Monday, June 12, 2017

Buying, Roasting, Grinding, Brewing: Coffee 101 -- How to Do it to Get the Optimal Stimulant and/or Health Effects

Don't worry you don't need a $10,000 machine to brew a better coffee.
Yeah, I know this is not the first article you've read to tell you how to brew the optimal coffee, I am pretty sure, however, that it's the first to define optimal by different, objectively measurable criteria such as the caffeine content, the amount of healthy chlorogenic acid in the brew, or the effect on blood lipids... sounds interesting?

Well, let's take a look at the not exactly all-encompassing data I came up with after dozens of hours I spend on Google Scholar and PubMed... what? Well, ok, maybe it were not so many dozens, after all ;-)
You can learn more about coffee and caffeine at the SuppVersity

For Caffeine, Timing Matters! 45 Min or More?

Caffeine Helps When Taken Intra-Workout, too

Coffee can Help You Get into Ketosis

Post-Workout Coffee Helps With DOMS

The Coffee³ Ad- vantage: Fat loss, Appetite & Mood

Quantifying the Benefits of Caffeine on Ex.
In order to present the results clearly, I decided to initially address different goals, like "maximizing caffeine content" individually to then, in a second step, investigate how/if there are interactions or contradictions between goals such as "make a high chlorogenic acid coffee" and "don't have that much caffeine in your coffee".

How to control the caffeine content of your coffee?

Please note that the amount of caffeine you will end up with is to a non-negligible extent determined by the type of beans you use. The data in Figure 1 exemplifies that, showing that more exotic types of coffee contain significantly less caffeine than the #2 global best sellers Arabica and Robusta (Hečimović 2011; Vignoli 2011).
Figure 1: Caffeine content (g/100g) of various coffee beans at different degrees of roasting (Hečimović 2011; Vignoli 2011).
So, the first determinant of the caffeine content of your coffee is the brand and product you choose (note: many of the cheaper products contain blends with a high content of cheap Robusta and little expensive Arabica). Practically speaking, this means that selecting a more exotic low caffeine coffee bean will never get you the same caffeine-power you'd get from a medium-roast Robusta bean.
Use your senses: Believe it or not, you can complement the data in Figure 1 with your senses. While you still won't be able to tell exactly how much extractable caffeine your coffee contains, deciding whether it tastes soft, hard, rio-ish or rio, will allow you to make some assumptions on whether your coffee is rather on the higher or lower end of the spectrum for its given type.

To makes those assumptions you can use my illustration of data from the Franca, et al. (2005) who found generally higher caffeine soft vs. hard coffees and (tainted) coffees with a medicine-like flavor.
Once you've determined the maximal caffeine content of your coffee by choosing the right bean and brand, grinding and brewing are the final steps to control the caffeine content. As early as in 1996, Bell et al. were able to show that the caffeine content of coffee solids to water volume ratio, the volume of coffee prepared, home versus store grinding, and drip/filtered versus boiling all affect the caffeine content of the final product.
Figure 2: Caffeine content of drip-prepared coffee according to the amount of coffee (in grams/ml water, i.e. 8/355 and 32/420) and the extent of store-grinding (left) and home-grinding (right | Bell 1996).
Modifying all these variables, the scientists produced coffees with caffeine contents per 177 ml (6 oz) ranging from 50 to 143 mg - obviously, depending on the ratio of coffee to water and the way the beans have been processed (coarse, medium or fine).
Where's 'cold brew' coffee? There's only a single study (a dissertation from the University of Mississippi by Jeffory Taylor Wallace from 2017) I could find that investigated the caffeine and chlorogenic acid content of cold-brewed coffee and found surprisingly little differences to regular coffee. There's minimally more chlorogenic acid in it and marginally less caffeine.
As you can see in Figure 2, a higher number of coffee solids (i.e. particles) and larger extent of grinding is associated with significantly higher caffeine contents in filtered coffee. Larger volumes of coffee prepared at a constant coffee-solids-to-water-ratio also yielded significantly higher caffeine contents. Homegrinding yielded caffeine contents similar to that of store-ground coffee. Boiled coffee had caffeine contents equal to or greater than filtered coffee, depending upon the length of boiling time. The variable caffeine contents in coffee resulting from the mode of preparation should be recognized and addressed by both food composition databases and epidemiologists.
Figure 3: Total extracted solids (~how thick is the coffee; correlates significantly with caffeine content R=0.91 according to Voilley, et al. 1981) of drip-prepared coffee as influenced by home-grinding (left) and caffeine content of boiled coffee (=coffee + hot water, no filter) as compared to drip/ filtered coffee (Bell 1996).
Needless to say that the scientists also evaluated the differential effect of filtering and simply boiling (mocha-style) the coffee (see Figure 3, right) -- with a surprising result: at least for brewing times of < 2min, using the differences to filtered coffee are less pronounced than you may have thought.

