a number of factors, including diet and disease, presenting quite a lot of sudden variation in response to changes in those factors.

LDL particles seem to be intimately involved with disease, but do not be so quick to conclude that they cause disease. Something so widespread and with so many functions in the human body could not be primarily an agent of disease that needs to be countered with statins. That makes no sense.

Looking at the totally of evidence linking cholesterol with health, it seems that cholesterol is extremely important for the human body, particularly when it is under attack. So the increases in LDL cholesterol associated with various diseases, notably heart disease, may not be because cholesterol is causing disease, but rather because cholesterol is being used to cope with disease.

LDL particles, and their content (including cholesterol), may be used by the body to cope with conditions that themselves cause heart disease, and end up being blamed in the process. The lipid hypothesis may be a classic case of reverse causation. A case in point is that of cholesterol responses to stress, particularly mental stress.

Grundy and Griffin () studied the effects of academic final examinations on serum cholesterol levels in 2 groups of medical students in the winter and spring semesters (see table below). During control periods, average cholesterol levels in the two groups were approximately 213 and 216 mg/dl. During the final examination periods, average cholesterol levels were 248 and 240 mg/dl. These measures were for winter and spring, respectively.



One could say that even the bigger increase from 213 to 248 is not that impressive in percentage terms, approximately 16 percent. However, HDL cholesterol does not go up significantly response to sustained (e.g., multi-day) stress, it actually goes down, so the increases reported can be safely assumed to be chiefly due to LDL cholesterol. For most people, LDL particles are the main carriers of cholesterol in the human body. Thus, in percentage terms, the increases in LDL cholesterol are about twice those reported for total cholesterol.

A 32-percent increase (16 x 2) in LDL cholesterol would not go unnoticed today. If one’s LDL cholesterol were to be normally 140 mg/dl, it would jump to 185 mg/dl with a 32-percent increase. It looks like the standard deviations were more than 30 in the study. (This is based on the standard errors reported, and assuming that the standard deviation equals the standard error multiplied by the square root of the sample size.) So we can guess that several people might go from 140 to 215 or more (this is LDL cholesterol, in mg/dl) in response to the stress from exams.

And the effects above were observed with young medical students, in response to the stress from exams. What about a middle-aged man or woman trying to cope with chronic mental stress for months or years, due to losing his or her job, while still having to provide for a family? Or someone who has just been promoted, and finds himself or herself overwhelmed with the new responsibilities?

Keep in mind that sustained dieting can be a major stressor for some people, particular when one gets to that point in the dieting process where he or she gets regularly into negative nitrogen balance (muscle loss). So you may have heard from people saying that, after months or years of successful dieting, their cholesterol levels are inexplicably going up. Well, this post provides one of many possible explanations for that.

The finding that cholesterol goes up with stress has been replicated many times. It has been known for a long time, with studies dating back to the 1950s. Wertlake and colleagues () observed an increase in average cholesterol levels from 214 to 238 (in mg/dl); also among medical students, in response to the mental and emotional stress of an examination week. A similar study to the one above.

Those enamored with the idea of standing up the whole day, thinking that this will make them healthy, should know that performing cognitively demanding tasks while standing up is a known stressor. It is often used in research where stress must be induced to create an experimental condition. Muldoon and colleagues () found that people performing a mental task while standing experienced an increase in serum cholesterol of approximately 22 points (in mg/dl).

What we are not adapted for is sitting down for long hours in very comfortable furniture (, ). But our anatomy clearly suggests adaptations for sitting down, particularly when engaging in activities that resemble tool-making, a hallmark of the human species. Among modern hunter-gatherers, tool-making is part of daily life, and typically it is much easier to accomplish sitting down than standing up.

Modern urbanites could be seen as engaging in activities that resemble tool-making when they produce things at work for internal or external customers, whether those things are tangible or intangible.

So, stress is associated with cholesterol levels, and particularly with LDL cholesterol levels. Diehard lipid hypothesis proponents may argue that this is how stress is associated with heart disease: stress increases cholesterol which increases heart disease. Others may argue that one of the reasons why LDL cholesterol levels are sometimes found to be associated with heart disease-related conditions, such as chronic stress, and other health conditions is that the body is using LDL cholesterol to cope with those conditions.

