"As ethics researchers, we had been running experiments examining various unethical behaviors, such as lying, stealing, and cheating," researchers Maryam Kouchaki of Harvard University and Isaac Smith of the University of Utah's David Eccles School of Business explain. "We noticed that experiments conducted in the morning seemed to systematically result in lower instances of unethical behavior."

This led the researchers to wonder: Is it easier to resist opportunities to lie, cheat, steal, and engage in other unethical behavior in the morning than in the afternoon?

Knowing that self-control can be depleted from a lack of rest and from making repeated decisions, Kouchacki and Smith wanted to examine whether normal activities during the day would be enough to deplete self-control and increase dishonest behavior.

In two experiments, college-age participants were shown various patterns of dots on a computer. For each pattern, they were asked to identify whether more dots were displayed on the left or right side of the screen. Importantly, participants were not given money for getting correct answers, but were instead given money based on which side of the screen they determined had more dots; they were paid 10 times the amount for selecting the right over the left. Participants therefore had a financial incentive to select the right, even if there were unmistakably more dots on the left, which would be a case of clear cheating.

In line with the hypothesis, participants tested between 8:00 am and 12:00 pm were less likely to cheat than those tested between 12:00 pm and 6:00pm -- a phenomenon the researchers call the "morning morality effect."

They also tested participants' moral awareness in both the morning and afternoon. After presenting them with word fragments such as "_ _RAL" and "E_ _ _ C_ _" the morning participants were more likely to form the words "moral" and "ethical," whereas the afternoon participants tended to form the words "coral" and "effects," lending further support to the morning morality effect.

The researchers found the same pattern of results when they tested a sample of online participants from across the United States. Participants were more likely to send a dishonest message to a virtual partner or to report having solved an unsolvable number-matching problem in the afternoon, compared to the morning.

They also discovered that the extent to which people behave unethically without feeling guilt or distress -- known as moral disengagement -- made a difference in how strong the morning morality effect was. Those participants with a higher propensity to morally disengage were likely to cheat in both the morning and the afternoon. But people who had a lower propensity to morally disengage -- those who might be expected to be more ethical in general -- were honest in the morning, but less so in the afternoon.

"Unfortunately, the most honest people, such as those less likely to morally disengage, may be the most susceptible to the negative consequences associated with the morning morality effect," the researchers write. "Our findings suggest that mere time of day can lead to a systematic failure of good people to act morally."

Kouchacki, a post-doctoral research fellow at Harvard University's Edmond J. Safra Center for Ethics, completed her doctoral studies at the University of Utah, where Smith is a current doctoral student. They note that their research results could have implications for organizations or businesses trying to reduce unethical behavior.

"For instance, organizations may need to be more vigilant about combating the unethical behavior of customers or employees in the afternoon than in the morning," the researchers explain. "Whether you are personally trying to manage your own temptations, or you are a parent, teacher, or leader worried about the unethical behavior of others, our research suggests that it can be important to take something as seemingly mundane as the time of day into account."

The University of Rochester study was published today in the Journal of Head Trauma Rehabilitation. If confirmed in subsequent research, the findings could alter the treatment and prognosis of women who suffer head injuries from sports, falls, car accidents or combat.

Several recent studies have confirmed what women and their physicians anecdotally have known for years: Women experience greater cognitive decline, poorer reaction times, more headaches, extended periods of depression, longer hospital stays and delayed return-to-work compared to men following head injury. Such results are particularly pronounced in women of childbearing age; girls who have not started their period and post-menopausal women have outcomes similar to men.

Few studies have explored why such differences occur, but senior author Jeffrey J. Bazarian, M.D., M.P.H. says it stands to reason that sex hormones such as estrogen and progesterone, which are highest in women of childbearing age, may play a role.

"I don't think doctors consider menstrual history when evaluating a patient after a concussion, but maybe we should," noted Bazarian, associate professor of Emergency Medicine at the University of Rochester School of Medicine and Dentistry who treats patients and conducts research on traumatic brain injury and long-term outcomes among athletes. "By taking into account the stage of their cycle at the time of injury we could better identify patients who might need more aggressive monitoring or treatment. It would also allow us to counsel women that they're more -- or less -- likely to feel poorly because of their menstrual phase."

Although media coverage tends to focus on concussions in male professional athletes, studies suggest that women have a higher incidence of head injuries than men playing sports with similar rules, such as ice hockey, soccer and basketball. Bazarian estimates that 70 percent of the patients he treats in the URMC Sport Concussion Clinic are young women. He believes the number is so high because they often need more follow-up care. In his experience, soccer is the most common sport leading to head injuries in women, but lacrosse, field hockey, cheerleading, volleyball and basketball can lead to injuries as well.

Sex hormone levels often change after a head injury, as women who have suffered a concussion and subsequently missed one or more periods can attest. According to Kathleen M. Hoeger, M.D., M.P.H., study co-author and professor of Obstetrics and Gynecology at the University of Rochester School of Medicine and Dentistry, any stressful event, like a hit to the head, can shut down the pituitary gland in the brain, which is the body's hormone generator. If the pituitary doesn't work, the level of estrogen and progesterone would drop quickly.

According to Bazarian, progesterone is known to have a calming effect on the brain and on mood. Knowing this, his team came up with the "withdrawal hypothesis": If a woman suffers a concussion in the premenstrual phase when progesterone levels are naturally high, an abrupt drop in progesterone after injury produces a kind of withdrawal which either contributes to or worsens post concussive symptoms like headache, nausea, dizziness and trouble concentrating. This may be why women recover differently than men, who have low pre-injury levels of the hormone.

Hoeger and Bazarian tested their theory by recruiting144 women ages 18 to 60 who arrived within four hours of a head hit at five emergency departments in upstate New York and one in Pennsylvania. Participants gave blood within six hours of injury and progesterone level determined the menstrual cycle phase at the time of injury. Based on the results, participants fell into three groups: 37 in the premenstrual/high progesterone group; 72 in the low progesterone group (progesterone is low in the two weeks directly after a period); and 35 in the birth control group based on self-reported use.

One month later, women in the premenstrual/high progesterone group were twice as likely to score in a worse percentile on standardized tests that measure concussion recovery and quality of life -- as defined by mobility, self-care, usual activity, pain and emotional health -- compared to women in the low progesterone group. Women in the premenstrual/high progesterone group also scored the lowest (average 65) on a health rating scale that went from 0, being the worst health imaginable, to 100, being the best. Women in the birth control group had the highest scores (average 77).

"If you get hit when progesterone is high and you experience a steep drop in the hormone, this is what makes you feel lousy and causes symptoms to linger," said Bazarian. "But, if you are injured when progesterone is already low, a hit to the head can't lower it any further, so there is less change in the way you feel."

The team suspected that women taking birth control pills, which contain synthetic hormones that mimic the action of progesterone, would have similar outcomes to women injured in the low progesterone phase of their cycle. As expected, there was no clear difference between these groups, as women taking birth control pills have a constant stream of sex hormones and don't experience a drop following a head hit, so long as they continue to take the pill.

"Women who are very athletic get several benefits from the pill; it protects their bones and keeps their periods predictable," noted Hoeger. "If larger studies confirm our data, this could be one more way in which the pill is helpful in athletic women, especially women who participate in sports like soccer that present lots of opportunities for head injuries."

In addition to determining menstrual cycle phase at the time of injury, Bazarian plans to scrutinize a woman's cycles after injury to make sure they are not disrupted. If they are, the woman should make an appointment with her gynecologist to discuss the change.