Supporting routine cognitive reactivity assessment in the perinatal period: psychometric testing of the Chinese version of the Leiden Index of Depression Sensitivity

doi:10.21203/rs.3.rs-1462687/v1

PDF

Research Article

Supporting routine cognitive reactivity assessment in the perinatal period: psychometric testing of the Chinese version of the Leiden Index of Depression Sensitivity

https://doi.org/10.21203/rs.3.rs-1462687/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Finding optimal forms to recognize perinatal depression and its vulnerable factors and making them more applicable to depression screening is critical. This study aims to evaluate the reliability and validity of the Chinese version of he Leiden Index of Depression Sensitivity (LEIDS-RR) among perinatal women in China and determine cut-off values for screening high-risk depression.

Methods

In the third trimester of pregnancy and at six-weeks postpartum, women completed the LEIDS-RR-CV and a diagnostic reference standard online. We assessed the LEIDS-RR-CV using classical test theory and item response theory (IRT). The test performance for cut-off scores to screen for high-risk depression at each time point was assessed using receiver operator characteristic analysis.

Results

In total, 396 (third trimester) and 321 (six-weeks postpartum) women participated. Cronbach’s alpha, two-week test-retest reliability, and marginal reliability for the scale were all greater than 0.8. It showed a five-factor model, and the cut-off values were 58 (third trimester) and 60 (six-weeks postpartum). The areas under the curve were acceptable (≥ 0.7), and the LEIDS-RR-CV was positively correlated with the total Edinburgh Postnatal Depression Scale (EPDS) score (r = 0.52 and 0.56, p = 0.00), indicating its predictive validity. IRT analysis confirmed its discriminative validity.

Conclusions

The LEIDS-RR-CV was shown to be reliable, valid, and can be used to quantify cognitive reactivity in perinatal Chinese women and for screening for high-risk depression during this period.

cognitive reactivity

perinatal care

depression

reliability and validity

item response theory

longitudinal study

The perinatal period represents a vulnerable time for developing mental health disorders, including perinatal depression[1]. In the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders, PND is defined as the occurrence of a major depressive episode during pregnancy (i.e., antenatal depression: onset during pregnancy) or within four weeks after childbirth (i.e., postpartum depression [PPD]), with the use of the specifier ‘‘with peripartum onset’’ for defining the depressive disorder[2] (American Psychological Association 2013).

During last decade, the prevalence of PND has shown a significant increasing trend. In 2020, a systematic review in mainland China, focusing on the prevalence of PND and its determinants, reported a 16.3% pooled prevalence of PND: 19.7% for antenatal depression and 14.8% for PPD [3]. PND is associated with various adverse health outcomes for both mothers and babies[1, 4]. For mothers, these include increased risk of preeclampsia, pregnancy and labor complications, and infanticide or suicide. For infants and newborns, these include preterm birth, low birth weight, and poor cognitive and emotional development [1, 4].

PND remains under-detected and under-treated worldwide, especially across developing countries [5, 6]. There is limited available data on diagnostic and treatment rates for PND in China. However, with the implementation of the three-child policy in China and the release of the “explore the work plan of characteristic services for depression prevention and treatment” blueprint, stakeholders have started to pay more attention to PND [7]. Thus, in line with prior research [8], there is a critical need to find better ways to recognize PND and its risk factors, and to make it more amenable to depression screening measures, especially during the risk period (prenatal period to one-year postpartum).

According to Beck’s cognitive model, one’s vulnerability to depression is marked by schemas or dysfunctional attitudes[9]. Teasdale (1988) explained that negative cognitions remain latent in some individuals but may be (re)activated by life events, stress, or even negative moods[10]. This concept of negative cognitions being (re)activated by negative moods is known as cognitive reactivity [11]. Studies have demonstrated that CR is an important cognitive vulnerability factor for the onset and maintenance of depressive symptoms, showing a moderate to high predictive power[12, 13]. Some researchers have used Dysfunctional Attitudes Scale (DAS) to explore the relationship between dysfunctional attitudes and PPD, having found that perinatal women with high dysfunctional attitudes are more vulnerable to PPD[14, 15] and such attitudes are indirectly related to symptoms of depression and anxiety in the infants[16]. However, the DAS has limitations that may have hindered the generalization of their findings[17].

To compensate for these, researchers have used the Leiden Index of Depression Sensitivity (LEIDS)—a self-reported inventory of CR without the use of mood induction[11]. This scale was later revised to LEIDS-R[18] and LEIDS-RR[19], the latter having been translated into several languages and validated in many countries[13, 20], including China[12]. The Chinese version of the LEIDS-RR (LEIDS-RR-CV) is a reliable and valid instrument for measuring CR in Chinese patients with clinical depression in remission and in the healthy Chinese population [21].

