To send content items to your account,
please confirm that you agree to abide by our usage policies.
If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account.
Find out more about sending content to .
To send content items to your Kindle, first ensure email@example.com
is added to your Approved Personal Document E-mail List under your Personal Document Settings
on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part
of your Kindle email address below.
Find out more about sending to your Kindle.
Note you can select to send to either the @free.kindle.com or @kindle.com variations.
‘@free.kindle.com’ emails are free but can only be sent to your device when it is connected to wi-fi.
‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.
One of the main capabilities of atom probe tomography (APT) is the ability to not only identify but also characterize early stages of precipitation at length scales that are not achievable by other techniques. One of the most popular methods to identify nanoscale clustering in APT data, based on the density-based spatial clustering of applications with noise (DBSCAN), is used extensively in many branches of research. However, it is common that not all of the steps leading to the selection of certain parameters used in the analysis are reported. Without knowing the rationale behind parameter selection, it may be difficult to compare cluster parameters obtained by different researchers. In this work, a simple open-source tool, PosgenPy, is used to justify cluster search parameter selection via providing a systematic sweep through parameter values with multiple randomizations to minimize a false-positive cluster ratio. The tool is applied to several different microstructures: a simulated material system and two experimental datasets from a low-alloy steel . The analyses show how values for the various parameters can be selected to ensure that the calculated cluster number density and cluster composition are accurate.
It has been well established that depressive disorders including perinatal depression are very heterogeneous, which partly explain the ineffectiveness of available treatments for many patients. Recent innovations in data science can help elucidate the nature of perinatal depression especially the heterogeneity in its presentation.
The present study aime to elucidate heterogeneous subtypes of PND and assess the effectiveness of a multicomponent cognitive behavioral therapy (CBT) across heterogenous subtypes of PND.
This study was conducted in 2005 in two rural areas of Rawalpindi, Pakistan. Out of a total of 3,898 women, 903 pregnant women were identifed with PND (using DSM-IV) and randomly assigned to intervention and control group. Baseline assessments included interviewer admininstered Hamilton Depression Scale (HDS) and social risk factors. Follow-up assessments were conducted at 6 months and 12 months post-intervention. Principle component analysis was run to reduce dimensionality of the HDS. Two step cluster analysis was then run to elucidate subtypes of PND using the dimensional scores. Thereafter, effectiveness of CBT was compared across these subtypes of PND using multilevel modelling.
Principle component analysis revealed a four component solution for the Hamilton depression rating scale. Using these dimensional scores, cluster analysis (average silhouette= 0.5) revealed a parsimonius four cluster soultion of participants with mild PND symptoms (n=326); predominant sleep problems (n=311) c) predominant atypical symptoms (n=80) and d) comorbid depressive and anxiety symptoms (n=186). CBT yielded moderate effect sizes across all these subtypes of PND (cohen’s d > 0.8).
Multicomponent CBT is effective across hetergeneous presentations of PND.
There is limited data on the dietary patterns of 5-year-old children in Asia. The study examined childhood dietary patterns and their maternal and child correlates in a multi-ethnic Asian cohort. Based on caregiver-reported 1-month quantitative FFQ of 777 children from the Growing Up in Singapore Towards healthy Outcomes cohort, cluster analysis identified two mutually exclusive clusters. Children in the ‘Unhealthy’ cluster (43·9 %) consumed more fries, processed meat, biscuits and ice cream, and less fish, fruits and vegetables compared with those in the ‘Healthy’ cluster (56·1 %). Children with mothers of lower educational attainment had twice the odds of being assigned to the ‘Unhealthy’ cluster (adjusted OR (95 % CI) = 2·19 (95 % CI 1·49–3·24)). Children of Malay and Indian ethnicities had higher odds of being assigned to the ‘Unhealthy’ cluster (adjusted OR = 25·46 (95 % CI 15·40, 42·10) and 4·03 (95 % CI 2·68–6·06), respectively), relative to Chinese ethnicity. In conclusion, this study identified two dietary patterns in children, labelled as the ‘Unhealthy’ and ‘Healthy’ clusters. Mothers’ educational attainment and ethnicity were two correlates that were associated with the children’s assignments to the clusters. These findings can assist in informing health promotion programmes targeted at Asian children.