Specifically regarding mental stress, a third argument has been put forth by Patterson and colleagues, who claimed that stress-mediated variations in blood lipid concentrations are a secondary result of decreased plasma volume. The cause, in their interpretation, was unspecified – “vascular fluid shifts”. However, when you look at the numbers reported in their study, you still see a marked increase in LDL cholesterol, even controlling for plasma volume. And this is all in response to “10 minutes of mental arithmetic with harassment” ().

I tend to think that the view that cholesterol increases with stress because cholesterol is used by the body to cope with stress is the closest to the truth. Among other things, stress increases the body’s overall protein demand, and cholesterol is used in the synthesis of many proteins. This includes proteins used for signaling, also known as hormones.

Cholesterol also seems to be a diet marker, tending to go up in high fat diets. This is easier to explain. High fat diets increase the demand for bile production, as bile is used in the digestion of fat. Most of the cholesterol produced by the human body is used to make bile.

The anatomy of a VAP test report

The vertical auto profile (VAP) test is an enhanced lipid profile test. It has been proposed, chiefly by the company Atherotech (), as a more complete test that relies on direct measurement of previously calculated lipid measures. The VAP test is particularly known for providing direct measurements of LDL cholesterol, instead of calculating them through equations ().

At the time of this writing, a typical VAP test report would provide direct measures of the cholesterol content of LDL, Lp(a), IDL, HDL, and VLDL particles. It would also provide additional measures referred to as secondary risk factors, notably particle density patterns and apolipoprotein concentrations. Finally, it would provide a customized risk summary and some basic recommendations for treatment. Below is the top part of a typical VAP test report (from Atherotech), showing measures of the cholesterol content of various particles. LDL cholesterol is combined for four particle subtypes, the small-dense subtypes 4 and 3, and the large-buoyant subtypes 2 and 1. A breakdown by LDL particle subtype is provided later in the VAP report.

In the table above, HDL cholesterol is categorized in two subtypes, the small-dense subtype 2, and the large-buoyant subtype 3. Interestingly, most of the HDL cholesterol in the table is supposedly of the least protective subtype, which seems to be a common finding in the general population. VLDL cholesterol is categorized in a similar way. IDL stands for intermediate-density lipoprotein; this is essentially a VLDL particle that has given off some of its content, particularly its triglyceride (or fat) cargo, but still remains in circulation.

Lp(a) is a special subtype of the LDL particle that is purported to be associated with markedly atherogenic factors. Mainstream medicine generally considers Lp(a) particles themselves to be atherogenic, which is highly debatable. Among other things, cardiovascular disease (CVD) risk and Lp(a) concentration follow a J-curve pattern, and Lp(a)’s range of variation in humans is very large. A blog post by Peter (Hyperlipid) has a figure right at the top that illustrates the former J-curve assertion (). The latter fact, related to range of variation, generally leads to a rather wide normal distribution of Lp(a) concentrations in most populations; meaning that a large number of individuals tend to fall outside Lp(a)’s optimal range and still have a low risk of developing CVD.

Below is the middle part of a typical VAP report, showing secondary risk factors, such as particle density patterns and apolipoprotein concentrations. LDL particle pattern A is considered to be the most protective, supposedly because large-buoyant LDL particles are less likely to penetrate the endothelial gaps, which are about 25 nm in diameter. Apolipoproteins are proteins that bind to fats for their transport in lipoproteins, to be used by various tissues for energy; free fatty acids also need to bind to proteins, notably albumin, to be transported to tissues for use as energy. Redundant particles and processes are everywhere in the human body!

Below is the bottom part of a typical VAP report, providing a risk summary and some basic recommendations. One of the recommendations is “to lower” the LDL target from 130mg/dL to 100mg/dL due to the presence of the checked emerging risk factors on the right, under “Considerations”. What that usually means in practice is a recommendation to take drugs, especially statins, to reduce LDL cholesterol levels. A recent post here and the discussion under it suggest that this would be a highly questionable recommendation in the vast majority of cases ().