To our knowledge, there have been no studies devoted to conducting a psychometric analysis of the LEIDS-RR and applying it to perinatal women, especially in China. This study aims to examine the psychometric performance of the LEIDS-RR-CV and explore its applicability to detect women at high risk of PND. Specifically, we used classical test theory (CTT) and item response theory (IRT) for the psychometric analyses.

2.1. Participants and procedure

This prospective study was conducted between September 2020 and March 2021. Using convenience sampling, we recruited pregnant women who met the standard of discharge from the obstetrical clinic and obstetrical inpatient department of four tertiary (level 3) hospitals in Fuzhou, Putian, and Quanzhou City, Fujian Province. Inclusion criteria were (1) ages ≥ 20 years, (2) in late pregnancy (i.e., gestational age ≥ 28 weeks) and with fetal survival diagnosed by type-B ultrasonic test, (3) having no depression (i.e., a score ≤ 12 on the 10-item Edinburgh Postnatal Depression Scale), (4) the possibility of being followed-up for up to six weeks postpartum, and (5) choosing to voluntarily participate in the study and signing the informed consent form. Exclusion criteria were (1) having mental disorders, (2) undergoing mental health treatment, and (3) having cognitive impairment. We determined the sample size based on a subject-to-item ratio of 5–10:1, assuming a non-response rate of 20%. This led to a final sample size of at least 320 women. Twenty-four additional eligible women were included at both time points when we conducted the assessments (i.e., in the third trimester of pregnancy [hereinafter, the third trimester] and at six-weeks postpartum), with an interval of two weeks. We calculated the correlation coefficients between the test scores at both time points.

Eligible women were asked to complete the study measures at both time points. They partook in both surveys through Wenjuanxing online platform (a popular online survey platform in China, https://www.wjx.cn/m/91639653.aspx / https://www.wjx.cn/m/91655868.aspx.). At each time, women were able to opt out of participation.

2.2. Measures

Women completed the following measures in the survey questionnaire at both assessment points.

2.2.1. The LEIDS-RR-CV

We used the 26-item self-reported LEIDS-RR-CV to measure CR. It comprises five subscales: hopelessness/suicidality, acceptance/coping, aggression, control/perfectionism, and avoidant coping. Items are rated on a five-point Likert scale ranging from 1–5 (“not at all” to “very strongly”). After reverse scoring the acceptance/coping subscale and summing up the scores for each item, higher total scores indicated greater CR. The LEIDS-RR-CV has been validated and has good psychometric properties [21]; Cronbach’s alpha was 0.92.

2.2.2. Edinburgh Postnatal Depression Scale

The 10-item self-reported Edinburgh Postnatal Depression Scale (EPDS) is the most used screening tool for common depression symptoms in pre- and postpartum women. Items are responded on a four-point scale ranging from 0–3, with higher total scores indicating more symptoms of depression (range: 0–30) [22]. The EPDS has been validated in many countries and has good psychometric properties in both pre- and postpartum contexts[23, 24]. When assessing potential for PND, a cut-off score of 13 or more is recommended[22].

2.2.3. Sociodemographic and clinical characteristics

We collected sociodemographic and clinical information of participants from the administrative database of the participating hospital; data included telephone number, maternal age, education level, only child or not, residential location, city, pregnancy intention, monthly household income (yuan, RMB), gravidity, parity, gestational week, and expected due date.

2.3. Ethics

The study was approved by the appropriate ethical committee; written informed consent was obtained from all participants. All methods in the study were performed in accordance with relevant institutional guidelines.

2.4. Statistical Analysis

Analyses were conducted using IBM SPSS Statistics, version 26.0, AMOS 25.0, and Stata 14.0. If α = 0.05, we deemed the difference as statistically significant (p < 0.05). We conducted the psychometric analysis of LEIDS-RR-CV based on CTT and IRT for both assessment time points.

CTT-based psychometric analysis

Regarding reliability, we used Cronbach’s alpha of the total scale to assess internal consistency. Test-retest reliability was evaluated using intraclass correlation coefficient (ICC).

Regarding validity, we conducted confirmatory factor analysis (CFA) to analyze the structural validity of the tool using maximum likelihood method. We evaluated the model’s goodness-of-fit using absolute and relative indices: normed χ2 (χ2/df; 1.0–3.0) and root mean square error of approximation (RMSEA; <0.08); goodness-of-fit index (GFI), adjusted goodness-of-fit index (AGFI), comparative fit index (CFI), and Tucker–Lewis Index (TLI) all being > 0.9[25].