Weedy rice (Oryza spp.) is one of the most troublesome weeds affecting rice (Oryza sativa L.) production in many countries. Weedy rice control is difficult in rice fields, because the weed and crop are phenotypically and morphologically similar. Weedy rice can be a source of genetic diversity for cultivated rice. Thus, this study aimed to characterize the morphological diversity of weedy rice in southern Brazil. Qualitative and quantitative traits of 249 accessions from eight rice-growing mesoregions in Rio Grande do Sul (RS) and Santa Catarina (SC) states were analyzed. For each accession, 24 morphological descriptors (14 qualitative and 10 quantitative) were evaluated. All 249 accessions from RS and SC are of indica lineage. Considering all the phenotypic traits evaluated, the accessions separated into 14 distinct groups. One of the largest groups consisted of plants that were predominantly tall with green leaves, intermediate shattering, and variable flowering time. Distinct subgroups exist within larger clusters, showing discernible phenotypic diversity within the main clusters. The variability in flowering time was high (77 to 110 d after emergence), indicating high potential for flowering synchrony with rice cultivars and, consequently, gene flow. This indicates the need to remove escapes when planting herbicide-resistant rice. Thus, weedy rice populations in southern Brazil are highly diverse, and this diversity could result in variable response to weed management.
Eating habits of lactating women can influence the nutrient composition of human milk, which in turn influences nutrient intake of breastfed infants. The aim of the present study was to identify food patterns and nutritional adequacy among lactating women in Europe. Data from a multicentre European longitudinal cohort (ATLAS study) were analysed to identify dietary patterns using cluster analysis. Dietary information from 180 lactating women was obtained using 3-d food diaries over the first 4 months of lactation. Four dietary patterns were identified: ‘vege-oils’, ‘fish-poultry’, ‘confectionery-salads’ and ‘mixed dishes’. Nutrition adequacy was not significantly different between clusters, but the ‘vege-oils’ cluster tended to yield the highest nutrition adequacy measured by Mean Adequacy Ratio. Compared with European dietary reference values (DRVs) for lactating women, women in all clusters had inadequate intakes of energy, pantothenic acid, folate, vitamin C, vitamin A, vitamin D, zinc, iodine, potassium and linoleic acid. Adequate intake for fibre and α-linolenic acid was only achieved in the ‘vege-oils’ cluster. Overall, fat intake was above DRVs. The present study showed that various dietary patterns do not adequately supply all nutrients, indicating a need to promote overall healthy dietary habits for European lactating women.
The present study aims were (1) to identify the proportion of terminally ill cancer patients with desire for hastened death (DHD) receiving specialized palliative care, (2) to identify the reasons for DHD, and (3) to classify patients with DHD into some interpretable subgroups.
Advanced cancer patients admitted to 23 inpatients hospices/palliative care units in 2017 were enrolled. Data were prospectively obtained by the primarily responsible physicians. The presence/absence of DHD and reasons for DHD were recorded. A cluster analysis was performed to identify patterns of subgroups in patients with DHD.
Data from 971 patients, whose Richmond Agitation–Sedation Scale score at admission was zero and who died in palliative care units, were analyzed. The average age was 72 years, common primary cancer sites were the gastrointestinal tract (31%) and the liver/biliary ducts/pancreas (19%). A total of 174 patients (18%: 95% confidence interval, 16–20) expressed DHD. Common reasons for DHD were dependency (45%), burden to others (28%), meaninglessness (24%), and inability to engage in pleasant activities (24%). We identified five clusters of patients with DHD: cluster 1 (35%, 61/173): “physical distress,” cluster 2 (21%, 37/173): “dependent and burdensome,” cluster 3 (19%, 33/173): “hopelessness,” cluster 4 (17%, 30/173): “profound fatigue,” and cluster 5 (7%, 12/173): “extensive existential suffering.”
A considerable number of patients expressed DHD and could be categorized into five subgroups. These findings may contribute to the development of therapeutic strategies.
Differences in psychiatric background and dose–response to asenapine in patients with schizophrenia were examined based on efficacy and safety, using data obtained in a double-blind, placebo-controlled trial.
