Research Question:

What is the effect of adding a 60‑second cold rinse at the end of a hot morning shower on resting heart rate measured ten minutes later in healthy adults?

Rationale and Background:

I have always loved taking a cold shower in the mornings. As an athlete most of my life, I found cold showers to be a productive way to start the day and get me going, but also a good form of recovery after a brutal workout. Even on rest days the quick shock wakes me up better than coffee. Because of that lived experience, I naturally wanted to dig deeper into what shower temperature really does to the body. That curiosity pushed me to pick this study and frame my own question.

In the original paper, a RCT of 3,018 Dutch adults was conducted where they would end a daily hot shower with 30-90s of 10-12°C water for 30 days. This was found to lead to 29% fewer sickness‑absence days during the 90 day follow‑up, while the total number of self‑reported illness days remained unchanged (Buijze et al., The Effect of Cold Showering on Health and Work). This shows us that brief cold rinses helped participants keep working through mild ailments rather than preventing illness itself. Essentially, it helped them power through on days where they weren’t feeling their best, but in the case of actual illness a cold shower didn’t provide significant help.

That drop in sick-leave is interesting, but the study never measured any solid physiology. Resting heart rate (HR) is an easy, smartwatch‑friendly marker that tracks how active the “rest-and-digest” side of our nervous system is. Lower HR usually means more parasympathetic activity, which lines up with feeling relaxed and ready to go again. Cold water seems to help here too. For example, in a 2012 study elite swimmers sat in 15°C water for five minutes after practice. The very next morning their heart‑rate variability was higher, showing a clear boost in parasympathetic tone and better recovery (Al Haddad et al., Effect of daily cold water immersion on heart rate variability and subjective ratings of well-being in highly trained swimmers). https://pubmed.ncbi.nlm.nih.gov/21941017/

This study aims to determine whether a brief, 60-second cold-water rinse appended to a conventional hot shower produces a measurable reduction in resting heart rate 10 minutes post-exposure in healthy adults. Demonstrating such an effect would provide an inexpensive, easily adoptable intervention with potential implications for occupational readiness and post-exercise recovery.

Hypotheses:

Null hypothesis:

Adding a 60‑second cold rinse to the end of a hot morning shower does not change resting heart rate measured ten minutes later compared with showering hot only.

Alternative hypothesis:

Adding a 60‑second cold rinse to the end of a hot morning shower lowers resting heart rate measured ten minutes later compared with showering hot only.

Experimental Design:

This is a short, manipulative experiment that compares resting heart rate (HR) after two shower conditions: (1) a normal hot shower only (control) and (2) the same hot shower immediately followed by a 60‑second cold rinse at 10-12°C (treatment). Each participant completes both conditions on separate mornings, so everyone acts as their own control. The order of conditions is randomized to cancel out day‑to‑day variation.

Shower setup: Participants shower at home using their usual routine (water = around 38-40°C, 5 min minimum). At the end, they switch the faucet to full‑cold (verified at 10-12°C) for exactly 60s, starting a phone timer as soon as the water hits their shoulders. Identical hot shower is taken by the control group but they do not switch to cold; they stand under hot water for an extra 60s instead, keeping total shower time equal.

Measuring DV: Resting HR is recorded 10 min after stepping out of the shower. Participants dry off and start a wearable HR monitor (Apple Watch) linked to a logging app. Once seated, they rest hands‑on‑lap, breathe normally, and record a 60s average HR. This single number is the outcome for that trial.

Because we actively change water temperature, this is a manipulative experiment.

Ways to Reduce Error: Avoid staying up late and get an adequate amount of rest every night, this will help with any inconsistencies that come from fatigue. Additionally ask participants to avoid caffeine, heavy meals, and vigorous activity for 2h before each trial. This will help with outliers as well as help with energy/HR deviations due to outside sources. To reduce recording bias, participants will label each trial only as “A” or “B” in the HR app– the analyst gets the key after data import. This keeps the person crunching numbers blinded to condition. Also, each person provides both a control and a treatment value, removing between‑person variability (blinding). Each person also serves as their own control by taking a normal hot shower on one morning (no cold rinse). Lastly, we can incorporate randomizing in this way: flip a coin for every participant to decide whether they try the cold‑rinse morning first or the hot‑only morning first. This kills any morning bias– maybe you slept badly the first night or had an exam the second day. Random order evens that out.

