Cultural Evolution

Conceptually, the KIDSCREEN instruments are based on the definition of QoL as a multidimensional construct covering physical, emotional, mental, social, and behavioral components of well-being and functioning as perceived by patients and/or other individuals. The KIDSCREEN project used a simultaneous approach to include 13 European countries in the cross-cultural harmonization and development of the measures. Content for the KIDSCREEN questionnaire was generated from a literature review [5 ], a Delphi exercise with experts in QoL measurement in children [11 (link)], and focus groups with children and parents [12 (link)]. Focus group work in the participating European countries led to the formulation of 2,505 statements which formed the original pool of possible items for the questionnaire. After an item reduction process involving redundancy rating and card sorting (Fig. 1), 179 items were selected to form the basis of a draft questionnaire for pilot testing. Administration in a pilot study with 3,019 children in seven European countries provided data which allowed for further item reduction using a combination of classical test theory (CTT) and item response theory (IRT) so as to define the final and definitive version of 52 items covering 10 dimensions of QoL [6 (link), 13 (link)]. From this version, the KIDSCREEN-27 was produced using basic item analyses, confirmatory and explorative factor analyses, and IRT [8 (link), 9 (link)] and the KIDSCREEN-10 was developed in turn from KIDSCREEN-27 using Rasch analysis [10 (link)].
Fig. 1

Flowchart showing development process of the KIDSCREEN tool

All three KIDSCREEN questionnaires were psychometrically tested using data obtained in a multicenter European study which included a sample of 22,827 children recruited in 13 countries [14 (link)]. Participants completed the KIDSCREEN-52 together with one or more other QoL instruments for children and adolescents, such as the pediatric quality of life inventory (PedsQL) [15 (link)], Child Health and Illness Profile-Adolescent Edition (CHIP-AE—in children aged 12 years and over) [16 ] or the youth quality of life instrument—surveillance version (YQOL-S) [17 (link)]. The reliability and validity of the 52-, 27-, and 10-item versions of KIDSCREEN were tested primarily using a CTT approach, though Rasch analysis was also used. Test–retest reliability was assessed in approximately 10 % of the overall sample by administering the questionnaire on two occasions 2 weeks apart. The instruments’ convergent and known groups’ validity was tested by examining correlations with similar instruments and investigating whether KIDSCREEN-27 and KIDSCREEN-52 discriminated between groups defined by differences in health status. The underlying structure of the 27- and 52-item versions was examined using factor analysis and the criterion validity of KIDSCREEN-10 and KIDSCREEN-27 was analyzed by determining the magnitude of correlations with the KIDSCREEN-52. All validity testing was carried out in both the self-complete and proxy versions. Further analyses were performed to determine the cross-cultural validity of the different language versions [9 (link)]. Population norms are available at http://www.kidscreen.org.
To test responsiveness and sensitivity to change in the KIDSCREEN instruments, they have been included in longitudinal studies which provide evidence of this property. One example of such a study was a 3-year follow-up study in Spain, which investigated changes in QoL in a representative, population-based sample of children and adolescents in Spain [18 (link)] and how changes in mental health affected QoL over the same period [19 (link)]. Another example is the German longitudinal study of mental health in children and adolescents [BELLA study, 20 ].

We used the National Institute of Environmental Health Science's Environmental Genome Project SNPs database [41] , which results from direct Sanger resequencing of environmental response genes in several populations. We considered all diallelic SNPs in 5.01 Mb of sequence from noncoding regions of 219 autosomal genes (Supplementary Table 8 in Text S1). These data have been the subject of many publications, including [17] (link),[23] (link),[27] (link),[42] (link). As an assessment of quality, additional high-coverage short-read sequencing has recently been performed across 8 samples in this data set. Over 26,000 sites, the SNP concordance between this next-generation sequencing and the original Sanger sequencing averages 99.5% (D. Nickerson, personal communication). Given the high quality of this data set, we do not incorporate sequencing error into our modeling. We believe such correction will be essential in future applications to less accurate short-read sequencing data, as inference based on the frequency spectrum is sensitive to rare alleles.
To estimate the ancestral allele, we aligned to the panTro2 build of the chimp genome [43] (link). Like other methods based on the unfolded AFS, our analysis is sensitive to errors in identifying the ancestral allele. We statistically corrected the AFS for ancestral misidentification [17] (link), using a context-dependent substitution model [44] (link). This procedure has been shown to perform better than aligning to multiple species [17] (link). To account for missing data and ease qualitative comparisons between populations, we projected all spectra down to 20 samples per population [5] (link) (Text S1).
The human-chimp divergence in the data is 1.13%. We assumed a divergence time of 6 My [45] (link) and a generation time of 25 years. This yielded an estimated neutral mutation rate of per site per generation, which is comparable to direct estimates [46] . There is some controversy as to the appropriate generation time to assume in human population genetic studies [47] (link),[48] (link). In particular, the human generation time may differ between cultures and may have changed during our biological and cultural evolution. The bootstrap uncertainties reported in Table 1 and Table 2 do not include systematic uncertainties in the human-chimp divergence or generation times. The generation time, however, formally cancels when converting between genetic and chronological times.