Patients with schizophrenia were classified into three clusters by a cluster analysis based on the Positive and Negative Symptom Scale (PANSS) subscores at baseline, using the data from a 6-week, double-blind, placebo-controlled trial. PANSS Marder factor scores were calculated for each cluster. The efficacy of 10 or 20 mg/day of asenapine on PANSS score was used as the primary endpoint, with the incidence of adverse events evaluated as the secondary endpoint.
A total of 529 asenapine-treated patients were classified into 3 clusters: Cluster-P with the higher scores in positive symptoms, disorganized thoughts, and hostility/excitement, Cluster-N with higher scores in negative symptoms, and Cluster-L with overall lower scores. In Cluster-N and Cluster-L, both 10 and 20 mg/day groups showed significant improvement in PANSS scores, while only the 20 mg/day group showed a significant difference in Cluster-P. Cluster-N and Cluster-L had differences in the incidence of adverse events, but this was not seen in Cluster-P.
The efficacy and safety of asenapine 10 and 20 mg/day differed between the 3 clusters of patients. This suggests that background information regarding baseline psychiatric symptoms may affect the therapeutic response in patients with schizophrenia.
Fibromyalgia (FM) is a chronic syndrome characterized by heterogeneous clinical manifestations, and knowing this variability can help to develop tailored treatments. To understand better the heterogeneity of FM the present cross-sectional study analyzed the role of several physical symptoms (pain, fatigue and poor sleep quality) and cognitive-affective variables related to pain (pain catastrophizing, pain vigilance, self-efficacy in pain management, and pain acceptance) in the configuration of clinical profiles. A sample of 161 women with FM fulfilled an interview and several self-report measures to explore physical symptoms, cognitive-affective variables, disability and psychopathology. To establish FM groups a hierarchical cluster analysis was performed. The findings revealed three clusters that differed in the grouping variables, Wilks’ λ = .17, F(14, 304) = 31.50, p < .001, ηp2 = .59. Group 1 (n = 72) was characterized by high physical and psychological affectation, Group 2 (n = 19) by low physical affectation and high pain self-efficacy, and Group 3 (n = 70) by moderate physical affectation and low pain catastrophizing. The external validation of the clusters was confirmed, Wilks’ λ = .72, F(4, 314) = 14.09, p < .001, ηp2 = .15, showing Group 1 the highest levels of FM impact and psychopathological distress. Considering the distinctive clinical characteristics of each subgroup therapeutic strategies addressed to the specific needs of each group were suggested. Assessing FM profiles may be key for a better understanding and approach of this syndrome.
There is ongoing debate regarding the relationship between clinical symptoms and cognition in schizophrenia spectrum disorders (SSD). The present study aimed to explore the potential relationships between symptoms, with an emphasis on negative symptoms, and social and non-social cognition.
Hierarchical cluster analysis with k-means optimisation was conducted to characterise clinical subgroups using the Scale for the Assessment of Negative Symptoms and Scale for the Assessment of Positive Symptoms in n = 130 SSD participants. Emergent clusters were compared on the MATRICS Consensus Cognitive Battery, which measures non-social cognition and emotion management as well as demographic and clinical variables. Spearman’s correlations were then used to investigate potential relationships between specific negative symptoms and emotion management and non-social cognition.
Four distinct clinical subgroups were identified: 1. high hallucinations, 2. mixed symptoms, 3. high negative symptoms, and 4. relatively asymptomatic. The high negative symptom subgroup was found to have significantly poorer emotion management than the high hallucination and relatively asymptomatic subgroups. No further differences between subgroups were observed. Correlation analyses revealed avolition-apathy and anhedonia-asociality were negatively correlated with emotion management, but not non-social cognition. Affective flattening and alogia were not associated with either emotion management or non-social cognition.
The present study identified associations between negative symptoms and emotion management within social cognition, but no domains of non-social cognition. This relationship may be specific to motivation, anhedonia and apathy, but not expressive deficits. This suggests that targeted interventions for social cognition may also result in parallel improvement in some specific negative symptoms.
The dominant market where information is discovered plays the role of price leader providing substantial market information to other markets. This study investigates the dynamic relationships of 30 cattle markets across regions, cattle types, and cash/futures markets. The comparison of many markets, using an error correction model, is accomplished with the introduction of a tournament with a hierarchical cluster analysis, which allows us to conclude that the leading price for the U.S. cattle markets is discovered in the futures markets for both feeder and fed cattle.