What do I think about VAP tests? I think that they are useful in that they provide a lot more information about one’s lipids than standard lipid profiles, and more information is better than less. On the other hand, I think that people should be very careful about what they do with that information. There are even more direct tests that I would recommend before a decision to take drugs is made (, ), if that decision is ever made at all.

Familial hypercholesteromia: Why rely on cholesterol levels when more direct measures are available?

There are two forms of familial hypercholesteromia (FH), namely heterozygous and homozygous FH. In heterozygous FH only one copy of the gene that causes it is present, inherited either from the father or the mother. In homozygous FH, which is the most lethal form, two copies of the gene are present. FH is associated with early-onset cardiovascular disease (CVD).

Homozygous FH may happen if both the father and mother have heterozygous or homozygous FH. If both the father and mother have heterozygous FH, the likelihood that at least one in four children will have homozygous FH will be high. If both parents have homozygous FH the likelihood that all children will have homozygous FH will be high.

In fact, in the latter case, homozygous FH in the children is almost certain. One case in which it won’t occur is if the combining FH gene from the father or mother mutates into a non-FH gene before it is used in the assembly of the genome of the child. A gene mutation in a specific locus, only for the father or mother, is an unlikely event, and would lead to heterozygous FH. Two gene mutations at once in the same locus, for the father and mother, is a very unlikely event.

By the way, despite what many are led to believe based on fictional characters in movies and series like the X-Men and Hulk, mutations in functional genes usually lead to harmful traits. In our evolutionary past, those traits would have been largely removed from the gene pool by selection, making them rare or nonexistent in modern humans. Today we have modern medicine; a double-edged sword.

Mutations leading to super-human traits are very, very unlikely. The myostatin gene, for example, suppresses muscle growth. And yet the mutations that lead to little or no secretion of the related myostatin protein are very uncommon. Obviously they have not been favored by selection, even though their holders are very muscular – e.g., Germany’s “Incredible Hulky” ().

Okay, back to FH. Xanthelasmas are relatively common among those who suffer from FH (see photo below, from Globalskinatlas.com). They are skin deposits of cholesterol, have a genetic basis, and are NOT always associated with FH. This is important – several people have xanthelasmas but not FH.

FH is a fairly rare disease, even in its heterozygous form, with an overall incidence of approximately 0.2 percent. That is, about 1 in 500 people in the general population will have it. Genetically related groups will see a much higher or lower rate of incidence, as the disease is strongly influenced by a genetic mutation. This genetic mutation is apparently in the LDL receptor gene, located on the short arm of chromosome 19.

The table below, from a study by Miltiadous and colleagues (), paints a broad picture of the differences one would typically see between heterozygous FH sufferers and non-FH controls.

The main difference is in total cholesterol and in the relatively large contribution of LDL to total cholesterol. A large difference is also seen in Apolipoprotein B (indicated as "Apo B"), which acts as a LDL transporter (not to be confused with a LDL receptor). The LDL cholesterol shown on the table is calculated through the Friedewald equation, which is notoriously imprecise at low triglyceride levels ().

Looking at the total cholesterol row on the table, and assuming that the numbers after the plus/minus signs are standard deviations, we can conclude that: (a) a little more than two-thirds of the heterozygous FH sufferers had total cholesterol levels falling in between 280 and 446; and (b) a little more than two-thirds of the non-FH controls had total cholesterol levels falling in between 135 and 225.

Keep in mind that about 13.5 percent {calculated as: (95-68)/2} of the non-FH controls had total cholesterol levels between 225 and 270. This is a nontrivial percentage; i.e., these may be a minority but are not rare individuals. Heterozygous FH sufferers are rare, at 0.2 percent of the general population. Moreover, about 2 percent of the non-FH controls had non-pathological total cholesterol levels between 270 and 315. That is not so rare either, amounting to an “incidence” 10 times higher than heterozygous FH.

