Why Is There Variability in Sleep Study Results?

Sleep Position, Sleep Stages, Alcohol, and Other Factors May Play a Role


If you have had more than one sleep study, with differing results, you might wonder: Why is there variability in sleep study results? Learn about some of the factors that might contribute to these differences and changes over time.

In general, sleep studies are most commonly done to assess for obstructive sleep apnea. One of the most important measurements is the apnea-hypopnea index (AHI), which indicates the number of times per hour on average that your airway collapses during sleep, causing either a drop in the oxygen levels of the blood or an awakening from sleep.

This is sometimes reported as the respiratory disturbance index (RDI). These measures are used to determine your severity of sleep apnea, based on the following categories: normal (<5), mild (5-15), moderate (15-30), and severe (>30). These indicators often vary night to night, so why is this?

First, it should be admitted that even if you had 10 nights of sleep studies in a row, you would have 10 different sets of numbers. This is due to factors that may impact sleep apnea that can change quite quickly. For example, sleep position commonly impacts both snoring and sleep apnea. If you spend a night sleeping more on your back, your AHI may worsen as gravity shifts your tongue and blocks the airway more frequently. Spend a little more time on your sides one night and the number may improve.

Another variable is the amount of REM sleep that occurs during a night of sleep. In REM sleep the body’s muscles are relaxed to prevent dream enactment.

During this sleep state, the airway can become more collapsible and worsen sleep apnea. It is normal for the amount of REM to vary somewhat from night to night and this will also affect the average AHI.

There are a few other factors that could vary over the short term. If you drink alcohol, the amount consumed before bedtime may have a role.

Alcohol acts as a muscle relaxant, contributing to airway relaxation and worsening snoring and sleep apnea. If you have nasal congestion from a cold or allergies, this will decrease airflow through the nose, may lead to unstable mouth breathing, and worsen your AHI.

Time can greatly affect your degree of sleep apnea, and comparing studies separated by months or years may identify dramatic changes. Simple aging leads to a loss of muscle tone throughout the body, including along the airway. Women who enter menopause will have more than 10 times the risk of sleep apnea due to hormonal changes. Weight gain in both women and men can crowd the airway and this can worsen sleep apnea, too. In general, a 10% change (up or down) in body weight is likely to impact your degree of sleep apnea.

Other interventions may also affect your sleep study results. For example, if you have undergone surgery to treat sleep apnea, this may affect your condition (hopefully in a favorable way). If you have a test performed with an oral appliance in place, this likewise may change your condition.

These treatments may alter your anatomy and impact the measures used to assess sleep apnea.

Finally, it is important compare apples to apples and oranges to oranges. This means that you should only compare like studies to one another. Formal diagnostic polysomnograms done in a sleep laboratory should not be compared to home sleep tests as the measurements and procedures differ significantly. In addition, testing centers and even sleep professionals may differ somewhat in how strictly they apply the study scoring rules. As a result, the interpretation of the data and the ultimate results may vary slightly.

Sleep study results will be least variable if as many of these conditions can be kept as constant as possible. Many of these subtle variations will not result in a different diagnosis. If you are concerned about the results of your study, or what it might mean for your treatment options, speak with your sleep specialist.


Kryger, M.H. et al. "Principles and Practice of Sleep Medicine." ExpertConsult, 5th edition, 2011.

Continue Reading