A more detailed analysis of the effects of different preparation methods can be found in Gloess et al. (2013) who compared coffee prepared according to different principles and with a broad spectrum of machines (espresso = strong black coffee made by forcing steam through ground coffee beans; lungo = 'espresso' made using much more water):
  • Figure 4: Caffeine content per 10ml of coffee brew (top), fatty acid content (relevant for effects on cholesterol, bottom-left), chlorogenic acid content (relevant for some of the health benefits, bottom-right) according to brewing method (details on the left | Gloess 2013).
    DE - Espresso from semi-automatic machine
  • DL - Lungo from semi-automatic machine
  • SE - Espresso from fully automatic machine
  • SL - Lungo from fully automatic machine
  • NE - Nespresso capsules (variety "Arpeggio") in the corresponding machine
  • Bia - Espresso from Bialetto espresso maker, Moka Express
  • Bo - French Press double-wall coffee maker Shin Bistro (Bodum, Switzerland).
  • KK - Karlsbader Kanne (KK) using the traditional Karlsbad method
  • F -  classic filter coffee was brewed using a paper filter and a coffee machine 
As you can see in Figure 4, there's a significant difference in the caffeine content depending on how the coffee is produced. What I find quite interesting is that the chlorogenic acid, yet not the fatty acid content (implications discussed later in this article) correlate with the caffeine content of the brew.
Table 1: Details of the extractions: the extraction method (for abbreviations—see text), the roast degree in Pt (measured with Colorette 3b, Probat), the milling degree and the amount of ground coffee per extraction in grams, the volume of water per cup or jug in milliliters, the extraction time in seconds, the extraction temperature in degrees centigrade, and the extraction pressure in bar are given | a std+mean; b,e according to manufacturer, c amount of coffee was automatically regulated by the machine, d boiler temperature, exact extraction temperature undefined (Gloess. 2013)
Whether and to which extent the different caffeine contents are related to the intricate differences in the degree of milling, the amount of ground coffee, the extraction time, temperature and pressure of the machines Gloess et al. used is beyond the scope of my review. For those of you who're geeky enough to want to figure that out, I've yet included Table 1, which provides the corresponding details of the preparation methods.
Relative quantities (g/ml) of caffeine in freshly brewed espresso coffee from 20 coffee vendors in Glasgow, UK. Data expressed as mean values (n = 3 samples for each coffee), std. error < 7% of mean values (Crozier 2012).
You prefer buying your coffee at Starbucks or elsewhere? Here's an overview of how much caffeine you will get if you buy freshly brewed espresso coffee from 20 randomly selected coffee brewers in Glasgow, UK (Crozier 2012).

In their study, Crozier, et al. analysed the caffeine and caffeoylquinic acid intake from commercial espresso coffee and found 6-fold differences in caffeine levels, and 4-fold differences in the caffeoylquinic acid:caffeine ratio, they ascribe to (a) variations in batch-to-batch bean composition, (b) possible blending of arabica with robusta beans, (c) roasting+grinding, and, predominantly, the number of beans used in the coffee-making/barista processes.
While caffeine may be the most obvious coffee ingredient one may want to modulate, it's only one out of several potentially health-relevant coffee ingredients.

How to minimize the potentially unhealthy fatty acid content?

