Changes in gustatory function during the course of pregnancy and postpartum

Authors


Dr N. Ochsenbein-Kölble, Obstetric Research Unit, Department of Obstetrics, University Hospital, Zurich, Frauenklinikstr. 10, CH-8091 Zurich, Switzerland.

Abstract

Objective  To investigate changes in gustatory function during pregnancy and postpartum.

Design  Prospective study.

Setting  Obstetric outpatient clinic, Zurich University Hospital.

Population  Forty-four pregnant women, 46 controls.

Methods  A threshold-like measure of gustatory function was obtained for different concentrations of the four basic tastes using ‘taste tablets’; subjects were also asked to rate the intensity and pleasantness/unpleasantness of suprathreshold taste solutions. Tests were performed in early pregnancy up to 14 weeks of gestational age, 8–12 weeks after the first and second examination and at least 6 weeks after birth. In parallel, controls were tested four times at intervals of 8–12 weeks.

Main outcome measures  Gustatory score (number of correctly identified tastes).

Results  Pregnant subjects exhibited a decreased gustatory function compared with non-pregnant controls, still being present after birth. These ‘objective’ changes were observed although pregnant women rated the intensity of suprathreshold taste stimuli not significantly different from controls. Pregnant women rated salty stimuli to be more pleasant during the second trimester while they rated the salty stimuli to be less pleasant during the first and third trimester, and after birth.

Conclusions  Pregnancy is accompanied by changes in gustatory function. Decreased gustatory sensitivity may allow pregnant and breastfeeding women to vary their diet in order to consume adequate electrolytes. The discrepancy between ‘objective’ and ‘subjective’ findings in gustatory function may relate to changes in central processing of gustatory information during pregnancy.

INTRODUCTION

Changes in the perception of tastes are experienced by a majority of pregnant women1–5 and was recently reported by 93% of pregnant women in a questionnaire study.6 Pregnant women most frequently reported cravings for sweet foods, especially chocolate, milk and fruits.1,3,4,7 The most common aversions were described for meats, eggs or sauces flavoured with oregano.1,7 In the course of pregnancy food preferences could change (e.g. sweet foods were generally preferred during the second trimester, whereas salty foods appeared to be preferred during the third trimester).2

Why some foods taste different during pregnancy is not known. Hook1 suggested that possible factors mediating the development of food aversions and cravings may be changes in taste and olfactory sensitivity or hormonal changes accompanying pregnancy. The direction of change in gustatory sensitivity during pregnancy, however, is unclear: an increase in sensitivity,8 a decrease in sensitivity9–12 and even ageusia13 have been reported. In addition, little information is available on measured gustatory function during pregnancy and the postpartum period.

It is not surprising that results in studies on gustatory function in pregnant women vary: (1) methods exhibited large variation, including, for example, the use of unvalidated test systems8,14; (2) studied populations have been small (n < 25)10; (3) results from all trimesters have been integrated in one ‘pregnancy group’10; (4) studies have been designed as cross sectional and not longitudinal investigations6,8,14,15; (5) there has been a lack of parallel control groups/missing specification of the day of the menstrual cycle on which controls were tested6,8,10,16; and (6) often only one or two gustatory qualities have been tested.8,10,15

To address the question whether gustatory function changes during the course of pregnancy and postpartum, this prospective study was performed in a large sample of pregnant women and non-pregnant controls in a known phase of their menstrual cycle using a validated test of gustatory function.17

METHODS

For the study, 63 pregnant women (age 29 [5.7] years, mean [SD]) and 59 non-pregnant women (age 32 [4.7] years) were recruited following a detailed explanation of the protocol and written informed consent. The major portion of the participants were nulliparae. The study was performed according to the Declaration of Helsinki regarding Biomedical Research Involving Human Subjects. It was approved by the Ethics Committee of the University of Zurich Medical School.

Subjects in the two groups were of similar height and weight [pregnant women: 165.0 cm, 62.6 kg (pre-pregnant weight); non-pregnant women: 167.5 cm, 60.8 kg]. Pregnant subjects were tested in early pregnancy up to 14 weeks of gestational age (n= 63), 8–12 weeks after the first (n= 56) and second (n= 45) examination and at least 6 weeks after birth (n= 44). In parallel, controls were tested four times after a pause of 8–12 weeks (first measurement: n= 59; second measurement: n= 57; third measurement: n= 53; fourth measurement: n= 46) in the second half of the menstrual cycle in order to avoid the early luteal phase that might influence taste perception,18,19 or, for those on oral contraception, not at the time of menstruation. The first, second, third and fourth measurements in pregnant women were compared with the corresponding measurement in controls.