What would happen if people with heterozygous FH were to replace refined carbohydrates and sugars with saturated fat and cholesterol in their diets? Very likely their already high total cholesterol would go up higher, in part because their HDL cholesterol would go up (). Still, how could they be sure that CVD progression would accelerate if they did that?

According to some studies, the higher HDL cholesterol would either be generally protective or associated with protective factors, even among those with FH (). One of those protective factors may be a more nutrient-dense diet, as many foods rich in cholesterol are very nutrient-dense – e.g., eggs, organ meats, and seafood.

This brings me to my main point in this post. It is mainstream practice to diagnose people with FH based on total and/or LDL cholesterol levels. But the main problem with FH is that it leads to early onset of CVD, which can be measured more directly through simple tests, such as intima-media thickness and related ultrasound plaque tests (). These are noninvasive tests, done in 5 minutes or so, and often covered by insurance.

Even if simple direct tests are not perfect, it seems utterly nonsensical to rely on cholesterol measures to diagnose and treat FH, given the possible overlap between pathological and non-pathological high total cholesterol levels.

The China Study II: Cholesterol seems to protect against cardiovascular disease

First of all, many thanks are due to Dr. Campbell and his collaborators for collecting and compiling the data used in this analysis. This data is from this site, created by those researchers to disseminate the data from a study often referred to as the “China Study II”. It has already been analyzed by other bloggers. Notable analyses have been conducted by Ricardo at Canibais e Reis, Stan at Heretic, and Denise at Raw Food SOS.

The analyses in this post differ from those other analyses in various aspects. One of them is that data for males and females were used separately for each county, instead of the totals per county. Only two data points per county were used (for males and females). This increased the sample size of the dataset without artificially reducing variance (for more details, see “Notes” at the end of the post), which is desirable since the dataset is relatively small. This also allowed for the test of commonsense assumptions (e.g., the protective effects of being female), which is always a good idea in a complex analysis because violation of commonsense assumption may suggest data collection or analysis error. On the other hand, it required the inclusion of a sex variable as a control variable in the analysis, which is no big deal.

The analysis was conducted using WarpPLS. Below is the model with the main results of the analysis. (Click on it to enlarge. Use the "CRTL" and "+" keys to zoom in, and CRTL" and "-" to zoom out.) The arrows explore associations between variables, which are shown within ovals. The meaning of each variable is the following: SexM1F2 = sex, with 1 assigned to males and 2 to females; HDLCHOL = HDL cholesterol; TOTCHOL = total cholesterol; MSCHIST = mortality from schistosomiasis infection; and MVASC = mortality from all cardiovascular diseases.

The variables to the left of MVASC are the main predictors of interest in the model – HDLCHOL and TOTCHOL. The ones to the right are control variables – SexM1F2 and MSCHIST. The path coefficients (indicated as beta coefficients) reflect the strength of the relationships. A negative beta means that the relationship is negative; i.e., an increase in a variable is associated with a decrease in the variable that it points to. The P values indicate the statistical significance of the relationship; a P lower than 0.05 generally means a significant relationship (95 percent or higher likelihood that the relationship is “real”).

In summary, this is what the model above is telling us:

- As HDL cholesterol increases, total cholesterol increases significantly (beta=0.48; P<0.01). This is to be expected, as HDL is a main component of total cholesterol, together with VLDL and LDL cholesterol.

- As total cholesterol increases, mortality from all cardiovascular diseases decreases significantly (beta=-0.25; P<0.01). This is to be expected if we assume that total cholesterol is in part an intervening variable between HDL cholesterol and mortality from all cardiovascular diseases. This assumption can be tested through a separate model (more below). Also, there is more to this story, as noted below.

- The effect of HDL cholesterol on mortality from all cardiovascular diseases is insignificant when we control for the effect of total cholesterol (beta=-0.08; P=0.26). This suggests that HDL’s protective role is subsumed by the variable total cholesterol, and also that it is possible that there is something else associated with total cholesterol that makes it protective. Otherwise the effect of total cholesterol might have been insignificant, and the effect of HDL cholesterol significant (the reverse of what we see here).

- Being female is significantly associated with a reduction in mortality from all cardiovascular diseases (beta=-0.16; P=0.01). This is to be expected. In other words, men are women with a few design flaws. (This situation reverses itself a bit after menopause.)

- Mortality from schistosomiasis infection is significantly and inversely associated with mortality from all cardiovascular diseases (beta=-0.28; P<0.01). This is probably due to those dying from schistosomiasis infection not being entered in the dataset as dying from cardiovascular diseases, and vice-versa.

Two other main components of total cholesterol, in addition to HDL cholesterol, are VLDL and LDL cholesterol. These are carried in particles, known as lipoproteins. VLDL cholesterol is usually represented as a fraction of triglycerides in cholesterol equations (e.g., the Friedewald and Iranian equations). It usually correlates inversely with HDL; that is, as HDL cholesterol increases, usually VLDL cholesterol decreases. Given this and the associations discussed above, it seems that LDL cholesterol is a good candidate for the possible “something else associated with total cholesterol that makes it protective”. But waidaminet! Is it possible that the demon particle, the LDL, serves any purpose other than giving us heart attacks?

The graph below shows the shape of the association between total cholesterol (TOTCHOL) and mortality from all cardiovascular diseases (MVASC). The values are provided in standardized format; e.g., 0 is the average, 1 is one standard deviation above the mean, and so on. The curve is the best-fitting S curve obtained by the software (an S curve is a slightly more complex curve than a U curve).

The graph below shows some of the data in unstandardized format, and organized differently. The data is grouped here in ranges of total cholesterol, which are shown on the horizontal axis. The lowest and highest ranges in the dataset are shown, to highlight the magnitude of the apparently protective effect. Here the two variables used to calculate mortality from all cardiovascular diseases (MVASC; see “Notes” at the end of this post) were added. Clearly the lowest mortality from all cardiovascular diseases is in the highest total cholesterol range, 172.5 to 180; and the highest mortality in the lowest total cholesterol range, 120 to 127.5. The difference is quite large; the mortality in the lowest range is approximately 3.3 times higher than in the highest.

The shape of the S-curve graph above suggests that there are other variables that are confounding the results a bit. Mortality from all cardiovascular diseases does seem to generally go down with increases in total cholesterol, but the smooth inflection point at the middle of the S-curve graph suggests a more complex variation pattern that may be influenced by other variables (e.g., smoking, dietary patterns, or even schistosomiasis infection; see “Notes” at the end of this post).

As mentioned before, total cholesterol is strongly influenced by HDL cholesterol, so below is the model with only HDL cholesterol (HDLCHOL) pointing at mortality from all cardiovascular diseases (MVASC), and the control variable sex (SexM1F2).

The graph above confirms the assumption that HDL’s protective role is subsumed by the variable total cholesterol. When the variable total cholesterol is removed from the model, as it was done above, the protective effect of HDL cholesterol becomes significant (beta=-0.27; P<0.01). The control variable sex (SexM1F2) was retained even in this targeted HDL effect model because of the expected confounding effect of sex; females generally tend to have higher HDL cholesterol and less cardiovascular disease than males.

Below, in the “Notes” section (after the “Reference”) are several notes, some of which are quite technical. Providing them separately hopefully has made the discussion above a bit easier to follow. The notes also point at some limitations of the analysis. This data needs to be analyzed from different angles, using multiple models, so that firmer conclusions can be reached. Still, the overall picture that seems to be emerging is at odds with previous beliefs based on the same dataset.

What could be increasing the apparently protective HDL and total cholesterol in this dataset? High consumption of animal foods, particularly foods rich in saturated fat and cholesterol, are strong candidates. Low consumption of vegetable oils rich in linoleic acid, and of foods rich in refined carbohydrates, are also good candidates. Maybe it is a combination of these.

We need more analyses!

Reference:

Kock, N. (2010). WarpPLS 1.0 User Manual. Laredo, Texas: ScriptWarp Systems.


Notes:

- The path coefficients (indicated as beta coefficients) reflect the strength of the relationships; they are a bit like standard univariate (or Pearson) correlation coefficients, except that they take into consideration multivariate relationships (they control for competing effects on each variable).