Item generation and domain identification proceeded in three phases. First, as part of a study focused on developing an intervention to improve leadership for evidence-based practice implementation [18 ], the investigative team developed items based on review of literature relating leader behaviors to implementation and organizational climate and culture change [32 (link),33 ]. Second, items were reviewed for relevance and content by subject matter experts, including a mental health program leader, an EBP trainer and Community Development Team consultant from the California Institute for Mental Health, and four mental health program managers. Third, potential items were reviewed by the investigative team and program managers for face validity and content validity. Twenty-nine items were developed that represented five potential content domains of implementation leadership: proactive EBP leadership, leader knowledge of EBP, leader support for EBP, perseverance in the face of EBP implementation challenges, and attention and role modeling related to EBP implementation.

To observe lipid accumulation in the fibroblasts at different treatments Oil Red O stain was used. Fibroblasts were cultured in a 6-well plate for oil red staining under two different culture medium conditions. One culture medium contained DMEM with 10% FBS and 1% penicillin-streptomycin and was termed as the proliferation medium, and the other contained DMEM with 1% penicillin-streptomycin and was termed as the differentiation medium. After the recommended incubation time, the cells were treated with COPA3, gently washed with PBS, and fixed with 10% formalin for 1 h at room temperature (24°C). Formalin was removed and washed with distilled water, and the cells were then incubated with 60% isopropanol for 5 min at room temperature. Isopropanol was discarded and the cells were stained with Oil Red O solution (0.5% Oil Red O in isopropanol). The Oil Red O solution was discarded, and the samples were washed with water. Hematoxylin was added to the cells; the cells were incubated for 1 min, washed with water, and observed under a microscope (Olympus). Lipid content was measured by direct extraction of Oil Red O from the stained cells using isopropanol and absorbance at 492 nm was determined using a microplate reader (Multiskan GO Microplate Spectrophotometer, Thermo Scientific, MA).

SMH, who was not connected to the FAIMER Institute nor previously known to Fellows, conducted semi-structured interviews in the fall and winter of 2019 using Zoom or Skype to elicit Fellows’ experiences and perceptions. We pilot-tested an interview guide (Online Appendix 1), which was adjusted following the first two interviews and the investigators’ evolving conceptualizations that arose from data analysis. These changes included the order of questions asked and a shorter introduction so that we could dive into the main questions more quickly and accommodate the interview timeframe. We also modified the interview guide for FAIMER Faculty by using different wording, as needed. For example, for Faculty, we used “your teaching;” for Fellows, we used “your learning”. We also probed if FAIMER Faculty had ever changed their teaching content or process to take cultural differences into account. We deliberately avoided interview questions that could suggest discrimination, judgment, or stereotyping. Interviews lasted from 43 to 74 minutes, were audio-recorded, and were transcribed verbatim by an external professional. SMH reviewed all transcripts for accuracy.
Data analysis was iterative and started alongside data collection. We began with an inductive technique to identify preliminary patterns and common threads which were later merged to best represent the data. By using constant comparative analysis, we conducted a holistic and cross-case analysis of similar codes and themes. We also actively sought out exemplar quotations, illustrating each theme. To make sense of the findings, we gradually progressed to interpretation; this allowed us to focus on meaningful themes and elaborate on the practical application of findings. Moreover, though our data analysis was “bottom-up” as we did not generate codes using a pre-existing theoretical framework [22 (link)], we drew upon TLT as the analysis and interpretation evolved, to provide a meaningful lens on the interconnected relationship between participants and the context for teaching and learning [23 ].
Two authors (SMH and YS) independently read and coded the first two interviews, after which they discussed key observations and preliminary findings. SMH continued analyzing the remaining transcripts with input from YS, which led to consensus on patterns and themes. When all 15 interviews with Fellows had been coded, SMH started coding the remaining five interviews with FAIMER Faculty. SMH and YS met regularly to refine and agree on identified codes, thematic groupings, and interpretations of the findings. RV and WB reviewed the themes and gave feedback on themes, exemplar quotations, and applications of the findings.