Childhood trauma is a vulnerability factor for the development of obsessive–compulsive disorder (OCD). Empirical findings suggest that trauma-related alterations in brain networks, especially in thalamus-related regions, have been observed in OCD patients. However, the relationship between childhood trauma and thalamic connectivity in patients with OCD remains unclear. The present study aimed to examine the impact of childhood trauma on thalamic functional connectivity in OCD patients.
Magnetic resonance imaging resting-state scans were acquired in 79 patients with OCD, including 22 patients with a high level of childhood trauma (OCD_HCT), 57 patients with a low level of childhood trauma (OCD_LCT) and 47 healthy controls. Seven thalamic subdivisions were chosen as regions of interest (ROIs) to examine the group difference in thalamic ROIs and whole-brain resting-state functional connectivity (rsFC).
We found significantly decreased caudate-thalamic rsFC in OCD patients as a whole group and also in OCD_LCT patients, compared with healthy controls. However, OCD_HCT patients exhibited increased thalamic rsFC with the prefrontal cortex when compared with both OCD_LCT patients and healthy controls.
Taken together, OCD patients with high and low levels of childhood trauma exhibit different pathological alterations in thalamic rsFC, suggesting that childhood trauma may be a predisposing factor for some OCD patients.
Subjects with bipolar disorder (BD) show heterogeneous cognitive profile and that not necessarily the disease will lead to unfavorable clinical outcomes. We aimed to identify clinical markers of severity among cognitive clusters in individuals with BD through data-driven methods.
We recruited 167 outpatients with BD and 100 unaffected volunteers from Brazil and Spain that underwent a neuropsychological assessment. Cognitive functions assessed were inhibitory control, processing speed, cognitive flexibility, verbal fluency, working memory, short- and long-term verbal memory. We performed hierarchical cluster analysis and discriminant function analysis to determine and confirm cognitive clusters, respectively. Then, we used classification and regression tree (CART) algorithm to determine clinical and sociodemographic variables of the previously defined cognitive clusters.
We identified three neuropsychological subgroups in individuals with BD: intact (35.3%), selectively impaired (34.7%), and severely impaired individuals (29.9%). The most important predictors of cognitive subgroups were years of education, the number of hospitalizations, and age, respectively. The model with CART algorithm showed sensitivity 45.8%, specificity 78.4%, balanced accuracy 62.1%, and the area under the ROC curve was 0.61. Of 10 attributes included in the model, only three variables were able to separate cognitive clusters in BD individuals: years of education, number of hospitalizations, and age.
These results corroborate with recent findings of neuropsychological heterogeneity in BD, and suggest an overlapping between premorbid and morbid aspects that influence distinct cognitive courses of the disease.
Few studies have derived data-driven dietary patterns in youth in the USA. This study examined data-driven dietary patterns and their associations with BMI measures in predominantly low-income, racial/ethnic minority US youth. Data were from baseline assessments of the four Childhood Obesity Prevention and Treatment Research (COPTR) Consortium trials: NET-Works (534 2–4-year-olds), GROW (610 3–5-year-olds), GOALS (241 7–11-year-olds) and IMPACT (360 10–13-year-olds). Weight and height were measured. Children/adult proxies completed three 24-h dietary recalls. Dietary patterns were derived for each site from twenty-four food/beverage groups using k-means cluster analysis. Multivariable linear regression models examined associations of dietary patterns with BMI and percentage of the 95th BMI percentile. Healthy (produce and whole grains) and Unhealthy (fried food, savoury snacks and desserts) patterns were found in NET-Works and GROW. GROW additionally had a dairy- and sugar-sweetened beverage-based pattern. GOALS had a similar Healthy pattern and a pattern resembling a traditional Mexican diet. Associations between dietary patterns and BMI were only observed in IMPACT. In IMPACT, youth in the Sandwich (cold cuts, refined grains, cheese and miscellaneous) compared with Mixed (whole grains and desserts) cluster had significantly higher BMI (β = 0·99 (95 % CI 0·01, 1·97)) and percentage of the 95th BMI percentile (β = 4·17 (95 % CI 0·11, 8·24)). Healthy and Unhealthy patterns were the most common dietary patterns in COPTR youth, but diets may differ according to age, race/ethnicity or geographic location. Public health messages focused on healthy dietary substitutions may help youth mimic a dietary pattern associated with lower BMI.
