Selection procedure
We designed a cross-sectional study. Twenty-six herds registered in the NBCHRS and with ≥15 cow-years from the three most beef cattle-dense regions of Norway were sampled by computerized systematic assignment. Fifteen herdsmen accepted to take part in the study. Three herds were excluded because of non-compliance in connection with the trimming and recording of claw lesions. Claw and limb disorders and claw length and shape in cows and heifers ≥ 18 months of age were recorded.
Study population
The final study population consisted of 12 herds: 6 in region I (Hedmark/Oppland), 3 in region II (Rogaland) and 3 in region III (Trøndelag). The total number of beef-cow herds with ≥ 15 cow-years within the three regions was 213, 71 and 169. Respectively 66, 14 and 45 of these herds were members of the NBCHRS. The total number of beef-cow herds with ≥ 15 cow-years in Norway was 1782 and 232 were members of the NBCHRS.
The mean number of cows and heifers in the study herds was 30 and the total number was 362; 28 heifers (more than 30 days from first calving) and 334 cows. Data from 337 of these animals were available from the NBCHRS. Housing, management, feeding and cow variables in each herd are in Table 1, 2 and 3.
The Hereford breed was present in 5 of the herds and Aberdeen Angus, Charolais, Limousine and Simmental in respectively 3, 6, 3 and 1 of the herds. In total 72 Herefords, 97 Charolais, 14 Aberdeen Angus, 20 Limousines, 15 Simmentals, 93 cross-breds and 26 animals of unknown breed were included in the study. Animals with more than 75% of one breed were counted as pure-bred.
Recording of data
All seven professional claw trimmers attended two courses covering observation and recording of lameness, claw trimming procedures, diagnosis, recording and treatment of claw lesions. Individual training was given to each claw trimmer at the initiation of the practical work. For practical reasons two claw trimmers cooperated in some of the herds. All trimmers had previously participated in a study of claw health in dairy cattle [15].
The cows were trimmed and examined once during the period from the 15th of January 2003 to pasture season started. The last herd was visited on the 12th of June. Lameness was assessed when the animal was moved to the trimming chute as absent (no notation), moderate (1) or severe (2) (Table 3). Pericarpal and peritarsal swellings and wounds were recorded as absent (no notation), swelling (1), wound (2) or both swelling and wound (3). Claw shapes were recorded as normal (no notation), asymmetric (1) or corkscrewed (2). Claw lesions were diagnosed on the basis of macroscopic examination before and after trimming to the correct claw shape. The trimming technique included levelling the two claws, aiming for symmetric bulbs. The axial and abaxial walls were both intended to be parts of the bearing surface and the two claws were trimmed flat and balanced with each other. The caudal 2/3 of the axial sole of both claws was dished out. Dermatitis, heel-horn erosions, haemorrhages of the white line and the sole, sole ulcers and white-line fissures were scored as absent (no notation), mild (1), moderate (2) or severe (3). Definitions in Table 4 are adapted from Bergsten [18]. We used only one variable for "dermatitis" because the occurrence of digital dermatitis is close to zero in Norway and we assumed that the recorded cases would be interdigital dermatitis. The presence of double sole, interdigital hyperplasia, horizontal and vertical fissures, interdigital phlegmon and papillomatous dermatitis was also recorded. The recording protocol did not differentiate between the inner and the outer claw because most lesions occur in the outer hind claw and the inner front claw [19,4]. The term "claw and limb disorders" includes lameness, pericarpal and peritarsal swellings or wounds and all claw lesions in this study, but not asymmetric or corkscrewed claws.
Identity, age, date for calving, breed and events as disease and carcass characteristics were extracted from the NBCHRS. When associations between age and claw lesions were assessed, the animals were separated in young (2–4 years), medium-aged (5–7 years) and old animals (8–10 years). Conformation class and fat cover class were defined according to the EUROP grading system as defined by the EU [20,21]. Data on housing system, environment, feeding and management were collected by visits and questionnaire survey by one animal husbandry adviser.
Statistical analyses
The statistical analyses were performed in the SAS-PC System® Version 9.1 for Windows at cow and herd level. PROC UNIVARIATE, PROC MEANS and PROC FREQ were used for the descriptive analyses.
Two different unconditional logistic regression models were performed using the presence or absence of infectious and laminitis-related claw lesions as dependent variable, and age group (young = 2–4 years, medium = 5–7 years, old = ≥8 years) as independent variable (PROC LOGISTIC).
For animals that had been slaughtered within two years after claw inspection, general linear models (GLM) with carcass weight, conformation class and fat cover class as dependent variables were performed with "disorder" (presence of any claw or limb disorder, yes/no) as independent variable and breed as adjusting variable. The fit of the models was assessed by the R2-values.
At herd level, univariate statistics were generated for all environmental factors in relation to the occurrence of 1) claw and limb disorders, 2) laminitis-related claw lesions and 3) infectious claw lesions. For laminitis-related claw lesions, several of the environmental factors were statistical significant. All these factors were included as independent variables (fixed effects) in a preliminary multivariate (GLM) model with the herd prevalence of laminitis-related claw lesions as dependent variables. Two-tailed tests were applied. The type III F-test was used as elimination criterion. The modelling was manually conducted by stepwise backward elimination of variables one by one, using a p-value of 0.05 as the level for exclusion from the model. The least square means were estimated for all levels of significant independent variables in the final model.
General linear model (GLM) with calving interval as dependent variable was performed using PROC GLM, with claw and limb disorders (yes/no) as independent variables and breed as adjusting variable.
Answers to all your Biology Questions