In a day and age in which people deliberately add pounds of butter to their coffee, it may sound ridiculous to "minimize the fatty acid content of coffee" (this may also impair the mouth-feel and taste). For long-term SuppVersity readers, the reasons should be obvious, though: the fatty acid content of unfiltered coffee has been linked with increased cholesterol levels (learn more).
Figure 6: Using instant coffee or a filter will yield extremely reduced levels of the potential cholesterol-raising fatty acids. The composition, on the other hand, doesn't vary between "regular" and instant coffee (Ratnayake 1993). 
Oh, I already gave it away... Well, ok. If you didn't notice: the main determinant of the fatty acid content is whether or not you use a coffee filter. While filtered coffee contains almost no fatty acids and has no, if not beneficial effects on serum cholesterol (again, you can learn more about that in my article about the effects of Bulletproof Coffee on cholesterol), espressi contain 60–160 mg lipids/150-ml cup and coffees that were filtered through a metal screener still 50 mg lipids/150-ml cup (that's vs. <7 mg lipids/150ml cup with filtered coffee).
Figure 7: Effects of 9 weeks of coffee consumption on total, HDL and LDL cholesterol (Bak. 1989).
Whether the relatively small effects scientists like Bak, et al. (1989) observed over the course of 9 weeks (see Figure 6) is practically relevant is yet IMHO questionable. It could, however, explain why some studies, like the "Tromsö heart study", found sign. effects on serum cholesterol (Thelle 1983), while others didn't. That there's an effect, by the way, doesn't stand to question: if you give people only ~2g of the pure lipid-enriched fraction from boiled coffee for six weeks, you will increase their total cholesterol levels by 23%, their LDL levels by 29% and their triglycerides by a whopping 55%... well, that's at least what Zock, et al. observed in their 1990 six-week study in healthy volunteers.

It's not the fat in general, it's the diterpenes you want to filter out

The most likely candidates to trigger this increase in cholesterol are cafestol and kahweol, two diterpenes of which Urgert, et al. (1995) say that 10mg of cafestol per day will elevate cholesterol by 5mg/dl (0.13 mmol/L). Scandinavian boiled coffee contained (mean +/- SD) 3.0 +/- 2.8 mg, French
press coffee 3.5 +/- 1.2mg, and Turkish/Greek coffee 3.9 +/- 3.2 mg of cafestol per cup. Practically speaking this means that the...
"[c]onsumption of five cups per day of any of these coffee types could thus elevate serum cholesterol by 8-10 mgl dL [, with Italian espresso, which contains 1.5 +/- 1.0 mg of cafestol per cup], five cups [will] theoretically rais[e] cholesterol by 4 mg/dL" (Urgert 1995). 
I am nevertheless skeptical if the scientists' hypothesis that "[h]igh chronic intake of French press coffee or Turkish/Greek coffee could increase serum cholesterol and thus coronary risk" significantly - but hey, you can annihilate the risk if you use filter coffee ;-)

How to maximize chlorogenic acid and other health promoters?