Based on their self-report, all subjects were in excellent health. They were instructed to refrain from smoking, drinking coffee and eating at least 1 hour before testing. Exclusion criteria were severe disease (e.g. bad adjusted endocrine disease or epilepsy), drug and nicotine abuse (>10 cigarettes per day) and known gustatory dysfunction. Beside two subjects of each group with appropriate, well-adjusted thyroxine supplements, none of the pregnant women or controls had a further medication than vitamin supplements.

All subjects completed a pre-test self-rated questionnaire on nausea and vomiting in the previous three days. They also completed a mood questionnaire20 and nausea profile21 (parts of these results from the nausea questionnaire have been published previously22).

Gustatory function was assessed during the first trimester of pregnancy and on average seven weeks after birth, and in controls at the first and fourth measurements, always by the same two investigators. The four basic tastes were presented in round white 4-mm sweetener-like tablets17 containing sucrose (‘sweet’), sodium chloride (‘salty’), citric acid (‘sour’) or caffeine (‘bitter’), each at five different strengths (in 50% increments) to quantitate gustatory function. Microcristalline cellulose (Avicel PH101, Lehmann & Voss & Co., Hamburg, Germany), polyethylene glycol 6000 (Lutrol E 6000, BASF, Germany) and Aerosil 200 (Degussa, Frankfurt, Germany) were used as excipients, as in standard tablets. Subjects received the 20 tablets (four tastants at five strengths, numbered 1–20), placed on the tongue using tweezers, in pseudorandom order. No placebo tablets were presented. They were instructed to chew the tablets until they identified the taste (‘sweet’, ‘salty’, ‘sour’, ‘bitter’, ‘no taste’) then rinse with low-sodium, non-carbonated mineral water. The tablets were presented at half-minute intervals over 15−20 minutes. The correct identified tablets were scored receiving a gustatory score (range 0–5) for each tastant.

Intensity and character of four suprathreshold solutions were tested on all four measurements in both groups. To analyse responses to suprathreshold stimuli, solutions of ‘sweet’, ‘salty’, ‘sour’ and ‘bitter’ (saccharose 10%, NaCl 7.5%, citric acid 5% and quinine 0.5%) were sprayed onto the tongue in random order. Subjects rinsed between stimulants and rated intensity and pleasantness/unpleasantness on 10-cm pencil-and-paper visual analogue scales (from ‘no sensation’ and ‘extremely unpleasant’, respectively, on the left to ‘extremely intense’ and ‘extremely pleasant’, respectively, on the right). Completion of this test took approximately 5 minutes.

Results are presented as the mean [SD] (except Fig. 1). Data were analysed using SPSS 12.0 for Windows. Results were submitted to analyses of variance for repeated measures (rm-ANOVA; within-subject factors ‘session’ [first, second and third trimester; postpartum]; between-subject factor ‘group’ [pregnant, non-pregnant]). Bonferroni tests were used for post hoc comparisons. The alpha level was 0.05.

Figure 1.

Changes of gustatory function, expressed as gustatory score (mean [SEM]), separately for the four tastes, ‘sweet’, ‘salty’, ‘sour’ and ‘bitter’, in non-pregnant controls (n= 46) and pregnant women (n= 44). Measurements were taken in the first trimester of pregnancy (on average, 12 weeks of gestational age) and on average seven weeks after birth.

RESULTS

A total of 44 pregnant women completed the measurements taken during the first trimester of pregnancy and the one, on average, performed seven weeks after birth. Forty-six non-pregnant controls completed the first and the fourth measurement and their results were compared with the results of the pregnant group.

Overall, gustatory function was significantly lower in pregnant women compared with controls (factor ‘group’: F[1,88] = 5.99, P= 0.016). In both groups there was no significant difference between measures taken during the first and the second session (P > 0.6). Individual analysis for the four tastes (Fig. 1) used revealed group differences only for ‘bitter’ (factor ‘group’: F[1,88] = 17.2, P < 0.001).

The two groups did not show any significant differences with regard to taste intensities (Table 1).

