Sleep disorders during pregnancy: an underestimated risk factor for gestational diabetes mellitus

Insomnia during pregnancy has a high incidence, reaching up to 61%, and deteriorates with the advance of gestational age. Reported insomnia has been associated with GDM, depression, pre-eclampsia, fetal growth restriction, preterm delivery, prolonged labor, cesarean section, and undiagnosed underlying psychiatric disease [6]. In treating insomnia, efforts should focus on educating pregnant women on sleep hygiene, exercise, and behavioral therapy, while the aid of a psychologist-psychotherapist may be essential. Few pharmaceutical substances, such as the antihistamine diphenhydramine, could be used in pregnancy [7].

The most common sleep-related breathing disorder during pregnancy is obstructive sleep apnea (OSA) or obstructive sleep apnea-hypopnea syndrome (OSAHS). Various questionnaires, such as Berlin Questionnaire (BQ), Epworth Sleepiness Scale (ESS), STOP and STOP-BANG, are used as screening tools; however, none of these has a high sensitivity [8]. Fatigue and daytime sleepiness are the most common clinical features while snoring, awakenings with dyspnea, nocturnal polyuria, and apneas observed by partners are also present. OSA in pregnancy has been associated with various complications, such as GDM, hypertension, pre-eclampsia, premature labor, cardiomyopathy, congestive heart failure, low birth weight, reduced Apgar score, postpartum neonatal intensive care unit (NICU) hospitalization and prolonged hospitalization. However, a common confounder is the coexistence of obesity. In addition to OSA, even simple snoring, the incidence of which increases in pregnancy, has been associated with GDM, hypertension, and small-for-gestational-age (SGA) fetuses [9].

Restless legs syndrome (RLS) is characterized by an unpleasant urge to move one’s legs and affects the onset and maintenance of sleep. Hemoglobin values below 11 g/dl, family history of RLS, and age ≥35 years are some known risk factors. Moreover, RLS has been associated with a higher risk of GDM, poor sleep quality, daytime dysfunction, drowsiness, pre-eclampsia, preterm delivery, low birth weight, and depression. The treatment of choice is non-pharmaceutical, including exercise, massage, and the use of pressure devices. Iron administration to restore the ferritin value to at least 75 mcg/l is considered essential [10].

Narcolepsy is characterized by sudden involuntary sleep episodes. Women with an increased BMI before and those with excessive weight gain during pregnancy face a higher risk of developing narcolepsy type I or narcolepsy with cataplexy, a sudden muscle weakness triggered by laughing, crying or terror. There are no official recommendations for the treatment of narcolepsy with first onset during pregnancy since all the available medications are not authorized for such use [11]. However, a recent consensus by sleep experts concluded that the risk of teratogenic effects from narcolepsy medications is extremely low [12].

Qiu et al. evaluated the association between sleep duration and snoring during pregnancy with the risk of GDM in 1290 pregnant women [13]. After adjustment for maternal age and ethnicity, women who reported a sleep duration of ≤4 h per night at the beginning of pregnancy had a 5.56 times greater risk of developing GDM than women with a sleep duration of 9 h. It also appeared that even an extra hour of sleep reduced the risk for GDM by 15%. Moreover, women who reported sleep duration ≥10 h had higher plasma glucose concentrations, while snoring was related to a 1.86-fold increased risk of GDM. Rawal et al. published a study of 2581 pregnant women evaluating the association of sleep duration in the 1st and 2nd trimesters with GDM development using structured questionnaires [14]. Sleep duration decreased gradually in ≤7 h from the 1st to the 2nd trimester, along with a reduction in daytime sleep intervals. There was no association of sleep duration in the 1st trimester with the onset of GDM. It was observed that non-obese pregnant women who slept more or less than 8–9 h in the 2nd trimester had twice the risk of developing GDM, indicating a U-shaped association between GDM and sleep duration. Followingly, Wang et al. studied sleep duration and quality in 12,506 Chinese women, showing that moderate and poor sleep quality during pregnancy was associated with a higher risk of GDM development (OR 1.62 and 1.77, respectively) [15]. According to two recent studies, GDM prevalence was prominent in women with poor sleep quality and nocturnal sleep duration <6 or 7 h (OR 1.96, 1.32, respectively) [16, 17].