Cluster analyses have become popular tools for data-driven classification in biological psychiatric research. However, these analyses are known to be sensitive to the chosen methods and/or modelling options, which may hamper generalizability and replicability of findings. To gain more insight into this problem, we used Specification-Curve Analysis (SCA) to investigate the influence of methodological variation on biomarker-based cluster-analysis results.
Proteomics data (31 biomarkers) were used from patients (n = 688) and healthy controls (n = 426) in the Netherlands Study of Depression and Anxiety. In SCAs, consistency of results was evaluated across 1200 k-means and hierarchical clustering analyses, each with a unique combination of the clustering algorithm, fit-index, and distance metric. Next, SCAs were run in simulated datasets with varying cluster numbers and noise/outlier levels to evaluate the effect of data properties on SCA outcomes.
The real data SCA showed no robust patterns of biological clustering in either the MDD or a combined MDD/healthy dataset. The simulation results showed that the correct number of clusters could be identified quite consistently across the 1200 model specifications, but that correct cluster identification became harder when the number of clusters and noise levels increased.
SCA can provide useful insights into the presence of clusters in biomarker data. However, SCA is likely to show inconsistent results in real-world biomarker datasets that are complex and contain considerable levels of noise. Here, the number and nature of the observed clusters may depend strongly on the chosen model-specification, precluding conclusions about the existence of biological clusters among psychiatric patients.
After a general introduction to multivariate statistical analyses, we focus on describing the task of multivariate classification, distinguishing its non-hierarchical and hierarchical forms. Focusing on hierarchical agglomerative classification methods (cluster analysis), we highlight the important decisions that must be made regarding the measurement of dissimilarity (distance) among objects. Following this, we explain the construction of dendrograms representing this hierarchical classification. We also briefly mention divisive classification methods, focusing on the TWINSPAN method. The methods described in this chapter are accompanied by a carefully-explained guide to the R code needed for their use, in this case employing the cluster package.
In 2015, a Chaplaincy Research Consortium generated a model of human spirituality in the palliative care context to further chaplaincy research. This article investigates the clinical fit of (a) the model's fundamental premise of universal human spirituality and (b) its 4 proposed stage descriptors (Discovery, Dialogue, Struggle, and Arrival).
First, we collected qualitative data from an interdisciplinary palliative care focus group. Participants (n = 5) shared responses to the statement “the human spirit has essential commonalities across [ … ] groups and [ … ] attributes.” Participants also shared vignettes of spiritual care, and 48 vignettes illustrating patients’ spiritual journeys were subsequently taken from the transcript of that group. Second, we invited different mixed discipline palliative care professionals (n = 9) to individually card sort these vignettes to the model's 4 stage descriptors; we conducted pattern analysis on the results. We then administered a third step, convening six physicians to complete the card sort again, this time allowing designation of cards to one or two of the 4 stage descriptors.
Focus group participants were supportive of the model's all-encompassing definition of spirituality. The concept of “connectedness” was a shared focus for all participants, connectedness and spirituality appearing almost synonymous. Pattern analysis of assigned 48 vignettes to the 4 stages showed stronger consensus around Discovery and Arrival than Struggle and Dialogue. Results of the additional card sort suggested Struggle and Dialogue involve oscillation and are harder to think of as a steady state as distinct from processes associated with Discovery or Arrival.
Significance of results
“Connectedness” is a productive concept for modeling human spiritual experience near the end of life. As one healthcare professional said: “this connectedness piece is [ … ] what I always look for … ” Although further work is needed to understand struggle and dialogue elements in peoples’ spiritual journeys, discovery and arrival shared consensus among participants.
Migration is often perceived as a challenge to the welfare state. To manage this challenge, advanced welfare states have established transgovernmental networks. This article examines how domestic factors condition the interaction of representatives of advanced welfare states when they cooperate on transnational welfare governance. Based on new survey data, it compares who interacts with whom in one of the oldest transgovernmental networks of the European Union (EU) – the network that deals with EU citizens' rights to cross-border welfare. First, the authors perform a welfare cluster analysis of EU-28 and test whether institutional similarity explains these interactions. Furthermore, they test whether the level and kind of migration explains interaction and examine the explanatory value of administrative capacity. To test what drives interactions, the study employs social network analysis and exponential random graph models. It finds that cooperation in networked welfare governance tends to be homophilous, and that political cleavages between sending and receiving member states are mirrored in network interactions. Domestic factors are key drivers when advanced welfare states interact.
