Longitudinal course of endocannabinoids and N-acylethanolamines in hair of mothers and their children in the first year postpartum: investigating the relevance of maternal childhood maltreatment experiences

Background Childhood maltreatment (CM) exerts long-lasting psychological and biological alterations in affected individuals and might also affect the endocannabinoid (eCB) system which modulates inflammation and the endocrine stress response. Here, we investigated the eCB system of women with and without CM and their infants using hair samples representing eCB levels accumulated during the last trimester of pregnancy and 10–12 months postpartum. Methods CM exposure was assessed with the Childhood Trauma Questionnaire. At both timepoints, 3 cm hair strands were collected from mothers and children (N = 170 resp. 150) to measure anandamide (AEA), 2/1-arachidonoylglycerol (2-AG/1-AG), stearoylethanolamide (SEA), oleoylethanolamide (OEA), and palmitoylethanolamide (PEA). Results Maternal hair levels of 2-AG/1-AG increased and SEA levels decreased from late pregnancy to one year postpartum. Maternal CM was associated with lower SEA levels in late pregnancy, but not one year later. In the children's hair, levels of 2-AG/1-AG increased while levels of SEA, OEA, and PEA decreased from late pregnancy to one year later. Maternal CM was not consistently associated with the eCB levels measured in children's hair. Conclusions We provide first evidence for longitudinal change in the eCB system of mothers and infants from pregnancy to one year later. While maternal CM influenced the maternal eCB system, we found no consistent intergenerational effects on early regulation of the eCB system in children. Longitudinal research on the importance of the eCB system for the course and immunoregulation of pregnancy as well as for the children's development.


Supplementary Information: Hair Processing and Endocannabinoid Analysis
While circulating eCB concentrations mirror a state and can fluctuate depending on circadian rhythmicity and acute stress (Vaughn et al., 2010), hair analyses provide the more reliable and stable measurement of eCB (Krumbholz et al., 2013), especially in the time around childbirth.Thus they allow for a retrospective measure of long-term eCB accumulation over weeks and months.Hair sampling took place directly after the psychodiagnostic interviews at t0 and t2 in mother-child-dyads that gave informed consent in sampling of biological specimen.
In total, hair was collected from N = 474 mothers and N = 331 children at t0 and N = 244 mothers and N = 237 children at t2.At t0 59 mother-child-dyads rejected hair sampling.In n = 143 mothers declined hair sampling in their infants, the newborns hair was to spares for sampling or infants were too fidgety.At t2 26 mother-child-dyads rejected hair sampling; n = 7 children had to spares hair or were too fidgety for hair sampling (for general drop out reasons between t0, t1 and t2 see Figure S1).Upon limited material, the analysis of the primary hair-based biological marker, steroid hormones, was prioritized (data not shown here).At both times of measurement eCB and NAE were quantified in all remaining hair specimen with sufficient material.At t0, sufficient material for additional eCB and NAE quantification was available for 150 mothers and 92 children; and at t2, eCB and NAE were measured in the hair of 148 mothers and 170 children.However, due to study drop outs, reduced available hair samples for eCB analysis and eCB values under technical detection limit only a subsample of N = 63 mothers and N = 45 children presented complete eCB and NAE data at both measurement points in time (see Figure S2 for details).
Hair collection.At t0 and t2, hair samples were collected and processed by trained academic staff, using laboratory gloves to avoid contamination of the hair with skin moisture.In mothers, optimally three hair strands (~3mm diameter each) were cut close to the scalp from the posterior vertex position.When this sampling location was not possible for the children due to sparse hair, infant hair samples were cut from locations with most hair, usually at the hairline beneath the ear.After parturition (t0), newborns' hair was washed with clear water to preclude contamination with blood or amniotic fluid.Hair samples were wrapped in aluminum foil and stored at −20 °C to minimize possible degradation effects and loss of biochemical compounds due to long-term storage and humidity.To avoid contamination of the hair with skin moisture, hair samples were collected and processed using laboratory gloves.
Pre-processing.Before shipment to the laboratory of Prof. Kirschbaum at TU Dresden for LC-MS/MS mass spectrometry analyses, the preprocessing of all hair samples was performed in the laboratory of Prof. Kolassa at Ulm University.First, hair strands of the same subject were pooled.In a standardized procedure, the 3cm-hair segment proximal to the scalp was cut.Due to an approximate adult hair growth of ~1 cm/month (Wennig, 2000), the 3cm-hair segment proximal to the scalp reflects maternal cumulative eCB concentration incorporated in the last three months.However, fetal/neonatal hair grows slower than in adults with ~1cm in during the three months of the third trimester of pregnancy (cf.Gareri & Koren, 2010).Thus, to display the metabolic activity during the last three months of prenatal development, hair of newborns collected at t0 was cut to 1 cm segments.At t2 both, the proximal 3cm-hair segment of mothers and children was used for analyses, reflecting month 10 to 12 postpartum.To our knowledge, studies of children's hair growth velocity, in particular for children aged around 12 months, are spares.First evidence suggests, that hair growth velocity increases over the first year after birth and aligns to the growth rate of adults (Pecoraro et al., 1964;de Kruiff et al., 2020): In a small sample of children aged 6 to 12 months hair growth velocity was around ~6.7 mm/ per month and with 12 to 24 months ~9.4 mm/ per month (c.f., de Kruiff et al., 2020).Thus, for eCB analysis one year after birth, for mothers and children likewise, the 3cm-hair segment proximal to the scalp was used to reflect eCB levels of the same time period allowing to take a look into the first year postpartum.Cut hair of mothers and children was weighed (range 4-6mg) and placed into Falcon tubes.The non-pulverized hair of mothers had an average weight of 5.2 mg (SD=0.9)at t0 and 5.2 mg (SD = 0.7) at t2. Infant hair had an average weight of 2.4 mg (SD=1.7)at t0 and 5.3 (SD = 0.5) at t2.For further sample details and missing data see Figure 1 in the main manuscript.Note: At t0, AEA was analyzed in a subsample of mothers only, due to its low abundance in hair AEA measures of many cases were below the technical detection limit.The maximal number of available cases varies: a n = 39; b n = 98; c n = 53; d n = 44; e n = 89; f n = 43; g n = 45.