Table 1.  Results from intensity ratings of the four different tastes in non-pregnant controls (n= 46) and pregnant women (n= 44). Measurements were taken at sessions 1–3 during pregnancy (on average, gestational weeks 12, 21 and 36) and seven weeks after birth (session 4).
Session GroupSweetSourSaltyBitter
MeanSDMeanSDMeanSDMeanSD
1Control6.91.88.41.37.41.79.11.3
Pregnant6.52.58.21.67.42.29.01.3
2Control6.41.87.51.66.71.68.71.3
Pregnant5.72.07.81.97.11.88.81.6
3Control6.42.17.21.57.11.78.71.3
Pregnant6.02.47.21.97.31.98.41.4
4Control6.91.58.61.27.31.89.11.2
Pregnant7.02.08.41.78.11.78.91.2

Pregnant women rated ‘salty’ stimuli to be more pleasant during the second trimester of pregnancy (Table 2) while they rated ‘salty’ to be less pleasant during the first and third trimester and after birth (‘session’*‘group’: F[3,249] = 6.17, P= 0.001). Pleasantness of ‘salty’ increased significantly from the first to the second trimester and decreased thereafter to reach its lowest value postpartum (P < 0.001). For ‘sweet’, ‘sour’ and ‘bitter’ no significant differences were detected in ratings of pleasantness/unpleasantness in the course of pregnancy and postpartum.

Table 2.  Results from ratings of pleasantness/unpleasantness of the four different tastes in non-pregnant controls (n= 46) and pregnant women (n= 44). Measurements were taken at sessions 1–3 during pregnancy (on average, gestational weeks 12, 21 and 36) and seven weeks after birth (session 4).
Session GroupSweetSourSaltyBitter
MeanSDMeanSDMeanSDMeanSD
1Control5.72.63.62.74.22.01.01.7
Pregnant6.22.63.72.73.72.51.11.9
2Control6.02.74.92.53.91.51.01.2
Pregnant6.12.04.92.94.82.50.81.2
3Control6.22.35.01.74.52.11.32.0
Pregnant6.02.34.22.73.32.31.41.5
4Control6.02.34.22.63.72.30.81.3
Pregnant6.62.43.52.83.12.21.51.9

Results from the mood scale exhibited no significant differences between the two groups or between the four sessions. However, there was a significant interaction between these two factors indicating that, with the exception of the second trimester, pregnant women rated themselves to be in a better mood than control subjects (‘session’*‘group’: F[3,249] = 4.48, P= 0.004).

DISCUSSION

This is the first longitudinal study on measured gustatory function in pregnancy and postpartum comparing data to non-pregnant controls. We found that gustatory function was significantly lower in pregnant women in the first trimester and in the postpartum period compared with controls. This was largely due to decreased ‘bitter’ sensitivity. However, pregnant women rated the stimuli to be as intense as controls.

Concerning decreased gustatory function in pregnancy, our results are in line with gustatory threshold findings reported by Kuga et al.6 They performed a study on gustatory function in the first, second and third trimester of 32 pregnant women who underwent sequential tests of gustatory function. Thirty healthy, non-pregnant women participated as controls. Measurements of gustatory thresholds using a filter paper disk test showed that pregnant women had significantly higher gustatory thresholds than non-pregnant women, with an especially marked decrease in gustatory function in the first trimester. In their study, however, controls were not tested in a longitudinal manner but only once in a cross-sectional design, the day of the menstrual cycle was not specified on which gustatory function was examined and no information was given on gustatory function postpartum. This seems important as gustatory function has been reported repeatedly to change in relation to the period of the menstrual cycle.15,16,19,23

While the detailed measures of gustatory sensitivity indicate a decrease of gustatory function in pregnancy and after birth, there was no significant difference between taste intensity ratings of pregnant women and controls. An explanation for the discrepancy between ‘objective’ and ‘subjective’ findings may refer to alterations in the central processing of gustatory information due to pregnancy, which has already been suggested in pregnant animals.24

‘Bitter’ sensitivity in the first trimester of pregnancy has been described as both, decreased6,14 and increased.8 In addition to Kuga et al.,6 Takahashi et al.14 confirmed the reduced responsiveness for ‘bitter’ in the first trimester of pregnancy compared with controls, although they failed to analyse their data statistically. Contrary to our results, Bhatia and Puri8 found an increased ‘bitter’ sensitivity in the first trimester. Disadvantages of their study, however, were the examination of only one taste quality, the use of phenylthiocarbamide, which is notorious for large inter-individual differences in sensitivity,25 the testing of different pregnant women in each trimester of pregnancy and like Takahashi et al.14 the failure to analyse their data statistically.

