ex150nostarbucks review: down 13lbs, 6 of which fat

If I had a time machine I would go back 6 months and smack myself.

Wow, this trial I ran for 45 days. Why so long? Well, I decided to prolong it at around 30 days because I thought I was seeing a trend, but I wasn’t sure. I’m glad I did it.

I almost forgot how to write these reports.

Previously on ex150

We have to get back to the fat loss!

As you can see on my progress page, I’ve been in somewhat of a fat loss-malaise for half a year now. It’s not that where was no fat loss, but it seemed slower. After protein refeeds, which I do regularly, it took forever for the weight to come off.

I still hit a new low (since 2017) of 235lbs in June. But then, after a 10 day protein refeed, I didn’t even get back down to those 235lbs for the entire next month.

What was happening?

I am the Very Model of a Major Ketard-General

Let me begin this section with a song.

I am the very model of a major ketard-general,
I've macros of all vegetable, animal, and mineral,
I'm adding fats to everything, have never feared cholesterol,
In terms of ketogenesis, my level's therapeutical.
I'm very well acquainted, too, with matters mitochondrial,
I understand the endocrine, from glucagon to cortisol,
About insulin resistance I am teeming with a lot o' news,
With many dreadful facts, foremost of seed oils and fructose abuse.

Why am I a major ketard-general? I think I might have blinded myself to the cause of The Malaise. In retrospect, I have no idea how I could’ve missed this.

It all started with this Twitter conversation:

Immediately after I posted the “I think I’m already in a deficit” tweet I realized it couldn’t be true. To quote myself: "Caloric deficit" literally means "fat loss."

Hence, if I wasn’t losing fat (and I wasn’t) I wasn’t in a caloric deficit.

A couple hours later while driving (I do most of my best thinking while driving!) it hit me like a ton of bricks.

The heavy cream Starbucks lattes.

In my original post on the macros of ex150, I wrote:

One of the coffees is often a Starbucks Latte breve (made with half & half). I couldn’t find any nutrition info on their website for breve, but the regular one is ~220kcal with 15g of carbs and 11g of fat. The breve will have fewer carbs and more fat, so I’m just counting it as one coffee here, because mine also have ~200kcal/cup

That’s technically already incorrect if you read it: the regular (=2% milk?) latte has ~200kcal/cup. The breve clearly would have more. Reading that post now, I just seem to brush that under the rug.

And then, a few months later, I switched to heavy cream lattes.

Back of the napkin math

Unfortunately, Starbucks’ website doesn’t show you the nutrition information for “modifications” like a latte breve, or a latte heavy cream… or, I just realized, even whole milk.

So let’s try to reverse-engineer it. I always got the 24oz, almost always iced.

Ok, 180kcals, 6g fat, 18g carbs, 12g protein. Seems pretty skim to me, not like whole milk.

Pretty close. This has a gram more fat, but the calories, carbs, and protein are spot on. So let’s assume that the 24oz cup contains 8oz of ice, maybe 2-3oz of espresso shots, and about 1.5 cup / 13 oz / 366g of milk. I confirmed with a quick search that this is typically what baristas who’ve worked at Starbucks used. Of course it’s subject to fluctuation, depending on the individual barista that day.

They seem to default to 2% milk, which I hadn’t even realized. I guess low-fat milk foams better?

In order to save images, I’m not going to screenshot all of the nutrition data. Here’s the summary:

2% milk: 183kcal / 7g fat / 18g carbs / 12g protein
Whole milk: 223kcal / 12g fat / 18g carbs / 12g protein
Half & half: 447kcal / 37g fat / 16g carbs / 11g protein

Do you see where this is going?

Heavy cream: 1213kcal / 129g fat / 10g carbs / 10g protein

Edit: As Scott points out in the comments, according to some sources (https://starbmag.com/how-many-calories-are-in-starbucks-heavy-cream/) the “Heavy Cream Latte” is just a normal latte with a few 1-ounce splashes of heavy cream depending on size, e.g. 3oz for my venti (24oz) size. I’m not sure if they gave me 2% + 3 splashes, half & half + 3 splashes (since I came from Breves for months). Might’ve depended on the barista and the location. A breve + 3oz of heavy cream would clock in around just under 1,000kcal per drink.

And of course, I wasn’t just drinking one of these per day. Luckily I am the most anal-retentive person I know, and I had logged 100% of my food & drink intake for the half year I wore the CGM.

On average, I had 1.3 lattes per day over 150 days of CGM usage.

1.3 * 1213kcal = 1576kcal.

So I was, on average, consuming an additional 1,500kcal/day that I had just brushed under the rug in my original macros post as “another coffee with 200kcal.” This is about 73% more cream than I was usually consuming - nearly doubling my already high cream intake!

That puts me at about 3,900kcal/day total, instead of the 2,300kcal I was originally recording.