If this is not the first coffee article you've read at the SuppVersity, you will know that coffee is regarded not by all, but by the majority of specialized researchers as a "functional food". One of its main active ingredients is chlorogenic acid (CGA or 5-CQA for "5-caffeoilquinic acid", which is the more precise chemical designation for "chlorogenic acid") of which you've seen in Figure 4, already that its content in the final brew is increased with the relatively novel fully- or semi-automated coffee brewing machines (those that work with whole beans, not the old filter coffee machines).
Figure 8: Chlorogenic acid content of selected espresso-based coffees bought in the UK (Crozier 2012).
As the data in Figure 8 tells you, this does not mean that the semi-automated machines at Starbucks and elsewhere would necessarily produce similarly superior chlorogenic acid contents.
Isn't roasting also bad because of acrylamide? You will be surprised to hear that, but studies show quite consistently that a medium roast time gives rise to the lowest levels of acrylamide (compared to light and dark roasting | Lantz 2006). It is also not true that instant coffee would have higher acrylamide levels than regular coffee - the simple reason almost all the acrylamide from the beans will actually end up in your coffee, no matter if you "water extract" it, as you would do it with a classic brew or pre-process the beans into a powder that will then dissolve in your coffee pot (Granby 2004). You don't have to be afraid, though, with 5-10µg/l you'd have to consume >20L of the highest acrylamide coffee per day to surpass the limit that's been associated with an increased cancer risk (>320L if you are concerned about your brain health | Tardiff 2010).
In fact, Starbucks, in particular, sucked in the previously introduced comparison by Crozier et al. (2012) - main competitor Costa Coffee (of which I can tell from personal experience that it also tastes much better), for example, contains 10.22x more of the purported health booster of which Tajik et al. (2017) write in their latest review of the literature that ...
"[i]t is postulated that CGA is able to exert pivotal roles on glucose and lipid metabolism regulation and on the related disorders, e.g. diabetes, cardiovascular disease (CVD), obesity, cancer, and hepatic steatosis. (Tajik 2017).
As the reviewers continue to point out, CGA has also been found to have a...
"[...] wide range of potential health benefits of CGA, including its anti-diabetic, anti-carcinogenic, anti-inflammatory and anti-obesity impacts, may provide a non-pharmacological and non-invasive approach for treatment or prevention of some chronic diseases" (Tajik 2017).
Just like you've learned it for caffeine, before, maximizing the CGA content in your home-brewed coffee starts with choosing a high CGA variety of beans. According to the results of the previously cited study by Vignoli et al. (2011) that's a problem for caffeine lovers, because (a) the high caffeine Robusta beans have lower CGA concentrations than the lower-caffeine Arabica beans and (b) the dark roasts which have the highest caffeine level, yield the lowest concentrations of CGA (note: most other studies found higher levels in Robusta vs. Arabica | cf. Ky 2001; Budryn 2009).
Figure 9: Comparison of CGA, caffeine and melanoidin content of diff. roasted Arabica and Robusta beans (Vignoli 2011).
The concentration of melanoidins (MLD), heterogeneous polymers with antioxidant, antimicrobial, anticariogenic, anti-inflammatory, antihypertensive, and antiglycative activities (Moreira 2012), on the other hand, is surprisingly unaffected by the degree of roasting - surprisingly, because melanoidins are Maillard products that form when sugars and amino acids combine (through the Maillard reaction) at high temperatures and low water activity.
The more realistic extraction method in Budryn's 2009 study suggests that Robusta beans are also the better choice for those of you who are looking to maximize the CGA content of their coffee. Roasting, on the other hand, reduced CGA in all studies.
Conflicting evidence suggests Robusta prevails for CGA, too: As with any natural product, you will always find studies showing opposing results for the caffeine and/or CGA content of different coffee varieties. Budryn, et al. (2009), for example, report slightly higher levels of CGA in Robusta vs. Arabica beans - that's interesting, because their way of extracting the CGA by either boiling or brewing the beans has a much greater resemblance to the regular practice of coffee making than the extraction method of Vignoli et al. whose data I've plotted for you in Figure 9. What Burdyn's data seems to confirm, though, is that roasting will reduce the amount of CGA significantly. Since people usually choose Robusta or Arabica-based on taste, focusing on light roasting may be the most practical way to maximize the CGA content in your coffee, anyway.
As pointed out in the previous box, the data in Figure 9 is not beyond doubt. With the relatively small difference in CGA content of Arabica and Robusta beans, I gravitate towards relying on the data from Budryn et al. (2009) to formulate the following recommendations:
  • If you're rather into the energizing effects of caffeine, you will buy a medium roast high  Robusta, low Arabica blend.
  • If you're rather into the potential health benefits of chlorogenic acid (and melanoidin), you buy a lightly roasted high Robusta, low Arabica blend.
For those of you who're more interested in the health benefits, the in vitro antioxidant activity of the soluble coffee samples from Vignoli et al. (2011) may come handy as well. In contrast to the mere content of CGA and MLD, the results of the ABTS, FRAP and Folin assays give you an estimate of the combined antioxidant effects of all pro- and anti-inflammatory molecules in the beans.
Figure 10: Total antioxidant (ABTS, FRAP) activity and phenol content (Folin) of Arabica and Robusta beans at different degrees of roasting; note: these results take the effects of all molecules in the beans into consideration (Vignoli 2011).
As the overview of the results in Figure 10 tells you, these analyses point toward a higher antioxidant (ABTS and FRAP) and phenol content (Folin) assay in those beans with the highest caffeine content: medium or dark roast Robusta beans.
Figure 11: Total phenolic compounds (a) and ABTS antioxidant capacity (b) extracted from ground coffee by filter, plunger, mocha, and espresso methods. Bars with different letters are significantly different (p < 0.05 | Pérez-Martínez 2010).
Similar results have been observed by Hečimović et al. (2011) in the exotic coffee varieties Minas, Cioccolatato, Vietnam, and Cherry - with a slight phenol advantage of Cherry and a visible antioxidant advantage of Vietnam over the other coffee varieties, respectively. And data from Pérez-Martínez, et al. (2010) tells you that boiling is the #1 for the total antioxidant activity while the hot steam extraction that's used for espressos produces the coffee with the highest total antioxidant activity (Figure 11).
Did you know that a coffee post-workout will reduce DOMS
Wow! That's a lot to process. Let's recap: The caffeine content can be maximized by choosing a medium or dark roast of high(er) caffeine beans (Robusta > Arabica > most exotic brands | large effect), by grinding the beans more thoroughly (large effect), by using more coffee per ml of water (or volume of steam for classic espressos | large effect), and boiling the coffee directly and/or using an espresso machine instead of filtering it (this is a surprisingly small effect, though, so don't worry about using a filter).

