Muscle Sympathetic Action Potential Firing Patterns During Normotensive and Hypertensive Pregnancy: A Longitudinal Assessment

Globally, gestational hypertension (GH) represents one leading cause of maternal-fetal morbidity and mortality. While its pathogenesis remains to be fully elucidated, accumulating evidence suggests that sympathetic neural dysregulation may play an important role^1,2. In a cross-sectional design, we recently demonstrated during late-pregnancy, that women who develop GH display a marked augmentation of sympathetic action potential (AP) firing per integrated burst, as well as total AP firing frequency, compared to women with normal pregnancies³. It remains unknown whether aberrant AP discharge is present also during early-pregnancy in women who subsequently develop GH, when increased muscle sympathetic nerve activity (MSNA) burst frequency is already manifest¹, and how such firing patterns influence resting blood pressure (BP) throughout gestation. Therefore, we investigated, prospectively, the patterning of resting multi-unit muscle sympathetic AP discharge throughout both normotensive and hypertensive pregnancies.

Twenty-four otherwise healthy, non-smokers who were planning to become pregnant or were within the first 8-weeks of pregnancy participated following informed consent. Women were tested longitudinally pre-pregnancy (mid-luteal phase; n=12), during early- (4-8 wks gestation; n=24) and late-pregnancy (32-36 wks gestation; n=24), and post-partum (6-10 wks post-delivery; n=22). Following term, women were divided into two groups: 1) those with healthy, normal pregnancies (NP; n=18; early-pregnancy: 31±5 yrs, 70±12 kg; late-pregnancy: 81±12 kg); and 2) those who developed GH following their late-pregnancy testing (GH; n=6; early-pregnancy: 30±4 yrs, 81±22 kg; late-pregnancy: 93±19 kg), diagnosed clinically, as previously detailed^1,3. None developed pre-eclampsia. The study received IRB approval (STU-012011-198). Parts of these data have been reported to address separate, unique hypotheses^1,3.

Heart rate (HR), BP (electrosphygmomanometry), and MSNA (microneurography) were collected for 6-minutes during supine rest. Resting MSNA was analyzed from the traditionally-measured, integrated neurogram and from our novel approach to extract APs from the filtered raw MSNA signal, using wavelet-based methodology³. Multi-unit AP firing patterns were quantified as the mean AP content per integrated burst (spikes/burst) and AP frequency (spikes/min).

Resting mean BP was unchanged throughout gestation in NP (all P>0.05); in GH, it increased from early- to late-pregnancy (P<0.05), and, thus, was greater than NP during late-pregnancy and post-partum (both P<0.05; Figure 1A). Resting HR increased during late-pregnancy in NP and GH (both P<0.05 vs. pre-), but no differences existed between groups throughout gestation (all P>0.05; Figure 1B). MSNA burst frequency increased during late-pregnancy in NP and GH (both P<0.05 vs. pre-); as anticipated¹, throughout early- and late-pregnancy, and post-partum, it was greater in GH (all P<0.05; Figure 1C). During pre-pregnancy, the mean AP content per burst was similar between groups (P>0.05); it increased during late-pregnancy in NP (P<0.05 vs. pre-), whereas, in GH, increased during both early- and late-pregnancy (both P<0.05 vs. pre-), before returning to pre-pregnancy levels in both groups post-partum (both P>0.05). Therefore, throughout early- and late-pregnancy, the mean AP content per burst was greater in GH (both P<0.05; Figure 1D). Furthermore, AP frequency increased during late-pregnancy in NP (both P<0.05 vs. pre-), whereas, in GH, it increased during both early- and late-pregnancy (all P<0.05 vs. pre-), before returning to pre-pregnancy levels in both groups post-partum (both P>0.05). Therefore, throughout early- and late-pregnancy, AP frequency was greater in GH (all P<0.05; Figure 1E). Finally, the change in AP frequency from early- to late-pregnancy was related directly to resting BP in late-pregnancy (R²=0.32; P<0.01; Figure 1F).

Figure 1.

(A) mean arterial blood pressure (BP), (B) heart rate, (C) muscle sympathetic nerve activity (MSNA) burst frequency, (D) multi-unit action potentials (AP) per integrated burst, and (E) AP frequency at supine rest during pre- (NP, n=9; GH, n=3), early- (NP, n=18; GH, n=6), late- (NP, n=18; GH, n=6), and post-pregnancy (NP, n=17; GH, n=5) in women with healthy, normal pregnancies (NP; n=18) and in women who developed gestational hypertension (GH; n=6). Data are presented as mean ± standard deviation. (F) Relationship between the change in AP frequency from early- to late-pregnancy and resting mean arterial BP in late-pregnancy in women with NP (n=17) and in women who developed GH (n=5). *Significantly different from pre-pregnancy, P<0.05. ^†Significantly different from early-pregnancy, P<0.05. ^αSignificantly different than NP, P<0.05. The effects of group and time were assessed with a repeated measures analysis using a mixed model design, with Bonferroni-correct post hoc procedures. In addition to the overall average effect, the slope and intercept were allowed to vary from participant to participant (random effect), thus providing a more robust analysis of longitudinal data sets through its ability to accommodate missing data points and model nonlinear, individual characteristics. Linear regression analyses were used to determine specific relationships between variables of interest. Statistical significance was set at P<0.05. The authors declare that all supporting data are available within the article.

We demonstrate for the first time in a longitudinal manner, the presence of aberrant multi-unit sympathetic AP discharge patterns during early- and late-pregnancy in women who subsequently develop GH. Specifically, women with hypertensive pregnancies display a marked augmentation of within-burst AP firing and total AP firing frequency during early-pregnancy, i.e., within 8-weeks of gestation, both of which remain exaggerated throughout gestation. Conversely, such indexes are not elevated during early-pregnancy in women who remain normotensive. Thus, the overall sympathetic burden in GH is dramatically larger when considering AP discharge behavior. During early-pregnancy, the traditional metric of MSNA burst frequency was ~66% higher in GH than NP, whereas total AP firing frequency (the product of burst frequency and the number of APs firing per burst) was ~240% higher. This level of excess in GH persisted (and even increased) at late-pregnancy and suggests that AP firing patterns represent a more sensitive measure to detect and/or predict pregnancy-related abnormalities.

Furthermore, changes in total AP firing frequency from early- to late-pregnancy were related directly to resting BP in late-pregnancy, providing evidence of a role for abnormal sympathetic AP patterning in the pathogenesis of GH. Notably, our findings suggest that the substrate for elevated BP – and a potential warning signal or biomarker for GH development – appears to be present as early as the first trimester. Considering the rise in global incidence⁴, and the elevated and lasting morbidity associated with GH diagnosis⁵, our results emphasize the importance of investigating potential sympatho-inhibitory strategies aimed at reversing such disturbance early-on during pregnancy in women at risk of GH.