Supplementary Descriptive Data
† As endocannabinoid data is skewed and not normal distributed, M and SD are no reliable measures and are only reported for reasons of comparability.For analysis and interpretation Med and IQR are used exclusively.According to the mild cut-off criterion of the CTQ (Bernstein & Fink, 1998), women without any CM experiences were classified as CM− and those with at least mild CM experiences in at least one CTQ subscale were categorized as CM+.CM Childhood Maltreatment; CTQ Childhood Trauma Questionnaire AEA anandamide, 2-AG/1-AG 2/1-arachidonoylglycerol; OEA N-oleoylethanolamide; SEA N-stearoylethanolamide; PEA N-palmitoylethanolamide.
Supplementary Table S2 .08Note: * p < .050,** p < .010,*** p < .001,two-tailed.All p-values were adjusted for multiple comparisons using the false discovery rate (FDR).Note that at t0, AEA was analyzed in a subsample of mothers only.Due to its low abundance in hair the AEA measures of many cases were below the technical detection limit.The maximal number of available cases varies: a n = 39; b n = 98; c n = 89; d n = 62 mother-child-dyads; e n = 58 mother-child-dyads; f n = 62 mother-child-dyads; g n = 107 mother-childdyads.# Maternal values depicted in rows and children's values depicted in columns.AEA anandamide; 2-AG/1-AG 2/1-arachidonoylglycerol; OEA N-oleoylethanolamide; SEA N-stearoylethanolamide; PEA N-palmitoylethanolamide. S4  .058Note: * p < .050,** p < .010,*** p < .001,two-tailed.Underlined p-values are significant after correction with false discovery rate (FDR).Bivariate correlations were computed with the maximal number of cases available: a n = 39; b n = 97; c n = 88.Note that no AEA levels were detectable in the hair of children at t0.Exposure to childhood maltreatment (CM) was assessed with the Childhood Trauma Questionnaire (CTQ, Bader et al., 2009).S6   In accordance with our preceding study (Koenig et al., 2018) and for reasons of completeness we report supplementary analyses on CM-related group differences in eCB and NAE in mothers and children.