The causes of decreased gustatory function in pregnancy have not yet been clarified. Nevertheless, a number of factors have been investigated, including the need for increased intake of salt and calories in the course of pregnancy,2,4,10,26 changes in serum concentrations of trace elements,6,14 or in oestrogen and progesterone levels.2,15,23 Both Kuga et al.6 and Duffy et al.16 proposed that taste changes in pregnancy were due to fluctuations in sex hormones. Others suggested that changes of gustatory function are due to the proximity of gustatory and uterine representations within the insular neocortex. While unsubstantiated by experimental data, it was hypothesised that ‘…menstrual or pregnancy stimulation of the uterus might shift the locus of neuronal activity within the insula to include adjacent gustatory neurons and consequently alter taste experiences’.27

As mentioned above, there were no significant differences for ratings of taste intensities between both groups in the course of pregnancy and postpartum. However, with regard to the pleasantness/unpleasantness of the tastes, pregnant women rated ‘salty’ to be most pleasant in the second trimester, and least pleasant after birth.

Others also reported pregnancy-related changes in hedonic ratings.16 They described an increase in ‘bitter’ intensity during the first trimester and a reduction in intensity ratings of ‘salty’ and ‘bitter’ in the second and third trimester. This was accompanied by a reduction in unpleasantness of ‘salty’, ‘bitter’ and ‘sour’. In a recent questionnaire-based study, in which pregnant women were asked which taste they most preferred during pregnancy, ‘sour’ was mentioned by approximately 66% of the women, with ‘salty’ being second with 46%.6

Another study found that pregnant women were less able to correctly identify differences between salt concentrations and preferred significantly stronger solutions than did non-pregnant women.10 Skinner et al.4 found that during pregnancy, but not after, salted food was preferred over unsalted and low-salt products, which confirms the present results. Regarding pleasantness/unpleasantness of ‘sweet’, pregnant women in our study as well as in the study by Brown and Toma10 exhibited no significant difference in the preference of sucrose solutions compared with controls.

Decreased gustatory function may enable pregnant women to consume adequate electrolytes for expanding their plasma volume and to facilitate weight gain by broadening their diet.1,16 The advantages of an increase of the pleasantness for ‘saltiness’ in pregnancy, as found in the present study, seems to be obvious, as an increased appetite for NaCl may facilitate increased intake of salts.28 In animal studies, NaCl restriction in pregnancy results in clinically significant changes (e.g. hypovolemia, smaller litters, compromised brain development, fewer live births/more stillbirths per litter).29–31

A potential increase in ‘bitter’ sensitivity may help to avoid poisons in the critical phase of embryonic development.1,32 Interestingly, pregnant women often describe abnormal taste function as increased ‘bitter’ sensitivity and decreased ‘salt’ sensitivity.33 In contrast to this, our own data in a large group showed a decrease in bitter sensitivity in the first trimester and postpartum. It is important to keep in mind that in animals bitter taste thresholds vary independently of toxicity thresholds, indicating that the bitter rejection response is just as likely to be elicited by a harmless bitter food as it is by a harmful one.34 Based on this observation, it has been hypothesised that the adaptiveness of the bitter rejection response depends upon the relative occurrence of bitter and potentially toxic compounds in an animal's diet.34 Because it is usually not expected to encounter potentially toxic compounds in the diet of humans, we hypothesise that the evolutionary benefit of an increased bitter sensitivity is not easily applicable to pregnant women. Increasing the intensity of ‘bitter’ sensation would increase the risk of hyperemesis35 that can harm both mother and fetus. Similar hypothesis has been built for a decrease of the liking of sweet and high-fat food/beverages,36 which may hamper sufficient nutrient intake in pregnancy or after birth. Thus, we propose that adaptive processes underlie the changes in taste function during pregnancy—they may allow pregnant and breastfeeding women to vary their diet in order to increase weight gain and support the child with all the necessary nutrients.

In conclusion, this is the first prospective study on measured gustatory function in pregnancy and postpartum. We found a decrease in gustatory function during pregnancy and after birth, mainly due to a decreased sensitivity for ‘bitter’. It supports the idea to allow pregnant and breastfeeding women to vary their diet in order to consume adequate electrolytes and to support the child with all the necessary nutrients. The discrepancy between ‘objective’ and ‘subjective’ findings in gustatory function may relate to pregnancy-related changes in central processing of gustatory information.

Accepted 7 September 2005

Ancillary