Journal of Correctional Health Care - April 2023 - 130

130
DAVIS ET AL.
chronic health mental health diagnoses or medications (Bai
et al., 2015; Clarke & Waring, 2012; Drach et al., 2016;
Khavajou et al., 2007; Massie et al., 2000; Wolff et al.,
2012) in relation to weight during incarceration. Others
looked at how factors such as low activity levels (Baldwin
et al., 2016; Clarke & Waring, 2012; Massie, 2000), commissary
intake (Clarke & Waring, 2012), and excess
calories provided in diets (Massie, 2000) impact weight.
Pooled Prevalence of Obesity in Incarcerated
Populations Included in Meta-Analysis
Pooled OB prevalence was calculated for males and
females using all 11 studies (Figs. 2 and 3). Among the
10 studies used to pool OB prevalence for females, the
overall estimate of the proportion of OB females was
39.8% (95% CI: 37.3-42.4). The proportion of total variation
across studies due to heterogeneity was moderate
at 39.9% (Q= 15.0, p= .092). Among the five studies
used to pool OB prevalence for males, the overall estimate
of the proportion of OB was 30.0% (95% CI:
25.9-34.5). The proportion of total variation due to
heterogeneity was high at 84.5% (Q= 25.8, p< .0001).
The estimated proportion of incarcerated females with
OB (39.8%) closely mirrors National Health and Nutrition
Examination Survey (NHANES) data for 2017-
2018 that found an age-adjusted prevalence of obesity
in adults of 42.4%, with no significant differences
between men and women among all adults or by age
group (Hales et al., 2020). The estimated proportion of
incarcerated males with OB at 30% was lower than the
NHANES 2017-2018 age-adjusted prevalence of obesity
for American males (42.4%; Hales et al., 2020).
Study Quality
Study quality was rated using Cochrane's GRADE
assessment (Ryan & Hill, 2016). Consideration of sampling
methods and study design helped assess risk of
bias with particular attention to unique dynamics of
people in prison and how BMI data were collected. Additional
criteria (i.e., directness, precision, and publication
bias) assessed outcomes, effect sizes, and relevance of
results and interventions to ensure quality of metaanalytic
evidence. Other factors including effect and
dose response were assessed further only if a study demonstrated
threatened validity. For example, studies with
self-reported BMI data are limited.
Each study was discussed and assessed as either very
low, low, or moderate quality by two researchers. None
of the reviewed studies met GRADE criteria for high quality.
GRADE assessment ratings are included in Table 2.
Discussion
This meta-analysis summarized existing evidence regarding
OB and weight change among incarcerated males
and females in the United States. Data regarding OB
and weight change came from various sources: state
prison databases, measurements taken on facility entry,
measurements obtained during study data collection
period, and self-reported data. Of the 11 included studies,
4 measured weight gain during a period of incarceration,
and the weight gains noted were not insignificant.
Our pooled prevalence estimates demonstrate that the
proportion of incarcerated females with OB mirrors
the age-adjusted prevalence of OB among females in
the United States. The estimated proportion of incarcerated
men with OB was significantly lower than the
2017-2018 NHANES age-adjusted prevalence of obesity
for males (Hales et al., 2020). Significant weight gain
during incarceration was well demonstrated for females
in several small studies; there is less robust data to support
weight gain during incarceration in men.
Our findings are congruent with two recently published
meta-analyses looking at weight gain among incarcerated
people. Gebremariam et al. (2018) reviewed weight
gain in prisons in high-income countries worldwide,
and their meta-analysis demonstrated an average gain
of 0.43 pounds per week. They noted that incarceration
was associated with weight gain, especially among
females. Choudhry et al. (2018) focused on incarcerated
men worldwide. They concluded that although prevalence
of OB ranged widely (8.1% to 55.6%), the rates
were similar to or lower than the general population of
the country the prison resides in.
Our study looks specifically at people incarcerated
in U.S. prisons and includes males and females, filling
a gap in current research and contributing data with public
health implications.
Stressful conditions, which are a continual part of daily
prison life, may affect weight gain (Fogel, 1993; Houle,
2014). The mechanisms for weight gain vary, but include
increased circulating levels of cortisol (Cohen et al.,
2006) that can lead to visceral OB (Tchernof & Despre´s,
2013). Stress increases human motivation for foods
higher in carbohydrates and saturated fats, which may
increase caloric intake (Boggiano et al., 2017).
Finally, correctional facilities structure all aspects of
a person's daily schedule, including time available for
physical activity (Houle, 2014). It is largely unknown
whether incarcerated people engage in enough physical
activity (Herbert et al., 2012; Houle, 2014). A Lancet
review article noted no data were available regarding
physical activity in U.S. prisons (Herbert et al.,
2012).
Multiple authors have expressed concern that lack of
exercise may be contributing to NCD burden in prisons
worldwide (Binswanger et al., 2009; Clarke & Waring,
2012; Elwood Martin et al., 2013; Johnson et al.,
2019). There is a high likelihood that incarcerated people
do not have enough structured time for physical
Journal of Correctional Health Care - April 2023

Table of Contents for the Digital Edition of Journal of Correctional Health Care - April 2023
