58 Food Choice
stressed (Willenbring, Levine, and Morley, 1986), and a similar decrease
in eating has been reported in Marines during combat situations (Popper
et al., 1989). One longitudinal study collected daily records of stress and
eating for 84 days from 158 subjects, and reported that eating less was the
predominant response to stress (Stone and Brownell, 1994). In contrast,
studies have also shown the reverse relationship. For example, animal studies
using interventions such as tail pinching, handling, and social disruption
to elicit stress have reported an increase in food intake (Morley, Levine,
and Rowland, 1983; Meisel et al., 1990), and a similar increase has been
reported in obese people (Pine, 1985). Further, in a study of students’ per-
ceptions of the relationship between stress and snacking, 73 percent of respon-
dents reported that stress increased their snacking while decreasing their
consumption of “meal type” foods (Oliver and Wardle, 1999). In addition,
a naturalistic study of work stress concluded that high-workload periods
were associated with greater intakes of energy, saturated fats, and sugar
(Wardle et al., 2000). Likewise, stress is considered to elicit eating as a
coping response (Spillman, 1990), to be a possible trigger for binge eating
(Herman and Polivy, 1975), and to be a precipitant of weight gain in both
adults and children (Bradley, 1985; Mellbin and Vuille, 1989).
At times, therefore, stress decreases food intake, while at others it
causes an increase. This contradictory relationship between stress and
eating has been called the “stress eating paradox” by Stone and Brownell
(1994). Greeno and Wing (1994) proposed two hypotheses to explain
this paradox: first, the general effect model, which predicts that stress
changes food intake generally, and second, the individual difference model,
which predicts that stress only causes changes in eating in vulnerable groups
of individuals.
More recent research has focused on the individual difference model,
and has examined whether the variability in the response to stress relates
to aspects of the individual. Some research has focused on the role of
gender. For example, Michaud et al. (1990) reported that exam stress was
related to an increase in eating in girls but not boys, and similar results
were reported by Rosenfield and Stevenson (1988). In line with this,
Zellner, Saito, and Gonzalez (2007) found that men in their stress group
ate less unhealthy snacks than men in the no-stress group. In contrast, Stone
and Brownell (1994) concluded from their longitudinal study that even
though men were more likely to eat less than more when they were under
stress, women were even more likely to show a reduction in their food intake,
particularly when stress levels were at their highest.