Increasing the temperature in the range from 88-98°C, on the other hand, has no significant effect on the caffeine content of your every morning favorite (Andueza 2003).

For the health benefits, things are a bit more complicated. If you want maximal chlorogenic acid levels, though, a no- or only lightly roasted coffee should be your first choice (Budryn 2009; Moon 2009; Vignoli 2011). Whether you should prefer Arabica or Robusta beans is not clear, I personally put more faith in the data from Budryn 2009 pointing to Robusta being the better CGA source. Compared to the effect of roasting the choice of the blend is yet negligible.

With all the hype around green (unroasted) coffee and its high CGA content, though, we must not forget that a medium roast Robusta bean (or high Robusta blend) would be your #1 source if you define "healthy" by the overall (in vitro) antioxidant effects of the brew - instead of looking at CGA, alone. Choosing higher water pressures in espresso machines and longer exposure times of coffee to hot water are two other proven means to increase the amount of antioxidants in your coffee (Andueza 2003) - with boiling having a general total antioxidant and the espresso method having an advantage in terms of the total phenol content of the final brew. Similarly, a darker roast, may have reduced CGA levels, its total antioxidant prowess, however, has been found to be increased (over lighter roasts) only recently (Jung 2017). Last but not least, you should be aware that health-fanatics who buy decaffeinated coffee will, in general, get lower amounts of antioxidants per serving than people who prefer regular coffee (Parras 2007).

