Table 1.
Characteristics of included studies.
| Author, year | Study design | Country | Cases/controlsa | Age (mean ± SD, years) | Gestational age studied (n) | Olfactory assessment tool(s) | Main finding(s) | Qualityb | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Cases | Controls | 1T | 2T | 3T | PP | |||||||
| Gilbert and Wysocki (1991) | Cross-sectional | United States | 13,610/277,278 | 32.23 | 30.05 | * | * | * | * | 6 “scratch and sniff” odorants (OD, OH, OI, OID, Odor-Evoked Memories) |
Frequency of odor-evoked memories and self-rated olfaction in cases were lower compared to controls. Cases also identified mercaptans at a higher rate, identified androstenone at a lower rate, rated androstenone and galaxolide as less intense, and rated isoamyl acetate and mercaptans as more intense than controls | Fair |
| Laska et al. (1996) | Longitudinal | Germany | 20/20 | 27.5 ± 3.6 | 26.2 ± 4.5 | 20 | 20 | 20 | 20 | Squeeze bottles (OT, OD, OI, OH, OID) | Cases were less sensitive in the 1T, but more sensitive in the 3T compared to controls. Overall, systematic changes in OI, OD (intensity discrimination), OH, and OID were not found. Differences between cases and controls in OI, OH, or OID were either odor or trimester-specific | Fair |
| Dastur (2000) c | Longitudinal | Canada | 19/18 | 30.2 ± 0.8 | 28.8 ± 1.9 | 19 | 19 | 19 | 19 | PEA threshold testd (OT) UPSIT (OID) |
Cases had lower OT scores than controls overall and at each trimester, but not in postpartum. 1T OT scores were significantly lower than 2T, 3T, and PP scores. 2T and 3T OT scores did not differ significantly, but both 2T and 3T OT scores were lower than PP OT scores. There were no significant differences in OID scores between cases and controls. There were no significant differences between 1T, 2T, 3T, and PP OID scores. Across all test sessions, OT and OID scores were significantly correlated. However, for any of the individual test sessions, OT and OID scores did not significantly correlate | Good |
| Kölble et al. (2001) c | Cross-sectional | Switzerland | 53/59 | 28.0 ± 5.6 | 32.0 ± 4.7 | 53 | 0 | 0 | 0 | SSTe (OT, OD, OID) VAS ratings of 10 “ natural odors” (OI, OH) |
Pooled sensitivity scores (i.e. sum of OT, OD, and OID scores) did not differ significantly between cases and controls. OID was significantly lower in cases than controls. OT, OD, and OI did not differ significantly between cases and controls. Odor-specific differences found with regards to OH (cigarette, rum and coffee) | Good |
| Savović et al. (2002) | Cross-sectional | Serbia | 20/20 | * | * | 20 | 0 | 0 | 0 | Olfactometer (OT, OID) | Cases had lower OT and OID, but difference non-significant | Fair |
| Olofsson et al. (2005) | Cross-sectional | Sweden | 15/15 | 30.7 ± 4.7 | 31.1 ± 4.3 | 0 | 15 | 0 | 0 | CSERP | Reported hypersensitivity in cases may be due to cognitive changes, not changes in sensory processing. Differences were non-significant (P-value > 0.05) | Fair |
| Cameron (2007) c | Cross-sectional | United States | 60/20 | 29.7 | 25.9 | 20 | 20 | 20 | 20 | UPSIT (OID, OI, OH) | OID, OI, and OT did not differ between cases and controls, but odor-specific changes in OID, OI, and OH were reported. Cases in their 1T reported higher overall OI and lower OT, but these differences were found to be statistically non-significant in a 1-way ANOVA | Fair |
| Ochsenbein-Kölble et al. (2007) a | Longitudinal | Switzerland | 38/46 | 29.0 ± 5.7 | 32.0 ± 4.7 | 38 | 38 | 38 | 38 | Sniffin’ Sticks (OT, OD, OID) VAS ratings of 10 “ natural odors” (OI, OH) |