Supplementary Table
The following results can be found in Table S6 and Figure S2 and S3 for the sample of N=63 mothers and N=45 children with complete hair data at t0 and t2 who were included in the longitudinal analyses.Supplementary Table S7 and Figure S4 report the results using the complete available sample at t0 (Nm = 150; Nc = 92) and t2 (Nm = 148; Nc = 170), respectively.
Maternal hair.Partially confirming our prior results from a bigger study cohort within the "My Childhood -Your Childhood" study (see Koenig et al., 2018), in N=63 mothers at t0 maternal SEA concentrations were significantly lower in CM+ as compared to CM-women.2-AG/1-AG was descriptively higher in CM+ women as compared to CM-women, but failed to reach significance.OEA and PEA were descriptively lower in CM+ women as compared to CMwomen, but failed to reach significance.At 12 months postpartum (t2) these effects vanished: eCB and NAE hair concentrations did neither differ between mothers with and without a history of CM.
Infant hair.eCB levels in N = 45 newborn's hair were analyzed with respect to a maternal history of CM.At t0, in newborns with CM+ mothers, OEA hair concentrations were significantly higher than in newborns without maternal CM, whereas 2-AG/1-AG hair concentrations were descriptively higher in newborns with CM+ mothers than in newborns of CM-mothers.For SEA and PEA no statistical differences were found regarding maternal CM status.At 12 months postpartum (t2) these effects vanished: eCB and NAE hair concentrations did neither differ between children with and without a maternal history of CM Underlined p-values are significant after correction with false discovery rate (FDR).According to the mild cut-off criterion of the CTQ (Bernstein & Fink, 1998) women without any CM experiences were classified as CM− and those with at least mild CM experiences in at least one CTQ subscale were categorized as CM+.Wilcoxon tests were computed using all available samples of mother-child-dyads with hair samples collected at t0 and t2.Note that at t0, AEA was analyzed in a subsample of mothers only.As many of the AEA measures were below the technical detection limit, AEA had to be excluded from Wilcoxon rank-sum tests as there was no sufficient data.For analysis and interpretation Med and IQR are used exclusively.CM Childhood Maltreatment; AEA anandamide, 2-AG/1-AG 2/1arachidonoylglycerol; OEA N-oleoylethanolamide; SEA N-stearoylethanolamide; PEA N-palmitoylethanolamide. S3.Comparison of N = 63mothers with (CM+) and without childhood maltreatment (CM-) experiences in late pregnancy (t0, A-D) and one year after birth (t2, E-H) regarding their eCB and NAE hair concentrations represented by boxplots with overlying bee-swarms.Endocannabinoids are presented as raw data in pg/mg.CM+, women with CM experiences; CM-, women without CM experiences; 2-AG/1-AG 2/1arachidonoylglycerol; OEA N-oleoylethanolamide; SEA N-stearoylethanolamide; PEA N-palmitoylethanolamide. S4.Comparison of N = 45 children with (CM+) and without mothers exposed to childhood maltreatment (CM-) experiences in late pregnancy (t0, A-D) and one year after birth (t2, E-H) regarding their eCB and NAE hair concentrations represented by boxplots with overlying bee-swarms.Endocannabinoids are presented as raw data in pg/mg.CM+, women with CM experiences; CM-, women without CM experiences; 2-AG/1-AG 2/1-arachidonoylglycerol; OEA N-oleoylethanolamide; SEA N-stearoylethanolamide; PEA N-palmitoylethanolamide.

.
Raw data of endocannabinoid hair concentration in pg/mg for N = 63 mothers and N = 45 children with hair data at both points of measurement (Bernstein & Fink, 1998)lyzed in a subsample of mothers only, due to its low abundance in hair AEA measures of many cases were below the technical detection limit.†Asendocannabinoiddata is skewed and not normal distributed, M and SD are no reliable measures and are only reported for reasons of comparability.For analysis and interpretation Med and IQR are used exclusively.According to the mild cut-off criterion of the CTQ(Bernstein & Fink, 1998), women without any CM experiences were classified as CM− and those with at least mild CM experiences in at least one CTQ subscale were categorized as CM+.CM Childhood Maltreatment; CTQ Childhood Trauma Questionnaire AEA anandamide, 2-AG/1-AG 2/1-arachidonoylglycerol; OEA N-oleoylethanolamide; SEA N-stearoylethanolamide; PEA Npalmitoylethanolamide.Supplementary TableS3.

Intercorrelations of endocannabinoids at each point of measurement.
Bivariate Spearman rank correlations (rS) between endocannabinoid concentrations in mothers and children Bivariate Spearman rank correlations (rS) between endocannabinoid concentrations in hair sampled at t0 and t2 in N = 63 mothers and N = 45 children * p < .050,** p < .010,*** p < .001,two-tailed.All p-values were adjusted for multiple comparisons with the false discovery rate (FDR).data: a N = 17 mothers; b N = 23; no AEA data available for children.
Bivariate Spearman rank correlations (rS) between endocannabinoid concentrations in hair and covariates in N = 45 children Due to many AEA measures under technical detection limit, there is not sufficient AEA data in children available for analyses.Note: * p < .050,** p < .010,*** p < .001,two-tailed.Italic p-values indicate a trend for significance (p < .100).Underlined p-values are significant after correction with false discovery rate (FDR).

Table S8 .
Wilcoxon rank-sum test on differences in endocannabinoid concentrations in N = 63 mothers and N = 45 children with complete data at both points of measurement depending on maternal CM load t0

Table S9 .
Wilcoxon rank-sum test on differences in endocannabinoid concentrations in mothers and children depending on maternal CM load for the complete available sample t0