Relationship BetweenCaffeine Consumption and Sleep and Sleepiness of Shift Workers

Danielle Lemieux


A_5odie.gif (12220 bytes)



This study was designed to sdemonstrate the possible effects of caffeine on both sleep and sleepiness at work. The shift workers who consume high doses of caffeine should experience less quality sleep, difficulty falling asleep, and less overall sleep than their co-workers.This study included 55 participants employed in a mill in Maine. (for more about this company, see ) Those in the study worked shift work. They completed a survey which included a caffeine inventory, The Stanford Sleepiness Scale, Epworth Sleepiness Scale, and questions regarding sleep habits. It was found that the caffeine consumers exhibited less self reported fatigue than their co-workers consuming no caffeine, this was demonstrated by three separate means. The caffeine consumers reportedly slept for a lesser amount of time than non caffeine consumers. Finally, the caffeine consumers had more trouble falling asleep than their co-workers. Other studies, such as Bruce and Lader (1986), suggest findings such as this could lead to use of low to moderate amounts of caffeine to control sleepiness and facilitate worker productivity, especially for night shift workers. Caution should be taken however, as there are negative effects of caffeine that need to be explored more deeply concerning both use and withdrawal.




The subjects in this study were  white males and females from a middle class background. The participants ranged in age from 22 to 59, the mean being 39.87. there were 55 participants from 3 shifts, (20 people from 6-2,19 from 2-10,& 16 from 10-6). All are currently employed by  a paper plate manufacturing company in  Maine.

See also Circadian Technologies




This study used a food diary (to measure the amount of caffeine consumed in the last 6, 12, and 24 hours and in the last week). It included the Stanford sleepiness scale (SSS) to measure sleepiness of the subjects while at work. The Epworth sleepiness scale (ESS) was also used to measure fatigue. Finally there were a few questions about sleep habits. A combination of these tools was used to assess the relationship between caffeine consumption and the overall sleep habits of the participant.



The questionnaires were administered at three different times, each of the three shifts filled out the questionnaires approximately four hours into (half way through) their shift. This study was conducted on the fifth day of their working that particular shift. They were approached during a 20 min break in the break room (cafeteria). The participants were each asked to fill out the packet of surveys during one of their breaks. They were given a food log which was later use to estimate caffeine intake. This questionnaire was used to divide the workers into two groups, users and nonusers. It was also used to look at differences in caffeine consumption among the three shifts.

Also included were a few questions about their sleep habits and sleepiness, this information was obtained in order to compile data concerning the participants sleep habits and to make comparisons. The questions included how many hours of sleep are normal in a day on the shift they were working; How long it took to fall asleep when in bed; if they could fall asleep at that time and for how long could they sleep. See Appendix C for a copy of these questions. The question was later scored using the following scale.

Question 3 "If you were alone now and had the opportunity, could you fall asleep? If yes, for how long could you sleep?"

5-10 min=1
10-20 min=2
20-30 min=3
30-40 min=4
30 min -1 hour=5
more than one hour=6

They were then asked to fill out the Stanford sleepiness scale to assess their level of alertness. This survey asked the participants to choose one of seven statements that best described their state of alertness at that time. The statements range from 1-feeling active and vital; alert; wide awake to 7-losing struggle to remain awake; sleep onset soon. Refer to Appendix D.

Finally they answered the Epworth Sleepiness Scale in which the participants were asked to rate on a scale of 0-3, the likelihood of falling asleep in different situations (e.g., while watching TV). See appendix E for a sample of this questionnaire. This survey was then scored and the data from these tools compared.




Thisresearcher intended to show a correlation between caffeine consumption and both sleepiness at work and sleep patterns of shift workers.

There were 55 participants  from three shifts, 20 from the first shift, 19 from second shift and 16 from the third. Of these workers, 31 were male and 23 female. All were between the ages of 18 and 65, with the mean being 39.87. Participants reported consuming between 0 and 2400mg of caffeine in the 24hours prior to answering the survey. The first shift consumed the least on average, 356.55mg, second shift consumed 572.32mg, the most of the three shifts, and the third shift consumed an average of 360.6mg per person. See chart below. A one-way analysis of variance(ANOVA) was run on each of the variables comparing the 3 shifts. No significant differences were found among the three groups in regards to caffeine consumption because of large variability.




Differences between shift and means of

caffeine consumption and time to fall asleep



mean time to fall

shift N caff.(mg) asleep (min)*

* indicates significant difference was found between shifts


The second shift had the hardest time falling asleep once in bed. Here, significant differences were found. F(2,52)=3.151,p=.051. When a post hoc comparison was performed, a Bonferroni was used, second shift(2-10) and third shift (10-6) were found to have a difference at .056 significance.





Difference in means of

time to fall asleep

by shift


Data were also compared using gender as the factor for the one-way ANOVA. No significant difference was found between these two groups, when looking at caffeine consumption F(1,52)=.013,p=.909, Stanford Scale F(1,51)=2.387,p=.129, Epworth Scale F(1,51)=1.248,p=.269, sleep F(1,52)=.186,p=.668, time to fall asleep F(1,52)=2.891,p=.095 and napping question (number 3) F(1,52)=3.321,p=.074.

Finally Paired-Sample T-Tests were run on 12 of the participants. Only 6 of the workers did not consume any caffeine in the previous 24 hours, as reported. Participants (n=6) who had consumed caffeine were randomly chosen to be matched with the 6 who had reported no caffeine intake. The two groups differed significantly in many aspects. First, marginal significance was found when looking at sleep onset, the caffeine consumers took longer to fall asleep than non consumers, t(11)=2.090,p=.061. Significance was also found when looking at the amount of sleep per day, those who consumed caffeine got less sleep, t(11)=2.247,p=.046. More sleepiness was found in non-caffeine participants through three different means, Stanford Sleepiness Scale t(11)2.27,p=.044, Epworth Sleepiness Scale t(11)=2.23,p=.048, and a question concerning the ability to nap at that point in time t(11)2.269,p=.044, all three were found to be significant.


The research hypothesis was supported, the caffeine consumers exhibited less fatigue than their co-workers consuming no caffeine, this was demonstrated by three separate means. The caffeine consumers reported sleeping for a lesser amount of time than non caffeine consumers. Finally, the caffeine consumers had more trouble falling asleep than their co-workers.



any questions, please e-mail Danielle @ eyes2.gif (2438 bytes)