Sunday, July 20, 2014

Nutrient Database (USDA)

I previously used some entries in the USDA nutrient database and organised them in such a way to compare the nutrient density of foods and food groups.  More recently I did a similar thing but used all the entries in the NUTTAB nutrient database (Australia) to form another nutrient database.  This time I used all (8463) the entries in the USDA database

To download the nutrient database click here.  Don't try and read it in the Google drive viewer because you won't be able to see much of it, just download it

Sunday, July 13, 2014

Comparing Nutrient Density with the Nutrient Database

One of the main purposes of the nutrient database was to compare the nutrient content of foods and food groups in a meaningful way - the amount of nutrients in 2000 kcal of food divided by the RDI.  All values in this post are based on that measure 

Whole Foods are Superior 

I firstly compared categorised the groups into whole foods, SAD meals, junk foods and extras*.  It’s clear from this that whole foods on average are superior and good sources of almost all nutrients, such that simply eating a balanced whole food diet is very likely to fulfil almost** all nutrient needs and more.  There’s something to be said for ‘just eat real food’ (JERF) 

In this very rough measure, the whole food average was only lacking in fluoride (which you get in the water supply), manganese, sodium (which you can easily add to foods) and vitamin A.  Manganese is mostly found in plant foods, particularly grains.  The AI is based on the median population intake and so I think our grain-based diet skews estimates on how much manganese we need.  Vitamin A is kind of surprising, but it’s really only found in sufficient quantities in vegetables, eggs, liver, dairy and some fruits; where certain vegetables and liver are extremely good sources of vitamin A.  This might be just because I took the medians of groups rather than the mean 

The weight and macronutrients are also interesting.  Whole foods are generally high in protein and less calorie dense than SAD meals and junk food, and the SAD meals is remarkably close (by accident) to the macronutrient ratio eaten by Australians (~17:33:45).  There’s also like a gradient where protein decreases and energy density increases as foods/meals get more processed with increasing amounts of added refined sugars, starches and fats.  That being said, the averaged protein in the whole foods category is too high
* Whole foods included: dairy (average of cheese and milk), eggs, fruit, aboriginal plant foods, legumes, offal, nuts and seeds, crustacea and molluscs, vegetables and meat (which is an average of beef, game and other meat, lamb, mutton, pork, poultry, veal and fish) 

SAD meals included: bread and bread products, breakfast cereals, flours, grains and starches, hamburgers, pizza and other takeaway products, noodles and pasta, dairy and meat alternatives, processed meats, asian restaurant foods, mediterranean restaurant foods, processed fish, crustacea and molluscs and soups 

Junk foods included: biscuits, cakes, slices and other battered products, pastries, pies and tarts, ice cream & edible ice products, yoghurts and dairy desserts, snack foods, chocolate based and sugar based confectionary 

NOTE: I averaged the meats together and dairy together because there were several groups of meat that would otherwise skew the average and also by averaging the groups of meat and dairy into two groups it resulted in 5 groups of animal foods and 5 groups of plant foods.  Within these groups: ‘milk’ was more nutrient dense than ‘cheese’; and ‘game and other meat’, ‘veal’ and ‘fish’ were more nutrient dense than other meats 

** There are some nutrients that aren’t widely found in foods in sufficient quantities, such as choline/betaine 

Which Whole Foods are More Nutrient Dense? 

By using the nutrient database we can also get an idea as to which food groups are more nutrient dense, which you can see in the table below arranged from most nutrient dense to least
Obviously there are problems with calculating nutrient density this way:
  • In the USDA nutrient database the same measurements was higher but particularly for vegetables (3.5) and fruits (1.24), which seems to be due to a lot of 0’s in the raw data where you would expect there to be something
  • Values like 106.90 for B12 in offal that inflates the average, although in this comparison those really high values didn’t make much difference in the ranking (except aboriginal plant foods, see first point)
  • This doesn’t account for some missing micronutrients (choline/betaine, K1, K2), bioavailability and other nutrients/beneficial compounds in foods

Tuesday, July 8, 2014

The Chowdhury Meta-Analysis

In March this year, another meta-analysis (by Chowdhury, et al) was published that looked at the relationship between fats and coronary heart disease (CHD) [1].  Its conclusions were that Current evidence does not clearly support cardiovascular guidelines that encourage high consumption of polyunsaturated fatty acids and low consumption of total saturated fats”.  As expected, the mainstream was very critical of the paper, but mainly regarding the inclusion/interpretation of observational studies and those related to omega 3 supplementation.  Anyway, I’m going to ignore all that and focus on their interpretation of the clinical trials that replace SFA with omega 6 PUFA 

They included randomised controlled trials with 50 or more total coronary outcomes.  This criterion excluded the unfavourable Rose Corn Oil Trial (18 events), but didn’t exclude the favourable STARS trial (7 by their assessment).  Not like that matters much if you’re simply going to do a quantitative assessment as these two trials are too small to have a noticeable effect on the outcome (a relative risk (RR) assessment).  Their description of the trials they included and the risk of bias assessments are shown below (click to enlarge)