Data Analysis Plan:

For this experiment, I plan to run a two‑tailed paired‑sample t‑test on resting heart rate values. Each participant supplies two numbers: (hot only) and (cold rinse) so the test factors in the natural difference in everyone’s heart rate and checks if the average difference is zero. Heart rate is continuous, usually close to normal, and the paired design helps to eliminate noise between subjects. Considering all of this, I think that this test will work most efficiently for our sample of 30 people.

Assumptions and Exploratory Data Analysis (EDA):

  • Independence: Each cold and hot value comes from the same person, and one person’s pair has nothing to do with anyone else’s.
  • Normality of the paired differences: the set of heart rate differences should look roughly bell‑shaped.
  • No extreme outliers: single wild points can wreck the mean and the t‑test. I doubt any extreme outliers would be present in this set of data because we are dealing with heart rate in healthy adults which shouldn’t deviate too much.
library(tidyverse)
## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr     1.1.4     ✔ readr     2.1.5
## ✔ forcats   1.0.0     ✔ stringr   1.5.1
## ✔ ggplot2   3.5.1     ✔ tibble    3.2.1
## ✔ lubridate 1.9.4     ✔ tidyr     1.3.1
## ✔ purrr     1.0.4     
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors
data_raw <- read_csv ("aawasthi.csv")
## New names:
## Rows: 30 Columns: 3
## ── Column specification
## ──────────────────────────────────────────────────────── Delimiter: "," dbl
## (3): ...1, Level 2: Hot to Cold, Level 1: Hot Only
## ℹ Use `spec()` to retrieve the full column specification for this data. ℹ
## Specify the column types or set `show_col_types = FALSE` to quiet this message.
## • `` -> `...1`
data <- data_raw |>
  select(`Level 2: Hot to Cold`, `Level 1: Hot Only`) |>
  rename(cold = `Level 2: Hot to Cold`, #renamed columns for easier and faster coding
         hot  = `Level 1: Hot Only`)

data <- data |>
  mutate(diff = cold - hot) #made column for the differences
head(data)
data_clean <- data |> filter(hot < 200, cold < 200)  #removing clear outlier 

ggplot(data_clean, aes(y = diff)) +
  geom_boxplot() +
  labs(y = "HR difference (cold − hot, bpm)",
       title = "Boxplot of paired HR differences") #boxplot visualization

ggplot(data_clean, aes(x = diff)) +
  geom_histogram(binwidth = 2) +
  labs(x = "HR difference (bpm)",
       title = "Histogram of paired differences") #histogram visualization

Interpretation of EDA:

I started by pulling aawasthi.csv into R and trimming the raw table down to the two columns we care about: Level 2: Hot to Cold and Level 1: Hot Only, then gave them shorter names (cold and hot) so the code reads clean. Creating a third column, diff = cold – hot, instantly shows the paired difference for every participant and keeps all the math in one tidy data frame.

A quick head() confirmed the import worked: six rows with plausible bpm values and the new diff column in place. Before plotting, I ran a sanity check for impossible heart rates. Anything above 200 bpm in a resting, healthy adult is clearly a typo, so I filtered out those rows (filter(hot <200, cold <200)). Only one record vanished, leaving 29 usable pairs.

The boxplot of diff has a compact IQR and no points dangling more than 1.5×IQR from the box—nice evidence that extreme outliers are gone. Most values sit slightly below zero, hinting that heart rate tends to drop after the cold rinse.

The histogram is roughly mound‑shaped with a gentle right tail but no spikes or double peaks. Additionally, just taking a glance at it we can see it looks normal and visually there is nothing wrong with it.

Because the paired differences appear symmetric and we have no crazy outliers, I kept the data on the original bpm scale– no log‑10 transform needed. Using raw beats/min keeps the final t‑test result easy to explain (“HR dropped by X bpm”), which is friendlier for a first‑year bio audience than talking about transformed units.

Overall, the EDA shows:

  • Clean sample of 29 paired observations
  • No remaining impossible values or extreme outliers
  • Distribution of differences looks approximately normal
  • Visual trend points toward lower HR after the cold rinse

These checks clear the way for a paired‑sample t‑test.

