Response to “Sleep Duration and Hypertension With Special Emphasis on Gender and Obesity”

by : DESY SETIAWATI S.P (14010059)

American Journal of Hypertension 26(11) November 2013 1363

Response to “Sleep Duration and Hypertension With Special
Emphasis on Gender and Obesity”

James E. Gangwisch1
To the Editor: I thank Dr Kawada for
raising interesting points related to
our study on sleep duration and risk
for hypertension in women, “Sleep
Duration and Hypertension With
Special Emphasis on Gender and
Obesity.” In analyses stratified by age,
we found sleep durations of ≤5 hours to
be associated with a higher prevalence
of hypertension in all age groups and
with higher incidence of hypertension
only in the younger age group (aged
<50 years). As Dr Kawada points out, in
an analysis of data from the Whitehall
II Study, Cappuccio and colleagues1
found sleep durations of ≤5 hours to be
associated with a significantly higher
prevalence of hypertension in women
but not with incidence of hypertension.
Our differing findings in regards to
the incidence of hypertension could be
explained by differences in the analyses
conducted. Cappuccio and colleagues
did not conduct analyses stratified by
age and the average age of their sample
was >50 years. The inclusion of older
adults in their analyses could have lowered
the odds ratio. The sample size in
the Whitehall II analysis was also much
smaller than that of the Nurses’ Health
Study,2 resulting in less statistical power.
We explored whether obesity, diabetes,
and hypercholesterolemia could
have acted as partial mediators in the
relationship between sleep duration
and hypertension by seeing whether
the inclusion of these variables in the
multivariate models sufficiently attenuated
the associations. Statistical analyses
alone cannot distinguish between a
confounding variable and a mediating
variable. However, results from experimental
studies showing that insufficient
sleep compromises insulin sensitivity
and increases hunger provide plausible
mechanisms for the potential for
mediation. Although sleep duration
has been shown to be associated with
diabetes incidence in previous analyses
from the Nurses’ Health Study and
diabetes was associated with increased
risk for hypertension prevalence and
incidence (Table 1) in the present study,
our results were not consistent with
diabetes acting as a partial mediator of
the relationships between sleep duration
and hypertension. Our results were
consistent with obesity acting as a partial
mediator. Sleep duration has been
shown to be associated with weight
gain in data from the Nurses’ Health
Study;3 in our analysis, body mass
index was associated with hypertension
prevalence and incidence (Table 1). As
Dr Kawada points out, we were not
able to control for the presence of sleep
apnea in our analyses; however, we
were able to control for snoring and
body weight, which are closely related
to sleep disordered breathing. Our
results lend support to the hypothesis
that inadequate sleep plays a role in the
etiology of hypertension in some young
women and that obesity acts as a partial
mediator of this relationship.
Funding
This work was supported by National
Institutes of Health grant HL091443
from the National Heart, Lung, and
Blood Institute.
Disclosure
The author declared no conflict of
interest.
References
1. Cappuccio FP, Stranges S, Kandala NB,
Miller MA, Taggart FM, Kumari M, Ferrie
JE, Shipley MJ, Brunner EJ, Marmot MG.
Gender-specific associations of short sleep
duration with prevalent and incident hypertension—
the Whitehall II Study. Hypertension
2007; 50:693–700.
2. Ayas NT, White DP, Al-Delaimy WK, Manson
JE, Stampfer MJ, Speizer FE, Patel S, Hu FB. A
prospective study of self-reported sleep duration
and incident diabetes in women. Diabetes
Care 2003; 26:380–384.
3. Patel SR, Malhotra A, White DP, Gottlieb DJ,
Hu FB. Association between reduced sleep
and weight gain in women. Am J Epidemiol
2006; 164:947–954.
Correspondence: James E. Gangwisch (jeg64@
columbia.edu).
1Columbia University, College of Physicians and
Surgeons, Department of Psychiatry, New York,
NY.
Initially submitted July 17, 2013; date of first
revision July 25, 2013; accepted for publication
July 26, 2013.
© American Journal of Hypertension, Ltd 2013.
All rights reserved. For Permissions, please
email: journals.permissions@oup.com
doi:10.1093/ajh/hpt157
1364 American Journal of Hypertension 26(11) November 2013
Response to Letter to Editor on Sleep and Hypertension
Table 1. Multivariate analyses using logistic and Cox proportional hazards models to explore the relationship between diabetes and BMI and
hypertension.
Diabetes and BMI
NHS-II, 2001, aged
37–54 years
NHS-I, 1986, aged
40–65 years
NHS-I, 2000, aged
54–79 years
Prevalent hypertension at baseline, odds ratio (95% CI)
Diabetes 5.65 (5.17–6.17) 3.98 (3.68–4.30) 3.36 (3.16–3.59)
BMI, kg/m2
<20 0.60 (0.53–0.68) 0.74 (0.69–0.81) 0.75 (0.70–0.81)
20–24.9 1.00 (Referent) 1.00 (Referent) 1.00 (Referent)
25–29.9 2.08 (1.98–2.19) 1.91 (1.83–1.99) 1.66 (1.60–1.73)
30–34.9 3.88 (3.66–4.11) 3.13 (2.92–3.36) 2.54 (2.38–2.71)
≥ 35 7.19 (6.76–7.65) 4.74 (4.47–5.04) 3.63 (3.43–3.85)
Incident hypertension over follow-up, hazard ratio (95% CI)
Diabetes 1.40 (1.26–1.57) 1.19 (1.04–1.35) 1.19 (1.08–1.31)
BMI, kg/m2
< 20 0.60 (0.52–0.68) 0.70 (0.61–0.79) 0.80 (0.73–0.88)
20–24.9 1.00 (Referent) 1.00 (Referent) 1.00 (Referent)
25–29.9 1.69 (1.60–1.78) 1.65 (1.56–1.74) 1.29 (1.23–1.36)
30–34.9 2.48 (2.34–2.64) 2.24 (2.04–2.46) 1.70 (1.56–1.85)
≥ 35 3.28 (3.06–3.52) 2.69 (2.48–2.92) 1.73 (1.60–1.88)
Controlling for age, race, Hispanic ethnicity, menopause, smoking, physical activity, alcohol, caffeine, Dietary Approaches to Stop Hypertension
diet, aspirin, acetaminophen, nonaspirin nonsteroidal antiinflammatory drugs, family history of hypertension, snoring, and shift work.
Abbreviations: BMI, body mass index; CI, confidence interval.