*Mixed poly-unsaturated intervention with linoleic acid as the primary fatty acid
Note: the age of the men in the FMHS was 34-54 (not 34-44)
NFMI = non-fatal myocardial infarction; FMI = fatal myocardial infarction; FCHD = fatal coronary heart disease; SCD = sudden cardiac death
Based on the trials, they calculated the RR for omega 6 supplementation to be 0.89 with a confidence interval of 0.71 to 1.12, making it not statistically significant.  But when they excluded the Sydney Diet Heart Study (SDHS) in a sensitivity analysis the RR was 0.81, which was significant

There are some problems with the meta-analysis: 

Under omega 6 supplementation, they included trials that did much more than simply replace SFA with omega 6.  The Oslo Diet Heart Study, Finnish Mental Hospital Study (FMHS) and STARS are the best examples of this as there were many other differences between the control group and the high omega 6 group that were almost always favorable to the high omega 6 group (such as less trans fats, more omega 3, fruit and vegetables, weight loss, etc).  But you also have other trials like the LA Veterans Administration Trial, which was mostly well done, except the control group was eating reheated butter as a main fat source and consequently had an insufficient intake of vitamin E (2.6 mg and the RDI is 10 mg) 

They only reported CHD outcomes and didn’t include total mortality.  Previous meta-analyses by Hooper, et al and Mozaffarian, et al suggest some benefit for CHD, but find no difference in total mortality.  What good are these interventions if CHD mortality decreases, but total mortality stays the same? 

Supportive of DHH for CHD Events
Supportive of DHH for CHD Mortality
Supportive of DHH for Total Mortality
Hooper (Cochrane)
RR = 0.82
CI = 0.66 to 1.02
RR = 0.92
CI = 0.73 to 1.15
RR = 1.02
CI = 0.88 to 1.18
(+5% PUFA)
RR = 0.81
CI = 0.70–0.95
RR = 0.80
CI = 0.65–0.98
RR = 0.98
CI = 0.89–1.08
* Significant difference 

Ultimately the problem with all the meta-analyses of this type is that they treat the trials as if they’re same, ignore the other differences between the groups and the overall quality of the trial, and then just run some statistics.  When you have such (mostly) poor and variable trials (with no further ones being done) I think a qualitative approach is the best way to discuss these trials and arrive at a conclusion 

Walter Willett made a similar point (if only he tell this to his colleagues): 

The controversy should serve as a warning about meta-analyses, Willett adds. Such studies compile the data from many individual studies to get a clearer result. "It looks like a sweeping summary of all the data, so it gets a lot of attention," Willett says. "But these days meta-analyses are often done by people who are not familiar with a field, who don't have the primary data or don't make the effort to get it." And while drug trials are often very similar in design, making it easy to combine their results, nutritional studies vary widely in the way they are set up. "Often the strengths and weaknesses of individual studies get lost," Willett says. "It's dangerous." [2]

Monday, June 30, 2014

Nutrient Database (NUTTAB)

Download the nutrient database here.  Don't try and read it in the Google drive viewer because you won't be able to see much of it, just download it

This is very similar to my old nutrient database (sourced from the USDA nutrient database), but for this I used all of the data from NUTTAB (2190 foods), which is a nutrient database for Australia.  The RDIs and AIs are from Australian NRVs

Unfortunately NUTTAB isn't as complete as the USDA database and doesn't include vitamin D, vitamin K and choline, but it does include more minerals such as chromium, iodine and molybdenum

Sunday, June 22, 2014

Nutrient Database

Download the nutrient database here.  Don't try and read it in the Google drive viewer, just download it.
All of the data I used was sourced from the USDA database and the RDIs and AIs are from Australian NRVs.


When I came around to looking at nutrients I had no way of knowing for sure if a particular food was high or low in it.  It’s easy to say a particular food is high or low in a given nutrient, but very rarely are these claims justified.  I wanted to find an objective way to know what nutrients are in food.

I also wanted to put the quantity of nutrients in context.  Rather than listing nutrients for every 100 grams of food, it’s more relevant to discuss nutrient density per calorie, as we are limited by the calories we can consume, rather than by weight.

Finally I wanted to compare foods on many nutrients.  When we list positive qualities of food, we often limit ourselves to a few nutrients at the most.  Functional foods add some nutrients and are then marketed as healthy.   There are many nutrients; nutrient dense foods are those that have at least sufficient quantities of most of them.  At this point I found one website that got as far as measuring nutrient density by calories and comparing it to the RDI.  What they didn’t do, which I think is important, is to list as many nutrients as possible for each food, to compare many foods, group like foods together and then compare those food groups.