Regarding criterion-related validity, we conducted Spearman correlation analysis between the total score of the EPDS and LEIDS-RR-CV. For known-group validity, we used independent t-tests to calculate the mean scores between participants with PPD (EPDS > 13) and those with no depressive symptoms (EPDS ≤ 12). We also used propensity score matching to select 50 (out of 396 participants in the third trimester) and 42 (out of 328 participants at six-weeks postpartum) non-depressed women to compare with women with depression.

Based on the EPDS criteria, we determined predictive validity by calculating the receiver operating characteristic (ROC), area under the curve (AUC), sensitivity, specificity, positive predictive value, negative predictive value, and percentage. These outcomes were also used to determine the optimal cut-off points for LEIDS-RR-CV in childbearing and postpartum women.

IRT-based psychometric analysis

We used Samejima’s graded response model for IRT analysis[26]. After meeting the assumption of unidimensionality[27], we calculated the discrimination parameter (a_i), difficulty parameters (β_i), item information value, item characteristic curves, and test information functions. Finally, we judged reliability by marginal reliability, the amount of information provided by individual items, and the entire scale.

3.1 Participants’ characteristics

414 women participated in this study in total. Of these, 396 women (20–42 years old) completed the questionnaire in the third trimester, and 321 women (20–42 years old) completed the questionnaire at six-weeks postpartum. Overall, the ratio of women who met the EPDS criteria for depression at the two time points were 13.1% and 13.4%, respectively. Additional sociodemographic and clinical characteristics of the participants are summarized in Table 1.

Table 1

The socio-demographic and clinical characteristics of participants
Variables	The third trimester women (n = 396) n(%)	The 6-week postpartum women (n = 321) n(%)
Educational level
Junior high school or below	53(13.4)	43(13.4)
Middle school or Technical school	63(15.9)	47(14.6)
College or above	280(70.7)	231(72.0)
Only child or not
Yes	72(18.2)	59(18.4)
No	324(81.8)	262(81.6)
Residential location
Urban	296(74.7)	246(76.6)
Rural	100(25.3)	75(23.4)
City
Fuzhou	174(43.9)	147(45.8)
Putian	71(17.9)	52(16.2)
Quanzhou	132(33.3)	115(35.8)
Other^*	19(4.8)	7(2.2)
Monthly household income (yuan, RMB)
< 3000	21(5.3)	14(4.4)
3000 ~ 4999	114(28.8)	89(27.7)
5000 ~ 8000	158(39.9)	136(42.4)
> 8000	103(26.0)	82(25.5)
Gravidity
1	173(43.7)	145(45.2)
2	153(38.6)	132(41.1)
≥ 3	70(17.7)	44(13.7)
Parity
0	201(50.8)	168(52.3)
1	157(39.6)	135(42.1)
≥ 2	38(9.6)	18(5.6)
Pregnancy intention
In plan	262(66.2)	217(67.6)
Out of plan	134(33.8)	104(32.4)
Note: Other* include Xiamen, Ningde, Sanming, Longyan, Zhangzhou, Nanping and other cities.

3.2 Reliability

Data from women who responded at both time points had a total Cronbach’s alpha of 0.928 and 0.982, respectively; marginal reliability of 0.943 and 0.981, respectively; and ICC of 0.934 and 0.862, respectively, with the result showing that the test-retest reliability of the total scale was good.

3.3 Structural validity

Through CFA, we observed that a five-factor structure was a good fit to the data at both time points (χ2/df = 2.511 and 2.586, respectively; GFI = 0.861 and 0.840, respectively; AGFI = 0.830 and 0.803, respectively; RMSEA = 0.066 and 0.070, respectively; CFI = 0.925 and 0.963, respectively; TLI = 0.915 and 0.957, respectively). The factors were (a) hopelessness/suicidality, (b) acceptance/coping, (c) aggression, (d) control/perfectionism, and (e) avoidant coping (Fig. 1).

3.4 Criterion-related validity

The total score for the LEIDS-RR-CV showed a positive correlation with the total score for the EPDS (r = 0.52 and 0.56 in the third trimester and at six-weeks postpartum, respectively; p = 0.00).

3.5 Known-group validity

Women with depression showed higher total scores for the LEIDS-RR-CV (third trimester: 69.08 ± 11.68; six-weeks postpartum: 74.88 ± 15.74) than women with no depression (53.68 ± 11.73 and 61.45 ± 14.84, respectively). The difference between the two groups was large and significant (t = 6.58 and 4.02, respectively; p < 0.001).