What's left to discuss? Oh, yes: The cholesterol issue: If you already have problems with your blood lipids, you may further want to consider using a classic paper filter to avoid exposing yourself to the cholesterol-raising coffee diterpenes... What? You want to know what tastes best? I am sorry, but you will have to decide that for yourself. Taste is simply too individual to make a general statement about "what's best" | Comment!
References:
  • Andueza, Susana, et al. "Influence of extraction temperature on the final quality of espresso coffee." Journal of the Science of Food and Agriculture 83.3 (2003): 240-248.
  • Bak, Annette AA, and Diederick E. Grobbee. "The effect on serum cholesterol levels of coffee brewed by filtering or boiling." New England Journal of Medicine 321.21 (1989): 1432-1437.
  • Bell, Leonard N., Clinton R. Wetzel, and Alexandra N. Grand. "Caffeine content in coffee as influenced by grinding and brewing techniques." Food Research International 29.8 (1996): 785-789.
  • Budryn, Grażyna, et al. "Effect of different extraction methods on the recovery of chlorogenic acids, caffeine and Maillard reaction products in coffee beans." European food research and technology 228.6 (2009): 913-922.
  • Crozier, Thomas WM, et al. "Espresso coffees, caffeine and chlorogenic acid intake: potential health implications." Food & function 3.1 (2012): 30-33.
  • Franca, Adriana S., Juliana CF Mendonça, and Sami D. Oliveira. "Composition of green and roasted coffees of different cup qualities." LWT-Food Science and Technology 38.7 (2005): 709-715.
  • Gloess, Alexia N., et al. "Comparison of nine common coffee extraction methods: instrumental and sensory analysis." European Food Research and Technology 236.4 (2013): 607-627.
  • Granby, Kit, and Sisse Fagt. "Analysis of acrylamide in coffee and dietary exposure to acrylamide from coffee." Analytica Chimica Acta 520.1 (2004): 177-182.
  • Hečimović, Ivana, et al. "Comparative study of polyphenols and caffeine in different coffee varieties affected by the degree of roasting." Food chemistry 129.3 (2011): 991-1000.
  • Jung, Soohan, et al. "Cellular Antioxidant and Anti-Inflammatory Effects of Coffee Extracts with Different Roasting Levels." Journal of Medicinal Food (2017).
  • Ky, C-L., et al. "Caffeine, trigonelline, chlorogenic acids and sucrose diversity in wild Coffea arabica L. and C. canephora P. accessions." Food chemistry 75.2 (2001): 223-230.
  • Lantz, Ingo, et al. "Studies on acrylamide levels in roasting, storage and brewing of coffee." Molecular nutrition & food research 50.11 (2006): 1039-1046.
  • Moon, Joon-Kwan, Hyui Sun Yoo, and Takayuki Shibamoto. "Role of roasting conditions in the level of chlorogenic acid content in coffee beans: correlation with coffee acidity." Journal of agricultural and food chemistry 57.12 (2009): 5365-5369.
  • Moreira, Ana SP, et al. "Coffee melanoidins: structures, mechanisms of formation and potential health impacts." Food & function 3.9 (2012): 903-915.
  • Parras, P., et al. "Antioxidant capacity of coffees of several origins brewed following three different procedures." Food chemistry 102.3 (2007): 582-592.
  • Pérez-Martínez, Mónica, et al. "Influence of brewing method and acidity regulators on the antioxidant capacity of coffee brews." Journal of agricultural and food chemistry 58.5 (2010): 2958-2965.
  • Ratnayake, W. M. N., et al. "Lipid content and composition of coffee brews prepared by different methods." Food and chemical toxicology 31.4 (1993): 263-269.
  • Tajik, Narges, et al. "The potential effects of chlorogenic acid, the main phenolic components in coffee, on health: a comprehensive review of the literature." European Journal of Nutrition (2017): 1-30.
  • Tardiff, Robert G., et al. "Estimation of safe dietary intake levels of acrylamide for humans." Food and Chemical Toxicology 48.2 (2010): 658-667.
  • Urgert, Rob, et al. "Levels of the cholesterol-elevating diterpenes cafestol and kahweol in various coffee brews." Journal of agricultural and food chemistry 43.8 (1995): 2167-2172.
  • Vignoli, J. A., D. G. Bassoli, and M. T. Benassi. "Antioxidant activity, polyphenols, caffeine and melanoidins in soluble coffee: The influence of processing conditions and raw material." Food Chemistry 124.3 (2011): 863-868.
  • Voilley, A., et al. "Influence of some processing conditions on the quality of coffee brew." Journal of Food Processing and Preservation 5.3 (1981): 135-143.
  • Wallace, Jeffory Taylor. An Analysis of the Acid Profile of Coffee Brews: Caffeine and Chlorogenic Acid Concentrations in Different Forms of Coffee Brew. Diss. University of Mississippi, 2017.
  • Zock, P. L., et al. "Effect of a lipid-rich fraction from boiled coffee on serum cholesterol." The Lancet 335.8700 (1990): 1235-1237.