- The R-squared values reflect the percentage of explained variance for certain variables; the higher they are, the better the model fit with the data. In complex and multi-factorial phenomena such as health-related phenomena, many would consider an R-squared of 0.20 as acceptable. Still, such an R-squared would mean that 80 percent of the variance for a particularly variable is unexplained by the data.

- The P values have been calculated using a nonparametric technique, a form of resampling called jackknifing, which does not require the assumption that the data is normally distributed to be met. This and other related techniques also tend to yield more reliable results for small samples, and samples with outliers (as long as the outliers are “good” data, and are not the result of measurement error).

- Colinearity is an important consideration in models that analyze the effect of multiple predictors on one single variable. This is particularly true for multiple regression models, where there is a temptation of adding many predictors to the model to see which ones come out as the “winners”. This often backfires, as colinearity can severely distort the results. Some multiple regression techniques, such as automated stepwise regression with backward elimination, are particularly vulnerable to this problem. Colinearity is not the same as correlation, and thus is defined and measured differently. Two predictor variables may be significantly correlated and still have low colinearity. A reasonably reliable measure of colinearity is the variance inflation factor. Colinearity was tested in this model, and was found to be low.

- An effort was made here to avoid multiple data points per county (even though this was available for some variables), because this could artificially reduce the variance for each variable, and potentially bias the results. The reason for this is that multiple answers from a single county would normally be somewhat correlated; a higher degree of intra-county correlation than inter-county correlation. The resulting bias would be difficult to control for, via one or more control variables. With only two data points per county, one for males and the other for females, one can control for intra-country correlation by adding a “dummy” sex variable to the analysis, as a control variable. This was done here.

- Mortality from schistosomiasis infection (MSCHIST) is a variable that tends to affect the results in a way that makes it more difficult to make sense of them. Generally this is true for any infectious diseases that significantly affect a population under study. The problem with infection is that people with otherwise good health or habits may get the infection, and people with bad health and habits may not. Since cholesterol is used by the human body to fight disease, it may go up, giving the impression that it is going up for some other reason. Perhaps instead of controlling for its effect, as done here, it would have been better to remove from the analysis those counties with deaths from schistosomiasis infection. (See also this post, and this one.)

- Different parts of the data were collected at different times. It seems that the mortality data is for the period 1986-88, and the rest of the data is for 1989. This may have biased the results somewhat, even though the time lag is not that long, especially if there were changes in certain health trends from one period to the other. For example, major migrations from one county to another could have significantly affected the results.

- The following measures were used, from this online dataset like the other measures. P002 HDLCHOL, for HDLCHOL; P001 TOTCHOL, for TOTCHOL; and M021 SCHISTOc, for MSCHIST.

- SexM1F2 is a “dummy” variable that was coded with 1 assigned to males and 2 to females. As such, it essentially measures the “degree of femaleness” of the respondents. Being female is generally protective against cardiovascular disease, a situation that reverts itself a bit after menopause.

- MVASC is a composite measure of the two following variables, provided as component measures of mortality from all cardiovascular diseases: M058 ALLVASCb (ages 0-34), and M059 ALLVASCc (ages 35-69). A couple of obvious problems: (a) they does not include data on people older than 69; and (b) they seem to capture a lot of diseases, including some that do not seem like typical cardiovascular diseases. A factor analysis was conducted, and the loadings and cross-loadings suggested good validity. Composite reliability was also good. So essentially MVASC is measured here as a “latent variable” with two “indicators”. Why do this? The reason is that it reduces the biasing effects of incomplete data and measurement error (e.g., exclusion of folks older than 69). By the way, there is always some measurement error in any dataset.

- This note is related to measurement error in connection with the indicators for MVASC. There is something odd about the variables M058 ALLVASCb (ages 0-34), and M059 ALLVASCc (ages 35-69). According to the dataset, mortality from cardiovascular diseases for ages 0-34 is typically higher than for 35-69, for many counties. Given the good validity and reliability for MVASC as a latent variable, it is possible that the values for these two indicator variables were simply swapped by mistake.