3.6 Predictive validity

The AUC of the ROC for the LEIDS-RR-CV of the data from participants at both time points was 0.861 (95% confidence interval [CI]: 0.815–0.908) and 0.691 (95% CI: 0.612–0.770), respectively (Fig. 2), indicating good predictive power for depression.

In the third trimester, the optimal cut-off value for the LEIDS-RR-CV to screen for high-risk groups for depression was 58 points, and the sensitivity and specificity were 0.904 and 0.730, respectively. At six-weeks postpartum, the optimal cut-off value was 60 points, and the sensitivity and specificity were 0.767 and 0.540, respectively.

3.7 IRT analysis of LEIDS-RR-CV

We observed a support for the unidimensionality assumption based on the percentage of variance accounting for the first factor (i.e., 37.59%: third trimester; 69.19%: six-weeks postpartum), as well as the eigenvalues of the first factor divided by the second, which were 3.59 and 13.84. Parametric estimates (a, β₁, β₂, β₃, β₄) for the 26 items from the IRT analysis are presented in Table 2.

Table 2

IRT parameter estimates for the LEIDS-RR-CV for perinatal women
Item	The third trimester women							The 6-week postpartum women
		Slopeα(SE)	Difficulty				Maximum value of TIF	Mean value of TIF*	Slope α(SE)	Difficulty				Maximum value of TIF	Mean value of TIF*
		Slopeα(SE)	β₁ (SE)	β₂ (SE)	β₃ (SE)	β₄ (SE)	Maximum value of TIF	Mean value of TIF*	Slope α(SE)	β₁ (SE)	β₂ (SE)	β₃ (SE)	β₄ (SE)	Maximum value of TIF	Mean value of TIF*
I1	1.12 (0.14)	-1.01 (0.16)	1.43 (0.19)	3.40 (0.45)	4.44 (0.66)	0.38	0.29	2.06 (0.18)	-1.22 (0.12)	0.06 (0.09)	1.08 (0.11)	2.32 (0.21)	1.22	0.77
I8	1.27 (0.16)	-0.48 (0.12)	1.27 (0.17)	3.24 (0.41)	/	0.46	0.33	2.45 (0.21)	-0.95 (0.10)	0.13 (0.09)	1.11 (0.10)	2.29 (0.20)	1.70	0.96
I11	1.65 (0.19)	0.01 (0.10)	1.44 (0.15)	2.58 (0.28)	3.82 (0.55)	0.81	0.49	2.44 (0.21)	-0.79 (0.09)	0.14 (0.09)	1.19 (0.11)	2.30 (0.20)	1.70	0.93
I24	1.46 (0.19)	0.64 (0.11)	1.92 (0.22)	3.09 (0.38)	4.17 (0.65)	0.65	0.36	2.67 (0.24)	-0.42 (0.08)	0.20 (0.09)	1.33 (0.11)	2.39 (0.22)	2.11	1.00
I25	1.03 (0.13)	-1.23 (0.19)	0.73 (0.15)	2.28 (0.29)	4.17 (0.60)	0.32	0.26	1.93 (0.18)	-1.04 (0.12)	-0.17 (0.09)	1.06 (0.11)	2.31 (0.21)	1.11	0.68
I4	1.85 (0.23)	0.51 (0.09)	1.92 (0.22)	3.91 (0.65)	/	0.93	0.49	2.90 (0.26)	-0.65 (0.08)	0.22 (0.09)	1.30 (0.10)	2.46 (0.23)	2.33	1.19
I13	1.93 (0.26)	1.00 (0.12)	2.10 (0.22)	3.45 (0.49)	/	1.07	0.50	3.06 (0.27)	-0.48 (0.08)	0.30 (0.09)	1.38 (0.11)	2.50 (0.24)	2.61	1.26
I16	2.29 (0.25)	0.20 (0.08)	1.50 (0.14)	2.39 (0.23)	/	1.52	0.67	2.89 (0.26)	-0.62 (0.08)	0.22 (0.09)	1.42 (0.11)	2.38 (0.21)	2.31	1.17
I17	2.02 (0.23)	0.09 (0.09)	1.44 (0.14)	2.41 (0.24)	3.32 (0.44)	1.22	0.67	2.51 (0.22)	-0.79 (0.09)	0.24 (0.09)	1.32 (0.11)	2.91 (0.34)	1.75	1.02
I20	2.24 (0.27)	0.63 (0.09)	1.94 (0.18)	2.98 (0.35)	/	1.41	0.65	2.99 (0.27)	-0.56 (0.08)	0.23 (0.08)	1.42 (0.11)	2.43 (0.22)	2.49	1.22
I3	2.34 (0.25)	0.02 (0.08)	1.30 (0.12)	2.33 (0.21)	3.12 (0.38)	1.61	0.84	3.94 (0.37)	-0.71 (0.08)	0.17 (0.08)	1.27 (0.09)	2.05 (0.16)	4.10	1.73
I5	2.89 (0.39)	1.00 (0.10)	1.78 (0.16)	/	/	2.33	0.59	3.65 (0.33)	-0.51 (0.08)	0.34 (0.08)	1.37 (0.10)	2.18 (0.18)	3.57	1.56
I12	2.44 (0.26)	0.08 (0.08)	1.17 (0.11)	2.11 (0.19)	3.03 (0.36)	1.71	0.89	3.64 (0.33)	-0.68 (0.08)	0.27 (0.08)	1.31 (0.10)	2.10 (0.17)	3.56	1.57
I23	2.55 (0.32)	0.90 (0.10)	1.65 (0.15)	2.71 (0.28)	/	1.90	0.72	4.71 (0.47)	-0.08 (0.07)	0.42 (0.08)	1.59 (0.11)	2.16 (0.18)	6.13	1.98
I26	2.98 (0.40)	1.01 (0.10)	1.86 (0.16)	2.83 (0.32)	/	2.47	0.90	4.76 (0.49)	0.01 (0.07)	0.41 (0.08)	1.58 (0.11)	2.16 (0.18)	6.52	1.94
I2	0.94 (0.13)	-0.74 (0.17)	1.05 (0.18)	3.58 (0.51)	5.88 (1.06)	0.26	0.21	2.72 (0.23)	-0.99 (0.10)	0.05 (0.08)	1.02 (0.10)	2.10 (0.17)	2.05	1.10
I6	0.97 (0.13)	-1.26 (0.21)	1.02 (0.17)	3.81 (0.54)	5.71 (1.01)	0.27	0.23	2.94 (0.25)	-1.02 (0.10)	0.06 (0.08)	0.95 (0.09)	2.13 (0.17)	2.38	1.23
I14	0.89 (0.12)	-1.44 (0.23)	0.51 (0.15)	2.64 (0.37)	4.70 (0.73)	0.24	0.21	2.28 (0.20)	-1.08 (0.11)	0.01 (0.09)	0.93 (0.10)	2.31 (0.20)	1.51	0.88
I22	1.00 (0.14)	-0.78 (0.16)	1.16 (0.19)	4.39 (0.67)	6.22 (1.26)	0.29	0.22	2.77 (0.24)	-0.76 (0.09)	0.05 (0.08)	1.12 (0.10)	2.24 (0.19)	2.16	1.10
I7	1.77 (0.18)	-0.56 (0.11)	0.80 (0.11)	2.34 (0.23)	3.21 (0.38)	0.94	0.59	3.04 (0.26)	-0.98 (0.09)	0.04 (0.08)	0.92 (0.09)	2.33 (0.20)	2.54	1.29
I9	2.55 (0.26)	-0.25 (0.08)	1.14 (0.11)	2.10 (0.18)	2.99 (0.35)	1.86	0.98	3.08 (0.27)	-0.89 (0.09)	0.11 (0.08)	1.18 (0.10)	2.52 (0.24)	2.54	1.33
I10	2.25 (0.24)	-0.10 (0.08)	1.19 (0.12)	2.25 (0.21)	3.46 (0.52)	1.43	0.82	3.31 (0.29)	-0.76 (0.08)	0.21 (0.08)	1.14 (0.09)	2.34 (0.21)	2.94	1.43
I15	2.14 (0.22)	-0.46 (0.09)	1.05 (0.11)	2.35 (0.22)	3.56 (0.54)	1.27	0.79	3.75 (0.33)	-0.74 (0.08)	0.16 (0.08)	1.10 (0.09)	2.17 (0.17)	3.73	1.66
I18	2.11 (0.22)	-0.20 (0.09)	1.21 (0.12)	2.17 (0.20)	3.06 (0.36)	1.32	0.74	3.57 (0.32)	-0.74 (0.08)	0.18 (0.08)	1.14 (0.09)	2.49 (0.25)	3.39	1.59
I19	1.91 (0.20)	-0.70 (0.11)	1.06(0.12)	2.51 (0.25)	/	0.99	0.60	3.66 (0.33)	-0.81 (0.08)	0.14 (0.08)	1.14 (0.09)	2.34 (0.21)	3.52	1.64
I21	2.07 (0.20)	-0.88 (0.11)	0.69 (0.10)	1.96 (0.18)	2.78 (0.30)	1.27	0.76	3.57 (0.31)	-0.86 (0.09)	0.09 (0.08)	1.07 (0.09)	2.05 (0.16)	3.38	1.56
Note. "/"indicates that only four options are selected and the other option is not selected.
TIF, test information function.
*Mean test information function of seven categories, that is, categories − 3, −2, − 1, 0, 1, 2, 3.
curve of the LEIDS-RR-CV in women curve of the LEIDS-RR-CV in women
of late pregnancy of 6 weeks postpartum

