Commentary on: ‘Finger pulse plethysmography predicts gestational hypertension, preeclampsia and gestational diabetes’

The development of cardiometabolic complications during gestation, such as preeclampsia and gestational diabetes, significantly increases the risk of maternal and neonatal morbidity and mortality. Several established risk factors, including increased maternal age, a higher BMI and a previous complicated pregnancy, are known to increase the risk of these cardiometabolic complications. However, prediction is often difficult in many women, especially in those who have not previously given birth (nulliparous women). The development of additional predictive indicators would be of significant benefit in obstetric medicine. In this issue, Shahrbabaki et al. [1] present data describing an association between pulse wave amplitude characteristics derived from finger photoplethysmography (PPG) measurements in early pregnancy and the development of pregnancy complications, specifically gestational hypertension, preeclampsia and gestational diabetes, in a cohort of nulliparous pregnant women. PPG is a noninvasive technique that assesses changes in capillary blood volume across cardiac cycles; the amplitude of the PPG signal can therefore be used as a surrogate of pulse wave amplitude, which reflects heart and blood vessel health. The potential usage of wearable devices in the monitoring and prediction of cardiovascular complications is an exciting field; for an excellent review on the utility of PPG in this context, the reader is referred to a recent article by Charlton et al. [2]. In the present study, overnight PPG measurements were made on a total of 2714 pregnant women in the first trimester of pregnancy. From these PPG recordings, the authors analysed changes in pulse wave amplitude over the nocturnal period, deriving parameters of pulse wave amplitude attenuation events (sustained reductions in pulse wave amplitude), thought to reflect an increase in sympathetic outflow to vessels and vasoconstriction. The authors determined both the mean duration of these attenuation events (PWAduration, in s) and mean depth of the attenuations (PWAdepth) in participating women. The authors report BMI and age-dependent associations with PWAduration, with stepwise increases in PWAduration seen in overweight and obese women, and higher values in participants aged 35 years or older as compared with those under 22 years. The authors also report higher PWAduration in women who were smokers or who drank alcohol in early pregnancy. However, it is notable that whilst differences between these groups were statistically significant, the actual differences were marginal, in the region of 1–2%. In addition, the relationships between PWAdepth and maternal demographics were less clear, with patterns inconsistent amongst the different groups of women. The authors then determined 'optimal' cut-off values for PWAduration and PWAdepth and used the resultant dichotomised data to calculate crude and adjusted odds ratios (ORs) for the associations between PWA attenuation events and gestational hypertension, preeclampsia and gestational diabetes. Women with higher PWAduration (>8.74 s) in the first trimester were more likely to develop gestational hypertension in later pregnancy (crude OR 1.81; adjusted OR 1.75; P < 0.001 for both). Associations between higher PWAduration with preeclampsia and gestational diabetes were only significant prior to adjustment, but were no longer significant after adjustment (made for race, categorised maternal BMI and age, gravidity, history of spontaneous deliveries or miscarriages, smoking and drinking habits, SBP and DBP, total sleeping time and number of apnoea/hypopnea events). Conversely, higher PWAdepth was not associated with preeclampsia and gestational diabetes prior to adjustment, but following adjustment there was a marginal, but significant, positive association. Whilst potentially interesting, there are several aspects of the data presented in this study that warrant mentioning here. Firstly, it is notable that the range of PWA parameter values reported in this study are narrow, with mean PWAduration between 7 and 12 s across the entire cohort, compared with a previous study in older, nonpregnant individuals where the values of a seemingly equivalent parameter (PWA-drop duration) were much broader, ranging between 4 and 28 s [3]. This prior study showed a clear and graded relationship across quartiles of PWA-drop durations, with higher values associated with increased rates of hypertension and diabetes [3]. It is not clear whether the differences between these studies might be due to methodological variation in the algorithms used to derive PWA attenuation parameters from PPG waveforms, or whether the narrower range of values in the present study reflects a specific feature of cardiovascular physiology of pregnant women. It is also not entirely clear from the article how the authors developed and validated the optimal cut-off values for the PWA attenuation parameters of interest; this may be particularly relevant given some of the borderline significance values reported in this study. Perhaps most importantly though, as noted above, both the absolute and relative differences in PWAduration and PWAdepth reported between different groups of women are small. As such, the prognostic value of these measurements for the prediction of pregnancy complications is questionable. What is promising from this study is the significant amount of data obtained from noninvasive home measurements in a large cohort of women in early pregnancy. If truly predictive parameters could be derived from PPG measurements or similar, this approach could provide important information to help clinicians determine which women may benefit from additional interventions to reduce the risk of pregnancy complications. The application of wearable technology and the future of home monitoring has the potential to significantly improve prediction and management of cardiometabolic complications; further developments in this space in pregnancy research are eagerly awaited. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.