Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I have the same issue, it's truly unfortunate. What's odd is I forget about what caffeine does to my sleep and after a few weeks/months of drinking it, I'm wondering why I'm so stressed, tired and can't get ANY sleep.

I stop drinking coffee and BAM, I sleep like a baby. It doesn't MATTER when I drink it, I can drink it at 6AM and I will not have a good deep sleep. I am unsure if this is coincidence, but I also notice I remember way less dreams when I am on caffeine than not. I also find it's a compounding effect which is why it's slightly annoying.

If I drink 1 cup, in 2 weeks, my sleep will be fine so I will think, okay, it's not the caffeine. Then I will continue drinking it for weeks and suddenly I haven't had a good nights rest in weeks and I'm wondering what is going on. Not having deep sleep for weeks really has a big impact on your stress levels, memory, emotional well being and general energy levels.

The annoying part is coffee is so good for productivity so I go through cycles (also you start to think it's the stress not the caffeine that's causing the sleep issue!)

Weeks of stressful work - drink more caffeine to get all the work done - bad sleep, bad mood, bad energy levels, aka all the negative affects from not having enough deep sleep.

Weeks of less stressful work, no caffeine, great sleep, great mood/energy levels, etc.

I've always convinced myself that not drinking caffeine for deep sleep is just placebo, but I've tested it so many times that it just can't be.

Is there a way to test if you're a slower metabolizer? I know my partner can drink 3 cups and she is totally fine, lucky her! I'm 100% convinced I am, but it would be cool to test by some sort of blood/urine test?



You'll have to do a genetic test. Nebula, 23andme, etc.

Then look at CYP1A2, see if you have the C/C genotype.

Definitely read the Genetic Life Hacks article that I linked to above.


Stay away from 23andme if you have any privacy concerns. I've worked with providers of DNA insights and advice that don't build their revenue model on selling your data. For example DNAPal.me


They do not sell your data; Facebook and Google also don't sell your data.

Your DNA is worthless[0] and impossible to hide. If someone did want your DNA there is nothing you could do to stop them. You leave it everywhere you go.

[0] except to your children you don't know you have


did you forget /s

They sell it to GlaxoSmithKline and others:

https://en.wikipedia.org/wiki/23andMe

not to mention the recent data breach


Yeah, but why not support companies doing "the right thing" and nudge the trend towards companies that respect and preserve the privacy interests of their customers.


They do respect it. You have to sign up for research and fill out a thousand surveys before any of that data is used for… research.


So why's everyone making such a fuss about 23&me this week?


Journalists are kind of bad at understanding how things actually work.


I really wish there were more outlets like how for the legal/news junkie circuit we have things like MeidasTouch or whatever its called and the other YouTuber journalists (they deserve that umofficial title because what they do is top shelf journalism or at least investigative YouTubing.


CYP1A2 is the whole gene- you need to look at marker rs762551 within CYP1A2. Both the C/C and A/C genotype are slow caffeine metabolizers. The most common genotype is A/A, which is a fast metabolizer.

https://www.ncbi.nlm.nih.gov/snp/rs762551


Do you have a link to one that shows A/C? I'm only seeing C/C (I'm A/C so curious). Thanks!


https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2014233/

This seems to be the original paper. I don't find this paper particularly meaningful, but the effects they did observe showed A/C and C/C to be about the same, and both different than A/A.

This is based on a 5 hour after caffeine ingestion blood test in smokers. They found no differences in the non-smokers, but those were urine tests taken at variable times (whenever they peed), which seems sketchy to me.

Based on this study, subsequent studies (e.g. https://pubmed.ncbi.nlm.nih.gov/16522833/) seem to group A/C and C/C together, and not look at them independently. C/C is rare enough that studies have trouble getting many individuals from that group.


So if C/C then slow metabolizer?


Correct. C/C is the genotype.

Marker rs762551, as another poster noted above.

For 23andme customers: https://you.23andme.com/tools/data/?query=rs762551


Thank you for sharing your link.

Groan, I'm A/C!!


That's great - this all came in time for me to run it by my doc today during an annual physical.


This sounds like the same cycle I go through, especially the forgetting part. I've found that cycling caffeine through the week (taking a break on weekends) and just not having too much even during the week can help maintain the productivity, but it also means spending my weekends in a tired daze. I think I just need to commit to not having any caffeine, or if I do, only taking it temporarily before stopping again.


So if I can fall asleep after moderate amounts (like, right away after intake), this prolly doesn't affect me, no?


Fascinating. What are your deep sleep totals on and off caffeine?




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: