Women recruited into QF2011 were administered the same questionnaires, and had their medical information gathered, as described in the M@NGO protocol above. QF2011 added several questionnaires and procedures to the M@NGO protocol at recruitment, 36 weeks pregnancy, and at the 6-12 weeks and 6 months postpartum assessments. The QF2011 protocol also included a repeat of the flood-related questionnaires at 12 months post-flood (January 2012). In addition, QF2011 added procedures conducted at the birth, at the face-to-face assessments when the infants were 16 months and 2½ years of age, and will be conducting further assessments at age 4 years.
In 2008, we modified the STORM32 scale to assess objective stress exposure for a new project: The Iowa Flood Study [
89]. By taking into account experiences unique to flooding, and removing ice storm-related items, we created the Iowa Flood 100 (IF100) scale which included the same four domains as STORM32 (Threat, Loss, Scope and Change) with a possible 25 points per domain.
For QF2011, we added items to each domain of the IF100 questionnaire to more fully represent the women’s flood-related experiences, and allotted a maximum of 50 points per domain, resulting in the QFOSS (Queensland Flood Objective Stress Scale) with a maximum possible score of 200 points (Additional file
1). This scale was administered to all women at recruitment. It was re-administered at 12 months post-flood (January 2012) to update the values of financial loss, damage to house and property, experiences dealing with insurance companies, etc. Although recruitment and 12 month QFOSS scores are kept separate for some analyses, most analyses will be conducted with a final QFOSS score that integrates information from both. Because the IF100 is embedded within the QFOSS, a Queensland IF100 (QIF100) score can also be obtained for direct comparison with the Iowa study.
A number of women were recruited into QF2011 more than 10 months after the floods (n = ~50). Given that a 12-month post-flood questionnaire was planned, which repeated the administration of the recruitment questionnaire, the steering committee chose to send these women a single questionnaire at 12 months post flood rather than have them repeat the same questionnaire 2 months apart. As such, these women lacked a “recruitment IES-R” score. In order to impute these missing values, multiple regression was used to estimate their recruitment IES-R scores from post flood questionnaire responses. We used the dataset of women who had completed both the recruitment and the 12-month post flood questionnaires, and created separate algorithms for each IES-R sub-scale (Intrusions, Avoidance and Hyperarousal). Using this set of complete data, we regressed each recruitment IES-R sub-scale score on 12-month post-flood IES-R sub-scale scores, and other related scores that maximized the variance explained (STAI, PDI, PDEQ, MHC, etc.). The models explained 35.5 % (Avoidance), 44.2 % (Intrusions), and 49.1 % (Hyperarousal) of the variance in the recruitment scores. The regression coefficients from these results were used to create the equations we used for imputing the missing IES-R sub-scale scores for women who had not completed a Recruitment questionnaire. Once the sub-scales had been successfully imputed, IES-R Total scores were computed as usual by summing the sub-scales.
Maternal psychological health, psychosocial factors and other maternal measures
Maternal anxiety (recruitment, 12 months post-flood, 6-12 weeks, 6 months, and 2½ years)
The State-Trait Anxiety Inventory (STAI; [
97]) is a valid and reliable self-report measure of how one generally feels (trait) or currently feels (state). We administered the State scale of the STAI at recruitment, 12 months post flood, and at 6-12 weeks and 6 months. At 2½ years we administered the Trait scale of the STAI. Each scale has 20 items and participants are asked to rate statements (e.g., “I am happy” and “I lack self-confidence”) on a 4-point Likert scale (rated from Almost Never to Almost Always).
Depression (6-12 weeks and 6 months postpartum)
As noted in the M@NGO protocol section, the EPDS (Edinburgh Postpartum Depression Scale) was administered as part of M@NGO to assess maternal depression in pregnancy and again at 6-12 weeks and 6 months postpartum.
Depression, Anxiety and Stress Scales (16 months, 2½ and 4 years).
For maternal mental health at later assessments, we administered the 21-item short form of the Depression Anxiety Stress Scales (DASS-21 [
98-
100]) which has three independent scales (i.e., anxiety, stress, and depression) and has been used widely in perinatal samples.
Maternal depression (2½ years)
Maternal depression was also assessed with the Centre for Epidemiologic Studies Depression Scale
(CES-D; [
101]) screening questionnaire. This is commonly used to assess feelings of depression in the general population. Women rated 20 statements (e.g., “I felt that I was just as good as other people” and “I felt depressed”) for how they felt during the past week on a 4-point scale ranging from “Rarely or none of the time (<1 day)” to “Most or all of the time (5-7 days).”
Positive mental health (recruitment and 12 months post-flood)
Positive mental health was assessed with the 14-item Mental Health Continuum – Short Form (MHC-SF [
102]), which provides scores on Social, Emotional, and Psychological Well-Being. This scale has excellent psychometric properties. Extreme scores on this scale can define subgroups of “flourishing” and “languishing”.
Coping style (recruitment, 12 months post-flood, 16 months, 2½ years, and 4 years)
The Brief COPE includes 28 items: two items for each for the 14 coping strategies [
103]. The Brief R-COPE assesses the frequency of 14 religious coping strategies which load onto positive and negative religious coping factors [
104].
Prior trauma (12 months post-flood)
Because we have found that childhood trauma is associated with lower diurnal cortisol values in adulthood [
105], women were asked to complete the Childhood Trauma Questionnaire (CTQ; [
106,
107]), and the 24-item Trauma History Questionnaire (THQ; [
108]) when the second set of salivary cortisol samples was taken. The CTQ provides information concerning instances of emotional, sexual, and/or physical abuse and emotional and physical neglect occurring before age 18 years. The THQ assesses whether individuals have experienced traumatic events such as crime, disasters, and physical or sexual assault after the age of 18.
Life events (6 months and 16 months, 2½ years, and 4 years)
Because for some women, the flooding may have been the least of their problems during the pregnancy, or the floods may have compounded a host of other events, we need to control for this potential confound. The Life Experience Survey (LES; [
109]) lists 57 life changes, such as death of a spouse or a promotion at work; we reduced the number of items to the 29 more common occurrences. Respondents first indicate whether the event occurred or not, and then rate the impact of the event (if it occurred) on a 7-point Likert scale ranging from "Extremely Negative" to "Extremely Positive" and are asked to indicate the month and year of any events. Women indicate events that occurred in the preceding 24, 25, 12, and 18 months, respectively at the 6 month postpartum, 16 months, 2½ years and (forthcoming) 4 year assessments.
Maternal-rated scales of child development
Infant temperament (6-12 weeks, 6 months, 16 months, 2½ years, and 4 years)
The Short Temperament Scale for Infants (STSI) was developed as part of the ‘Australian Temperament Project: A Series of Studies of Australian Temperament, Development and Behavior in Australian Children’ [
121]. The STSI is a 30-item questionnaire in which parents rate the occurrence of common infant behaviors on a 6-point scale ranging from 1 (almost never) to 6 (almost always). The items yield five scales which measure the temperament dimensions of approach-withdrawal, rhythmicity, cooperation-manageability, activity-reactivity, and irritability. An overall “easy/difficult” score is calculated as the mean of the approach-avoidance, cooperation-manageability, and irritability scales. Infants scoring one standard deviation above and below the normative mean are classified as “difficult” and “easy” respectively. The STSI was administered to mothers at 6-12 weeks and 6 months. The companion Short Temperament Scale for Toddlers (STST; [
122]) was administered at 16 months and 2½ years, and we will administer the version for older children, the Short Temperament Scale for Children (STSC; [
122]), at 4 years.
Language abilities (16 months and 2½ years)
Language development was assessed using the short form of the MacArthur-Bates Communicative Development Inventory: Words and Sentences (MCDI; [
124]) at 16 months. The mother was given a list of 100 words and asked to indicate whether her child “understands” (receptive vocabulary) or “says” (productive vocabulary) each word. At 2
½ years an age-appropriate version of the MCDI-III was used that contains 100 words to assess the child’s productive vocabulary; the scale asks additional questions about how the child combines words into sentences and how they use language to ask questions, and about mean utterance length. This maternal report helps to avoid situational and temperamental factors that can affect children’s performance at face-to-face assessments, such as lack of interest or cooperation in the tasks, perhaps due to illness and/or anxiety [
125].
Infant handedness (16 months, 2½ years, and 4 years)
Following procedures described by Glover et al. [
116] we ask mothers which hand their child usually uses for the following five activities: drawing or coloring, throwing a ball, hitting things, stacking blocks, and using a spoon.
Social-emotional development (16 months)
The Brief Infant-Toddler Social Emotional Assessment (BITSEA; [
126]) is a 42-item questionnaire that aims to identify children at risk for, or currently experiencing, social-emotional/behavioral problems and/or delays in social-emotional competence. It addresses four domains: Externalizing, Internalizing, Dysregulation, and Competence. This tool allows mothers to provide information about how their children behave in different situations.
Face-to-face assessments: mother and child (16 month, 2½ years, and 4 years)
At 4 years, we plan to assess TOM using the Unexpected Contents Task [
142] which, similar to the Diverse Desires tasks, tests the child’s ability to make a judgment about what other characters know and what they are likely to do. The child is shown what is inside a distinctive container (e.g., pencils in a Smarties box) and has to judge what another character, who did not see what was in the box, believes to be inside the container.
Biological samples: mother, birth, and child
Data analysis and power
Our analyses of the QF2011 data will proceed in a highly structured manner, according to our underlying explanatory model. This model posits that objective stress and cognitive appraisal predict peritraumatic stress responses, which predict both subjective stress reaction (PTSD symptoms) and maternal HPA axis response. We hypothesize that at least some of these aspects of maternal stress may influence maternal mental health and may also influence variables within the placenta and fetus (cord blood); thereby influencing fetal/child physical/physiological development (growth & body composition, fingerprints or finger length ratios, HPA axis) and child neurodevelopment (cognitive, behavioral, and motor development). We further suggest that at each level of the cascade there may be genetic moderation and epigenetic mediation; and also moderation of outcomes by pre-disaster maternal characteristics (e.g., depression, socioeconomic status), and by the sex of the infant. Finally, our results may indicate a moderating effect from prenatal care (MGP versus standard care) and maternal caretaking behaviors. Below, we present a sampling of statistical approaches we will take to the data analysis.
The QF2011 study noted an initial formal withdrawal rate of 15.9 % between recruitment and the first face-to-face assessments of infants at 16 months of age, the withdrawal rate at the 2½ year assessment dropped to 3.7 %. Moreover, of the 191 families remaining in the study at 16 months, 91.6 % completed some portion of the assessment. At age 2½ years, 175 families provided data. Currently, a sample of 209 families can be contacted for the 4 year assessment which began in 2015. Using a conservative participation rate of 80 %, it is estimated that data will be available for approximately 167 families at this assessment.