Cases in their 3T and PP had significantly lower TDI scores compared to controls. OT was significantly lower (lower sensitivity per authors) in pregnant women in the 3T and PP compared to controls. OID did not change significantly across sessions and did not differ compared to controls. Pregnant individuals rated clove as significantly less pleasant than controls. Among pregnant women, acetic acid was rated as more pleasant during the 2T and 3T, but these ratings did not differ between pregnant women and controls when considering all four sessions. Coffee was rated significantly more unpleasant by pregnant women than controls during the 1T, but was not rated significantly more unpleasant in the 2T, 3T, or PP compared to controls | Good |
| Cameron (2014a) | Cross-sectional | United States | 17/32 | 25 | 20 | * | * | * | * | PEA threshold testd (OT) | There was no significant difference in OT measurements between cases and controls | Fair |
| Cameron (2014b) | Longitudinal | United States | 20/22 | 31 | 26 | 20 | 20 | 20 | 0 | PEA threshold testd (OT) Signal detection task (Sensitivity, Response Bias) |
There were no significant differences in OT measurements between trimesters or between cases and controls. There were also no significant differences in sensitivity or response bias between trimesters or between cases and controls. However, there was a tendency for more response bias earlier in pregnancy | Fair |
| Kyung-yeon (2014) c | Longitudinal | Korea | 50/40 | 30.3 ± 3.3 | 28.5 ± 6.6 | 50 | * | * | 29 | SSTf (OT, OD, OID) | No significant differences in OT, OD, OID or TDI | Good |
| Şimşek et al. (2015) c | Cross-sectional | Turkey | 92/30 | 27.4 ± 5.3 | 26.0 ± 4.2 | 31 | 30 | 31 | 0 | B-SITg (OID) | 1T cases had significantly lower median identification scores compared to 2T, 3T, and controls. 1T women identified leather, pine, and soot significantly less well than controls. Both 2T and 3T OID scores did not differ significantly from those of controls | Good |
| Nwankwo et al. (2017) c | Cross-sectional | Nigeria | 70/70 | 30.5 ± 3.9 | 28.5 ± 6.6 | * | * | * | * | SST (OT, OD, OID) | OID and TDI were significantly lower in cases compared to controls | Good |
| Fornazieri et al. (2019) c | Cross-sectional | Brazil | 124/50 | 26.4 ± 6.26 | 26.4 ± 5.0 | 47 | 33 | 44 | 32 | UPSITh (OID) VAS ratings of 4 common UPSIT odors (OH, OI) |
No difference in OID between 1T, 2T, 3T, or and controls. There were several significant odor-specific differences (P-value < 0.05) in OID, OH, and OI between various groups | Good |
SD, standard deviation; 1T, 1st Trimester; 2T, 2nd Trimester; 3T, 3rd Trimester; PP, postpartum; OT, odor threshold; OD, odor discrimination; OID, odor identification; OI, odor intensity; OH, Odor Hedonics; TDI, sum of the OT, OD, and OID scores as measured using Sniffin’ Sticks; VAS, Visual Analogue Scale; SST, Sniffin’ Sticks; UPSIT, University of Pennsylvania Smell Identification Test; PEA, Phenyl Ethyl Alcohol; B-SIT, Brief Smell Identification Test; CSERP, chemosensory event-related potentials.
*Not specified.
aAll cases were pregnant women, and all controls listed were healthy, non-pregnant women.
bQuality was determined using a modified Newcastle-Ottawa Scale. Scores include good, fair, and poor.
cIncluded in meta-analysis.
dPEA threshold test performed as outlined in Doty et al. (2000).
eAll Sniffin’s Sticks tests included the version with 16 odorants.
fKorean adaptation of the SST.
gTurkish adaption of the B-SIT.
hPortugese adaptation of the UPSIT.