Now that’s a confounding variable if I’ve ever seen one. In my first post on my understanding of CICO and caloric deficits/surpluses - which is still one of my favorite posts, by the way, check it out if you haven’t! - I explain it with this simple image:

In short, my argument went, you wanted to fix your fuel partitioning instead of “reducing calories in” or “increasing calories out.” That way, you’d be satiated and would naturally eat at a steep external deficit, allowing your body to make it up via the internal energy supply: body fat.

Well, turns out that even if you get your fuel partitioning sorted by eating ex150, it is still possible to completely saturate your metabolism with the fat you’re consuming.

At this point I suspect that flushing my system with excess cream every day eliminated my body’s need to use up any of that body fat, and that all my experiments were completely confounded by this, to varying degrees, for over 6 months.

Here is a reconstruction, from CGM and other records, of my phases of latte consumption:

The spikes, by the way, are not from the lattes - those are my protein refeeds.

If you squint, you could argue that the 2% lattes actually didn’t make a difference. That’s likely the reason I initially brushed them under the rug. At ~200kcal, they really aren’t much more than a cup of my regular old coffee that I drink at home.

But the Breve phase really seems slow. And long. There is still some fat loss, but it’s slowed way down. Recovering from refeeds takes much longer now. There’s a whole 2-month period where I’m basically not losing any weight on net. Overall, I only lost 15lbs from February to June. Unacceptable!

Finally, about 1.5 months of heavy cream lattes. If you discount the initial water weight loss, I go from 243lbs to 241lbs over the entire month of July. That’s not quite nothing, but it’s not exactly impressive fat loss.

How did I miss this?

Honestly, this is so bizarre to me. I feel like one of those people who “do low-carb” but then mention they ate waffles with syrup and honey and bananas for breakfast, seemingly not realizing that’s entirely carbs.

I’m the guy who has a script to calculate his macros. I double, triple, and quadruple checked it dozens of times. I got regular RMR scans to verify my metabolism hadn’t slowed down. I did the Ray Peat body temperature thing, usually coming out in the 98s, aka just fine.

What calories?

At the end of the day, it was probably a combination of being a Major Ketard-General and it kind of sneaking up on me.

My initial assessment was correct - a standard, 2% milk latte has less than 200kcals. It’s therefore reasonable to just count it as “one more coffee.”

The breve already clocks in at 450kcal. At an average of 1.3 a day, that’s nearly 600kcal - getting into the significant territory.

And then when I switched to heavy cream, I don’t know if I ever did the math. I don’t think so. I do remember thinking that it would surely be more calories, but not bothering to do the math - and Starbucks doesn’t provide the information for anything but 2% milk.

That’s the sneaking up part.

I think the other part is that I had my keto blinders on. Surely heavy cream was better than that sweet, sugary, carby milk, right?

Maybe the CGM even helped trick me here. I originally switched to breve lattes after discovering that they spiked my blood glucose less than the regular ones. But a rare, small spike probably doesn’t make a difference for long-term fat loss, whereas consuming an extra 1,500kcal can, even if they come from heavy cream.

So with all this preamble, let’s get to..

The ex150nostarbucks trial

There’s a very curious shape to the curve. As usual, I’m including a few days before the start of the trial as context.

I was at around 243lbs when I did a 24h protein refeed between back-to-back DEXA scans. That’s the big spike, peaking at 248.9lbs.

It took 5 days for the water weight to come back down.

Then, for 9 days (!), I completely plateaud. This was pretty frustrating. Was my Starbucks theory wrong? Surely cutting 1,500kcal/day should make a huge difference?

14 days into the trial, things started moving. I started seeing some near-daily weight drops.

There was a small bump the day after I took the Kraft test, but nothing too bad. Could even be just a coincidence. I expected the effect to be larger.

And this is why I extended the ex150nostarbucks trial to 45 days. Since it took 14 whole days (5 for water weight, 9 for the plateau) to even get to the “real” fat loss, was I just seeing a short-term trend? I’d had many of those, even over the 6 months of The Malaise. I’d need a longer stretch. 45 days would basically give me an entire 30 day trial, unadulterated, on top of the 14 day wash-out period.

There was a little bit of a weight-loss hiccup when I took a road trip over the Labor Day weekend. I’m quite sure I didn’t exceed my calories on that, as I counted everything the whole weekend and didn’t deviate whatsoever.

But driving for hours a day, the heat, and maybe the physical stress of travel might’ve triggered some water retention. In any case, it seems to have cost me 4-5 days. So nothing too crazy.

My favorite thing after the plateau: it just went so linearly and smoothly. Not a single plateau over maybe 2-3 days. Even the road trip wasn’t a big setback since I knew I hadn’t deviated from the diet. And it came right back down.

I ended at 235.45lbs this morning - back to my all-time-low since 2017. Wooo!

About that plateau… it seems familiar

During this 9 day plateau, I wondered. Hm, weird. I’d seen this type of plateau before, hadn’t I?

Oh yes, in pretty much every single one of the case studies I just wrote up over the last few months. Check out the last case study and then click through to the previous ones, and look at the graphs.