89% and 100% (in the third trimester and at six-weeks postpartum, respectively) of the items showed very high discrimination (≥ 1). The β_ik values for all items were between − 1.44 (item 14) and 6.22 (item 22) in the third trimester and between − 1.22 (item 1) and 2.91 (item 17) at six-weeks postpartum, indicating a broad range of information at both time points. The results also revealed no disordinal nor reversal of β_ik. Moreover, TIF maximum values ranged from 0.178 (item 2) to 2.81 (item 16) in the third trimester and from 1.110 (item 25) to 6.515 (item 26) at six-weeks postpartum. Further, most (92.30%) item characteristic curves were well shaped; the peaks of the five curves did not overlap; and curves 2, 3, and 4 were normally distributed (Appendices A and B).

To our knowledge, this was the first study to validate the LEIDS-RR-CV in a sample of perinatal women. We conducted a comprehensive psychometric evaluation based on CTT and IRT of the LEIDS-RR-CV that confirmed the predictive effect of CR on depression and that the scale has satisfactory structural, criterion-related, known-group, and predictive validity, robust internal consistency, and test-retest reliability. Therefore, the 26-item LEIDS-RR-CV can serve as a valid and reliable scale to quantify CR in perinatal women in China, and for screening women at high-risk for depression during perinatal periods.

In our study, the LEIDS-RR-CV showed a Cronbach’s alpha, ICC, and marginal reliability with values > 0.7, indicating a satisfactory internal consistency and temporal stability. Further, through CFA, we confirmed that a five-factor structure provided a good fit to the data, similar to the results of the original version of the scale[19] and those of the Chinese version, which was validated on undergraduate students and depression remission women[21, 28]. However, our results differed from those of the Spanish version of the LEIDS-RR, which presented a four-factor model[20], and the Persian version, which involved a sample of patients with depression from the general population that showed a six-factor model[13]. Thus, our finding extends the scientific evidence on the application of the LEIDS-RR in different samples and cultural contexts. Specifically, we showed that the LEIDS-RR-CV can be used to measure CR and its subscales (i.e., hopelessness/suicidality, acceptance/coping, aggression, control/perfectionism, and avoidant coping) in perinatal Chinese women.

As expected, the total score for CR was higher in women with depression than in women without depression. This was also confirmed by the IRT analysis [29]: the LEIDS-RR-CV has high discrimination and performed well over diverse individuals with low and high CR. The findings showed that the LEIDS-RR-CV has high discrimination and is expected to perform well when applied to perinatal women with various levels of CR.

We observed a significant correlation between our data for the LEIDS-RR-CV and the scores for the EPDS, indicating good criterion-related validity. When we used the LEIDS-RR-CV to predict future depression across the perinatal period, it showed acceptable test performance (i.e., AUC values > 0.7). Therefore, the 26-item LEIDS-RR-CV can be used to identify women at high-risk for depression and to predict future depression during the perinatal period, thus being clinically useful for stakeholders interested in the accurate assessment of CR in this population.

To our knowledge, our study first attempts to establish optimal cut-off scores on the LEIDS-RR-CV among perinatal women; the optimal assessment scores were 58 and 60 points for the third trimester and the six-weeks postpartum, respectively. Thus, women assessed in the third trimester, showing a total score for CR ≥ 58, were at a higher risk for depression than those with results below this value; at six-weeks postpartum, this cut-off value was ≥ 60. Specifically, our results showed that the cut-off score for this scale varied in individuals according to their particular context [30]. We suggest future scholars explore how the LEIDS-RR-CV performs when used in different clinical settings and subgroups. Our findings on the optimal cut-off score of CR in perinatal women may improve the ability of healthcare professionals to quickly recognize patients with depressive thoughts and those at risk of clinical PND. Knowledge on this index may also prove helpful in developing referral, intervention, and treatment strategies to reduce the incidence of PND.