Bipolar disorder (BD) is associated with social cognition (SC) impairments even during remission periods although a large heterogeneity has been described. Our aim was to explore the existence of different profiles on SC in euthymic patients with BD, and further explore the potential impact of distinct variables on SC.
Hierarchical cluster analysis was conducted using three SC domains [Theory of Mind (ToM), Emotional Intelligence (EI) and Attributional Bias (AB)]. The sample comprised of 131 individuals, 71 patients with BD and 60 healthy control subjects who were compared in terms of SC performance, demographic, clinical, and neurocognitive variables. A logistic regression model was used to estimate the effect of SC-associated risk factors.
A two-cluster solution was identified with an adjusted-performance group (N = 48, 67.6%) and a low-performance group (N = 23, 32.4%) with mild deficits in ToM and AB domains and with moderate difficulties in EI. Patients with low SC performance were mostly males, showed lower estimated IQ, higher subthreshold depressive symptoms, longer illness duration, and poorer visual memory and attention. Low estimated IQ (OR 0.920, 95% CI 0.863–0.981), male gender (OR 5.661, 95% CI 1.473–21.762), and longer illness duration (OR 1.085, 95% CI 1.006–1.171) contributed the most to the patients clustering. The model explained up to 35% of the variance in SC performance.
Our results confirmed the existence of two discrete profiles of SC among BD. Nearly two-thirds of patients exhibited adjusted social cognitive abilities. Longer illness duration, male gender, and lower estimated IQ were associated with low SC performance.
Identifying profiles of people with mental and substance use disorders who use emergency departments may help guide the development of interventions more appropriate to their particular characteristics and needs.
To develop a typology for the frequency of visits to the emergency department for mental health reasons based on the Andersen model.
Questionnaires were completed by patients who attended an emergency department (n = 320), recruited in Quebec (Canada), and administrative data were obtained related to sociodemographic/socioeconomic characteristics, mental health diagnoses including alcohol and drug use, and emergency department and mental health service utilization. A cluster analysis was performed, identifying needs, predisposing and enabling factors that differentiated subclasses of participants according to frequency of emergency department visits for mental health reasons.
Four classes were identified. Class 1 comprised individuals with moderate emergency department use and low use of other health services; mostly young, economically disadvantaged males with substance use disorders. Class 2 comprised individuals with high emergency department and specialized health service use, with multiple mental and substance use disorders. Class 3 comprised middle-aged, economically advantaged females with common mental disorders, who made moderate use of emergency departments but consulted general practitioners. Class 4 comprised older individuals with multiple chronic physical illnesses co-occurring with mental disorders, who made moderate use of the emergency department, but mainly consulted general practitioners.
The study found heterogeneity in emergency department use for mental health reasons, as each of the four classes represented distinct needs, predisposing and enabling factors. As such, interventions should be tailored to different classes of patients who use emergency departments, based on their characteristics.
Rosa x odorata (sect. Chinenses, Rosaceae) is an important species distributed only in Yunnan Province, China. There is an abundance of wild variation within the species. Using 22 germplasm resources collected from the wild, as well as R. chinensis var. spontanea, R. chinensis ‘Old Blush’ and R. lucidissima, this study involved morphological variation analysis, inter-trait correlation analysis, principal component analysis and clustering analysis based on 16 morphological traits. This study identified a high degree of morphological diversity in R. x odorata germplasm resources and the variation coefficients had a distribution range from 18.00 to 184.04%. The flower colour had the highest degree of variation, while leaflet length/width had the lowest degree of variation. Inter-trait correlation analysis revealed that there was an extremely significant positive correlation between leaflet length and leaflet width. There was also a significant positive correlation between the number of petals and duration of blooming, and the L* and a* values of flower colour were significantly negatively correlated. Principal component analysis screened five principal components with the highest cumulative contribution rate (81.679%) to population variance. Among the 16 morphological traits, style length, sepal width, flower diameter, flower colour, leaflet length and leaflet width were important indices that influenced the morphology of R. x odorata. This study offers guidance for the further development and utilization of R. x odorata germplasm resources.