Primary Statistical Analysis

ttest <- t.test(data_clean$cold, data_clean$hot,
                    paired      = TRUE,
                    alternative = "two.sided")

ttest  
## 
##  Paired t-test
## 
## data:  data_clean$cold and data_clean$hot
## t = -3.3009, df = 28, p-value = 0.002634
## alternative hypothesis: true mean difference is not equal to 0
## 95 percent confidence interval:
##  -8.946167 -2.094651
## sample estimates:
## mean difference 
##       -5.520409

Data Visualization

vis <- data_clean |>
  mutate(id = row_number()) |>
  pivot_longer(c(hot, cold),
               names_to  = "condition",
               values_to = "hr")

# decided to go with a paired‑line plot
ggplot(vis, aes(x = condition, y = hr, group = id)) +
  geom_line(colour = "grey60") +
  geom_point(size = 3) +
  labs(title = "Resting Heart Rate 10 min After Shower",
       x = NULL,
       y = "Heart rate (bpm)") +
  theme_linedraw(base_size = 12)

Conclusions

The numbers back up what my morning routine has been telling me for years. Paired‑sample t‑test tells a clear story: a 60‑second cold rinse at the end of a normal hot shower lowers resting heart rate ten minutes later. The statistics back it up—t = ‑3.30 with 28 degrees of freedom and a p‑value of 0.0026. Because the p‑value is well below the usual 0.05 cut‑off, we reject the null hypothesis that “nothing changes.” The average paired drop was ‑5.5 beats per minute, and the 95% confidence interval (‑8.95 bpm to ‑2.09 bpm) sits entirely below zero. In other words, every plausible true difference points to a slower heart after the cold rinse. That slower rate suggests a short‑lived boost in parasympathetic activity, matching what cold‑exposure studies often report and offering a possible physiological link to the reduced sick‑leave days Buijze’s study saw in a much larger trial; if the heart is calmer and the autonomic system is tilted toward recovery, it makes sense that people might feel more energetic and resilient at work.

This study is a tidy proof‑of‑concept, yet it carries clear caveats. First, the sample was small, young, and healthy, limiting how far we can generalize. Second, each participant tried the cold rinse only once; we do not know whether the benefit compounds, plateaus, or even backfires with daily use. Third, showers and heart‑rate recordings happened at home, so faucet temperatures, timer accuracy, and posture could all vary. Finally, everyone knew when the water turned icy; even though the physiological change is objective, an “I feel refreshed” mindset could contribute.

The next step is to scale up and tighten control. A larger, multi‑week study—ideally in a lab or with smart‑shower valves that guarantee temperature—could confirm whether the heart‑rate drop sticks around and whether it translates to measurable gains in mood, sleep quality, or work performance. Varying the rinse length (15,30,60,90s) could pinpoint the minimum effective dose. Adding heart‑rate variability, blood pressure, and salivary cortisol would flesh out the autonomic story, while pairing the rinse with post‑workout recovery in athletes could test its usefulness beyond simple morning routines. If those studies echo our findings, the humble cold finish might earn a spot alongside stretching and coffee as a low‑cost tool for daily reset and resilience.

Limitations -Small, homogeneous sample. Our 29 participants were young and healthy. Older adults, children, or people with heart conditions might react differently.

-Single exposure per person. Each volunteer tried one hot‑only and one cold‑rinse shower. We don’t yet know if the heart‑rate drop grows, stays steady, or disappears with daily use.

-Home‑based protocol. Showers happened in personal bathrooms. Water temperature, shower length, and timing were self‑reported, leaving room for measurement error.

-No participant blinding. Everyone felt the cold water, so a “placebo” effect or excitement could influence results, even if the heart‑rate drop is real.

-One physiological marker. We tracked resting heart rate only. Heart‑rate variability, blood pressure, or stress hormones would give a fuller autonomic picture.

Future directions - Longer, larger RCT. Follow hundreds of participants for 4–8 weeks, controlling shower temperature in a lab or with smart valves. Track heart rate, heart‑rate variability, blood pressure, mood, and sleep to see if benefits stick.

  • Exercise recovery study. Test cold rinses immediately after workouts in athletes and record next‑day performance and muscle soreness.

Citations

Buijze,J.C., Veenstra,E., Vermeulen,H.M.W., Loogman,M., Giele,J.L., Ruetgerink,J.J.M., … & Bloem,R.M. (2016). The effect of cold showering on health and work: A randomized controlled trial. PLOS ONE, 11(9), e0161749. https://doi.org/10.1371/journal.pone.0161749

Al Haddad,H., Laursen,P.B., Ahmaidi,S., & Buchheit,M. (2010). Influence of cold water immersion on post‑exercise parasympathetic reactivation. European Journal of Applied Physiology, 109(7), 1189–1197. https://doi.org/10.1007/s00421‑010‑1459‑9

Disclaimer: This project analyzes simulated data. The questions and hypotheses are real, but the results and conclusions are simulated.