Everybody loses a ton of water weight in the first few days. That’s expected when going to a very ketogenic diet. Most people lost weight after that, albeit more slowly. But for almost all of them, there’s quite a long plateau after that first water weight.

I wonder what it is. The body recalibrating something for fat loss? Glycogen depletion?

Hunger pangs?!

Day 4 of ex150nostarbucks, I got really bad hunger pangs in the evening. My stomach started to rumble. This was very surprising, because I haven’t really felt “hungry” in years, basically since starting keto. The only time was when I did prolonged, multi-day fasts.

But I suppose suddenly cutting daily calories by 1,500kcal took my body some time to adapt?

The feeling would go away when I ate my evening cream. I think I instinctively drank more coffee with more cream the next few days, and the feeling went away. It came back a few times when I didn’t drink enough coffee cream throughout the day, typically around 8pm or so.

Grass-fed, then back to ground chuck

Since the previous trial was ex150grassfed, I thought I’d continue with grass-fed meat for the first 14 days. That way, the type of beef wouldn’t be the difference.

Of course, it took exactly 14 days for the water weight and the plateau to go by, so now I accidentally confounded it even worse - did going back to ground chuck make all the difference? I mean, probably not, right? FML. Science is hard. Both types were 80 lean/20 fat.

Questions & Takeaways

All this calorie math is a vague estimate with lots of averaging

How much cream exactly do they put into a Starbucks 24oz iced latte? Probably depends on the day, the person, and so on.

I also took the average over 150 days to come up with the 1.3/day number. There were probably weeks where I averaged <1, and others where I averaged >2.

Plus, this is all reverse-engineered from what I think is the 2% regular latte on their website.

Is The Malaise over now?

Time will tell, I guess, but the trend for the last 30 days has been very positive. Accidentally consuming 1,500kcal/day more than I thought is definitely by far the best explanation for The Malaise I have so far.

Why did cement-truck satiety not work?

Good question. I actually remember getting cement-truck satiety from a heavy cream latte the very first time I bought one. It took me nearly an hour to finish the 24oz coffee. Usually, I finish them in 10-15 minutes.

Theories:

• Cement-truck satiety works better with whipped, not liquid cream

• The espresso flavor makes it too delicious

• Pouring it over ice makes it too delicious

• Since I spread the lattes out, e.g. morning and lunchtime, with whipped cream for dinner: maybe the satiety only lasts a few hours, not the whole day?

What does "ad-libitum" mean now?

ex150 is “ad-libitum heavy cream.” This accidental experiment makes me wonder what exactly ad-lib even means.

Before the breve/heavy cream lattes, I was perfectly satiated every day eating ~2,300kcal/day.

But apparently I can also consume about 4,000kcal/day without any trouble, at least after a very brief adaptation period (those first few heavy cream lattes).

So is ad-lib anything between 2,300 and 4,000kcal/day? How many more of those heavy cream lattes could I drink? Probably a few, if I tried. Maybe I could go down to 2,000kcal if I tried? Or even 1,500?

So you’re saying it’s all calories?

Well, I didn’t gain any fat on nearly 4,000kcal/day for 1.5 months. I even lost a little. So was my “calories out” even higher than that?

You tell me. But clearly I just lost a slight amount of weight consuming nearly 4,000kcal/day. Does that sound like “it’s all calories?”

That said, apparently you also can’t flush your system with infinite fatty acids and expect to still lose a large amount of body fat.

Insulin is necessary for fat gain

This whole experience seems to validate the insulin hypothesis. I can apparently consume about 4,000kcal/day on ex150 and not gain any weight, even slightly lose weight.

But when I up the protein, I gain insane amounts of fat. In the 10 day protein refeed I did in June, I gained 7lbs of fat. Not total weight, that’s what stuck after the water weight came back down. Gaining 0.7lb of fat per day is pretty nuts.

Ben Bikman (a keto researcher) has said that you cannot physically gain fat in absence of insulin. I think this experiment strongly supports that view.

What happens to the “calories” instead? You tell me, bud. I’m just presenting the facts.

What’s next?

First, I’m just super happy that I’m back at 235lbs, my record-since-2017 that I previously hit in early June.

Next up, I’m planning 14 days of ex150collagen - I’ll be supplementing with 40g of collagen protein per day.

Why collagen? It has a different amino acid profile to whey or muscle meat, and is supposed to be less insulinogenic. That also means it’s not a complete protein - if you eat only collagen as your protein, you will die. Luckily I’m keeping the 150g of beef from regular ex150, so I should be fine.

I had previously tried to slightly increase my protein, and had seen no weight loss. But I now know that this experiment was confounded by a pretty severe change in caloric intake, as I had already switched to breve lattes.

So the collagen is another attempt at getting to more reasonable/adequate protein levels, so I can hopefully turn ex150 from an intervention-style diet that requires refeeds, into something more sustainable.

Wish me luck.