Zhong et al. evaluated the association between poor sleep quality and duration in early pregnancy with GDM in 4066 pregnant women [18]. They filled out a structured questionnaire before the 16th gestational week to assess the duration and quality of morning and nocturnal sleep, with a repeat of the questionnaire at 24–28 weeks, when the OGTT was also performed. Poor sleep quality in early pregnancy combined with long sleep duration (≥8.5 h) was associated with an increased risk of GDM (OR 2.27). Another study, including 48,787 pregnant women, concluded that sleep duration less than 5 and more than 10 h was associated with a higher risk of GDM (OR 1.30 and 1.21, respectively) [19]. Wang et al. studied 1300 women and showed that sleep duration <7 or >10 h and poor sleep quality were associated with a higher risk of GDM (OR 4.28, 4.42 and 1.75, respectively). The combination of poor sleep quality and duration of <7 h was related to a greater risk of GDM (OR 12.71) [20]. Peivandi et al. concluded that women with poor sleep quality had significantly higher BMI levels and a higher risk of developing GDM (OR 2.99) [21]. Furthermore, according to another study, no difference was detected in GDM prevalence in women with poor sleep quality, while short sleep duration was associated with an increased risk for GDM [22]. Song et al. concluded that a positive BQ index increased the risk of GDM development from 27.5% to 66.7%, indicating an interaction of OSA with GDM [23]. Conclusively, sleep duration and risk of GDM present a U-shaped association, indicating that the optimal sleep duration during pregnancy is 8–9 h. The various findings are summarized at Table 1.

Actigraphy is a non-invasive and objective diagnostic method of sleep duration and quality [24]. Twedt et al. showed that shorter sleep duration has been related to worse glycemic control [25]. Nocturnal sleep duration <7 h and later sleep midpoint were associated with statistically significant increased risk for GDM [26]. Finally, in another study short sleep duration was related to impaired glucose tolerance [27]. The findings are summarized at Table 2.

Polysomnography (PSG) is the diagnostic gold standard for the majority of sleep disorders. Reutrakul et al. investigated the relationship between OSA and GDM using PSG [28]. GDM was related to shorter sleep duration, higher microarousal index, 4-fold ΑΗΙ, and increased risk for OSA (OR 6.6) compared to women without GDM. OSA was correlated with increased insulin resistance, pancreatic beta-cell dysfunction, BMI, and FPG values in obese pregnant women with GDM [29]. Chirakalwasan et al. measured the effect of a CPAP device in obese women with GDM and OSA [30]. Insulin sensitivity and secretion were improved in the CPAP subgroup. Alonso–Fernandez et al. evaluated 177 pregnant women with and without GDM in the 3rd trimester of pregnancy. Although women with GDM had greater BMI and increased insulin resistance, OSA prevalence did not differ between the two groups [31].

Bublitz et al. focused on the role of inflammation in OSA and GDM. They conducted a sleep study in 23 women with GDM and detected increased levels of interleukins 6, 8, and TNF-a. The researchers suggested that there is an elevated inflammatory profile in this population [32]. Balserak et al. concluded that OSA prevalence, snoring frequency, and REM sleep duration, were significantly higher in women with GDM [33]. In another study in women with GDM, the elevated ΑΗΙ was related to increased nocturnal glucose levels [34]. Sanapo et al. proved that women with OSA presented increased fasting glucose levels, C-peptide, and insulin resistance [35]. Serednytskyy et al. evaluated the coexistence of OSA and GDM and supported that they were related to increased sympathetic activity and IL-1β levels [36]. Conclusively, GDM was more common in women with OSA, while there is also a higher incidence of OSA in pregnant women with GDM. The findings are summarized at Table 2.

The incidence of GDM was higher in women diagnosed with narcolepsy type I before or during pregnancy compared to women diagnosed postpartum [37]. Besides, Calvo–Ferrandiz et al. concluded that GDM was more frequent in women with narcolepsy with cataplexy [38]. In a nationwide study performed in the USA, 7742 women with narcolepsy were detected (88% were diagnosed with narcolepsy type II or narcolepsy without cataplexy). However, no association between narcolepsy and GDM was observed [39].

Weschenfelder et al. evaluated 235 women with a history of GDM and concluded that pregnant women raising children and having unfavorable work conditions had a higher need for long-acting insulin therapy [40].

A recent study revealed a higher frequency of RLS in women with GDM [41]. Bisson et al. compared 26 pregnant women with GDM with 26 controls, concluding that there was a higher incidence of RLS in the GDM subgroup (46% vs 19%) [42].