The Process

I got all the data directly from the USDA nutrient database.  After I got enough I arranged the rows by nutrient.  Initially the foods were measured by 100g so I made a new sheet and changed the quantity of food to equal 2,000 calories.  I picked those nutrients that had an RDI, made a new sheet, and divided the quantity of nutrients (in 2,000 calories of food) by the RDI for each nutrient and each food.  I separated the foods into food groups, got the medians for the nutrient/RDI values for each nutrient, and then compared the food groups on this basis.

(The RDI is the amount of a nutrient so 97.5% of the population aren’t deficient in it.)

No measurement is perfect.  I used median rather than mean (average) because there were some foods with ridiculously huge nutrient/RDI values.  Liver for instance has 81 times the RDI for vitamin A.  Liver is grouped with other organ meats and shellfish, both of which don’t have very much vitamin A.  If I used the mean, the number would have suggested that the average organ meat and shellfish group has 9 times the RDI for vitamin A, which is true, but misrepresents that group.  Although in appropriately representing the group, I have misrepresented liver as the highest vitamin A containing food.  It’s a trade-off.  I suppose the knowledge of any foods that are extraordinarily high in a nutrient is fairly well known.  There’s nothing to stop you looking within the food groups themselves.

I made some assumptions and manipulated the data a little:-

·         I took all fibre to yield 2 calories
·         I often used higher alternatives of the RDI and those for 70kg adult men
·         For vitamin A I took the vitamin A RAE (retinol activity equivalent), converted it to IU by multiplying by 3.33 and used that measurement for the vitamin A row
·         Short chain omega 6/3 refers to 18 carbon long fatty acids
·         Short chain omega 6 came from 18:2 undifferentiated minus other 18:2 isomers, such as trans LA and rumenic acid, to approximate the amount of LA (18:2) and then I added GLA (18:3).  The approximation of LA is slightly high because there usually wasn’t data from the other isomers to subtract, but because GLA was often 0 because it was rarely measured directly it sort of balances out
·         Short chain omega 3 came from 18:3 undifferentiated minus other 18:3 isomers such as GLA or conjugated linolenic acids, to approximate the amount of ALA (18:3) and then I added stearidonic acid (18:4).  The approximation of ALA is slightly high because there usually wasn’t data from the other isomers to subtract
·         Long chain omega 6/3 refers to 20+ carbon long fatty acids
·         Long chain omega 6 came from 20:3 and 20:4 undifferentiated.  In each of these there is an omega 3 isomer, but they weren’t as common as DGLA and AA.  I also added adrenic acid (22:4).
·         Long chain omega 3 is simply the addition of EPA (20:5), DPA (22:5) and DHA (22:6)
·         Milk was measured for rumenic acid (18:2 i), but not for CLAs, so I used the rumenic acid for milk instead of the CLAs
·         Methionine and cysteine, and phenylalanine and tyrosine were added together for the RDI measurements because the RDI was for each group.  This is because in each group the essential amino acid can be synthesised into the non-essential amino acid, which is more commonly used by the body
·         Histidine is an essential amino acid, but we don’t need much of it and so the RDI is low.  The very low RDI caused histidine/RDI ratios of 100+ to be the norm.  I thought it was useless information and removed it


I remember reading that to measure each food cost the USDA $2,000.  Getting that data wasn’t easy and therein lies the problem.  Even although the data was the best I found it was still incomplete.  Most of the incompleteness came from the more difficult tasks such as the isomers of fats.  So the following were missing to a degree:-

·         Biotin (vitamin B7) was missing from all the foods.  The RDI for biotin is 30 µg, which is a really small amount.  Biotin is found in roughly the same foods as vitamin B6, which could act as a proxy for biotin on the spread sheet.
·         Vitamin K2 (menaquinoe-4) was missing from almost all foods.  Vitamin K1 is also measured in µg and we eat much less K2 than K1 (about 10 times) so the K2 presence in food may have been difficult to measure.  Also vitamin K, especially K2, is a very under researched and not a well-known vitamin
·         Fats such as CLA (rumenic acid), EPA, DPA, DHA and fat soluble vitamins such as A, D and E often were measured as 0 in animals that must obviously contain some.  EPA and DHA are essential fats and the fat soluble vitamins are essential nutrients that would both be incorporated into the animal’s tissues.  Even grain feeding in feedlots doesn’t entirely deplete these fats
·         Isomers of 18:2, 18:3, 20:3 and 20:4 were rarely measured individually

What this also doesn’t measure or account for is:-
·         The bioavailability of nutrients
·         Other nutrients that don’t have an RDI
·         The variation of nutrients in plants that depends on soil (minerals) and health (vitamins)
·         The variation of nutrients in animals that depends on what they eat
·         Just meeting the RDI doesn’t necessarily mean healthy, the amount of a nutrient so that 97.5 of the population have optimal blood levels may be considerably higher