4.1 Limitations

First, there were some characteristics of our research that may have introduced sample bias and affected the generalizability of our results, that is, the use of convenience sampling; all women in our sample were recruited from Fujian province, southern China; and we used data collected only through online self-reported questionnaires. Second, additional evidence on the validity of the scale is needed, particularly regarding its measurement invariance; hence, differential item function analysis may be an important next step to confirm that the items in the scale function similarly across different groups of women. Finally, although we established a cut-off value for CR to identify groups of women at high risk for depression at both the third trimester and six-weeks postpartum, we have not defined the cut-off values for mild, moderate, and severe CR scores—a suggestion for future research.

4.2 Implications

This study attempts to make a valid tool for indirectly predicting depression to aid healthcare providers or public health service providers manage, specifically by prevention, perinatal depression in China, where there seems to be no such tool to predict this construct. We recommend testing and applying the LEIDS-RR with a larger representative group of perinatal women in different countries, then comparing CR between western and Chinese groups. The LEIDS-RR will facilitate the design of psychological interventions aimed at curbing CT in perinatal women and is crucial to the effective development and evaluation of interventions to decrease CR, even PND, in the cross-cultural context.

This study extended the psychometric analysis and scientific evidence on the applicability of the LEIDS-RR-CV in perinatal Chinese women. The scale showed adequate reliability and validity, and can be used to measure CR to predict PND in perinatal women and to screen for women at high risk for depression.

Ethics approval and consent to participate

All study participants provided written informed consent, and the study design was approved by the ethics committee at Fujian Medical University (No 2017024).

Consent for publication

Not applicable.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests

We affirm that we do not have any financial or other interest in the results of the study.

Acknowledgements

We gratefully acknowledge all women participants, without them, it is not possible to complete the study. We also sincerely thank the English editing services from Editage.

Funding

This research was supported by the Startup Fund for High-level Talents of the Fujian Medical University, with grant number [XRCZX2017011] and General Project of Fujian Provincial Nature Science Foundation, with grant number [2021J01133126].

Authors' contributions

FFH: study design, scale validation, revision and writing of the manuscript

YQF: data analysis, and writing of the manuscript.

MFZ and YAL: validation of the scale.

JHZ: conducted data analysis.

HLH, SQH and JL: data collection and proofreading of the paper.

Dagher RK, Bruckheim HE, Colpe LJ, Edwards E, White DB: Perinatal Depression: Challenges and Opportunities. J Womens Health (Larchmt) 2021, 30(2):154–159.
American Psychological Association (2013) Diagnostic and statistical manual of mental disorders, 5th ed. Washington, DC: American Psychological Association.
Nisar A, Yin J, Waqas A, Bai X, Wang D, Rahman A, Li X: Prevalence of perinatal depression and its determinants in Mainland China: A systematic review and meta-analysis. J Affect Disord 2020, 277:1022–1037.
Howard LM, Khalifeh H: Perinatal mental health: a review of progress and challenges. World Psychiatry 2020, 19(3):313–327.
Nillni YI, Mehralizade A, Mayer L, Milanovic S: Treatment of depression, anxiety, and trauma-related disorders during the perinatal period: A systematic review. Clin Psychol Rev 2018, 66:136–148.
Vesga-López O, Blanco C, Keyes K, Olfson M, Grant BF, Hasin DS: Psychiatric disorders in pregnant and postpartum women in the United States. Arch Gen Psychiatry 2008, 65(7):805–815.
National Health Commission. Notice of the general office of the National Health Commission on exploring and developing special services for the prevention and treatment of depression and Alzheimer's disease (2020) http://www.nhc.gov.cn/jkj/s7914/202009/a63d8f82eb. Accessed 27 November 2021.
O’Connor E, Senger CA, Henninger M, Gaynes BN, Coppola E, Soulsby Weyrich M: U.S. Preventive Services Task Force Evidence Syntheses, formerly Systematic Evidence Reviews. In: Interventions to Prevent Perinatal Depression: A Systematic Evidence Review for the US Preventive Services Task Force. edn. Rockville (MD): Agency for Healthcare Research and Quality (US); 2019.
Beck AT (1967) Depression: Clinical, Experimental, and Theoretical Aspects. Harper and Row, New York.
Teadale JD (1988) Cognitive vulnerability to persistent depression. Cogn. Emot. 2,247–274.
Van der Does W: Cognitive reactivity to sad mood: structure and validity of a new measure. Behav Res Ther 2002, 40(1):105–120.
Huang FF, Wen ZP, Li Q, Chen B, Weng WJ: Factors influencing cognitive reactivity among young adults at high risk for depression in China: a cross-sectional study. BMC Public Health 2020, 20(1):703.
Ostovar S, Md Nor MB, Griffiths MD, Chermahini SA: Cognitive Reactivity: Cultural Adaptation and Psychometric Testing of the Persian Version of the Leiden Index of Depression Sensitivity Revised (LEIDS-R) in an Iranian Sample. Int J Ment Health Addict 2017, 15(3):621–638.
Fonseca A, Monteiro F, Canavarro MC: Dysfunctional beliefs towards motherhood and postpartum depressive and anxiety symptoms: Uncovering the role of experiential avoidance. J Clin Psychol 2018, 74(12):2134–2144.
Fonseca A, Canavarro MC: Cognitive correlates of women's postpartum depression risk and symptoms: the contribution of dysfunctional beliefs and negative thoughts. J Ment Health 2020, 29(6):614–622.
Dunning EE, McArthur BA, Abramson LY, Alloy LB: Linking Maternal Depression to Adolescent Internalizing Symptoms: Transmission of Cognitive Vulnerabilities. J Youth Adolesc 2021, 50(2):324–335.
Figueroa CA, Mocking RJT, Mahmoud GA, Koeter MW, Bockting CL, van der Does W, Ruhe HG, Schene AH: The measurement of cognitive reactivity to sad mood in patients remitted from major depressive disorder. Br J Clin Psychol 2018, 57(3):313–327.
Van rDoes AJW, Williams, JMG (2003) Leiden index of depress sensitivity–revised (LEIDS-R).From http://www.dousa.nl/publications_depression.htm#LEIDS.
Solis E, Antypa N, Conijn JM, Kelderman H, Van der Does W: Psychometric properties of the Leiden Index of Depression Sensitivity (LEIDS). Psychol Assess 2017, 29(2):158–171.
Senín-Calderón C, Perona-Garcelán S, Ruíz-Veguilla M, Rodríguez-Testal JF: Leiden Index of Depression Sensitivity-Revised (LEIDS-R): Spanish validation proposal. Int J Clin Health Psychol 2017, 17(2):139–150.
Huang FF, Li H, Lei Y, Li Q, Chen B: Adaptation and validation of the Chinese version of the modified Leiden index of depression sensitivity. J Affect Disord 2019, 256:458–467.
Cox JL, Holden JM, Sagovsky R: Detection of postnatal depression. Development of the 10-item Edinburgh Postnatal Depression Scale. Br J Psychiatry 1987, 150:782–786.
Levis B, Negeri Z, Sun Y, Benedetti A, Thombs BD: Accuracy of the Edinburgh Postnatal Depression Scale (EPDS) for screening to detect major depression among pregnant and postpartum women: systematic review and meta-analysis of individual participant data. Bmj 2020, 371:m4022.
Shrestha SD, Pradhan R, Tran TD, Gualano RC, Fisher JR: Reliability and validity of the Edinburgh Postnatal Depression Scale (EPDS) for detecting perinatal common mental disorders (PCMDs) among women in low-and lower-middle-income countries: a systematic review. BMC Pregnancy Childbirth 2016, 16:72.
Brunton RJ, Dryer R, Saliba A, Kohlhoff J: The initial development of the Pregnancy-related Anxiety Scale. Women Birth 2019, 32(1):e118-e130.
Hambleton RK, Linden W (1997) Handbook of Modern Item Response Theory[M]. City: Handbook of Modern Item Response Theory.
Hambleton RK (1991) Fundamentals of Item Response Theory[J]. BMS bulletin of sociological methodology.
Huang FF, Chen WT, Lin YA, Hong YT, Chen B: Cognitive reactivity among high-risk individuals at the first and recurrent episode of depression symptomology: A structural equation modelling analysis. Int J Ment Health Nurs 2021, 30(1):334–345.
Costa DS, Dieng M, Cust AE, Butow PN, Kasparian NA: Psychometric properties of the Fear of Cancer Recurrence Inventory: an item response theory approach. Psychooncology 2016, 25(7):832–838.
Reilly N, Hadzi-Pavlovic D, Loxton D, Black E, Mule V, Austin MP: Supporting routine psychosocial assessment in the perinatal period: The concurrent and predictive validity of the Antenatal Risk Questionnaire-Revised. Women Birth 2021.

No competing interests reported.

appendix.docx

PDF

Reviewers invited by journal
19 May, 2022
Editor assigned by journal
12 May, 2022
Editor invited by journal
25 Mar, 2022
Submission checks completed at journal
25 Mar, 2022
First submitted to journal
17 Mar, 2022

You are reading this latest preprint version

Supporting routine cognitive reactivity assessment in the perinatal period: psychometric testing of the Chinese version of the Leiden Index of Depression Sensitivity

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

1. Background

2. Methods