Associations of automated body condition scores at dry-off and through early lactation with milk yield of Holstein cows

Abstract

The objective of this study was to analyze the associations of body condition score (BCS) and BCS change (∆BCS) during the dry period and the first 100 d of lactation with daily milk yield. Examining the involvement of health status in the associations between BCS and milk yield was a secondary objective of this research. Data included 12,042 lactations in 7,626 Holstein cows calving between April 2019 and January 2022 in a commercial dairy operation located in Colorado, USA. BCSs were generated daily by an automated BCS camera system located at the exit of the milking parlor. The assessment points selected for this study were dry-off (BCSdry), calving (BCS1), 7 DIM (BCS7), 14 DIM (BCS14), 21 DIM (BCS21), and nadir (nBCS; defined as the lowest daily BCS from calving to 100 DIM). Subsequently, these BCS were categorized considering quartiles (Q1 = 25% lowest BCS; Q4 = 25% greatest BCS), separately for primiparous and multiparous cows. Changes in BCS were calculated from dry-off to calving (multiparous); and from calving to 7 DIM, 14 DIM, 21 DIM, and nadir and assigned into quartile categories considering Q1 as the 25% of cows with the greatest decrease of BCS. Lactations were classified based on the number of health events before nadir as healthy, affected by one event, or having multiple events. Data were examined in primiparous and multiparous cows separately using ANOVA. The least square means for daily milk at 60 DIM and 305 DIM were calculated by category of BCS and ∆BCS at multiple time points and time periods. Subsequently, lactation curves were created by BCS and ∆BCS categories and by health status. Multivariable models included calving season and BCS1 as covariables. The largest differences in milk yield among categories of BCS and ∆BCS were identified for BCS originated at nadir and for the ∆BCS between calving and nadir. The differences in average daily milk yield between cows in the lowest and the greatest nBCS category (Q1 vs. Q4) were 3.3 kg/d (60 DIM) and 3.4 kg/d (305 DIM) for primiparous cows and 2.4 kg/d (60 DIM) and 2.1 kg/d (305 DIM) for multiparous cows. During the period from calving to nadir, primiparous cows in Q1 (greatest decrease of BCS) produced 4.3 kg/d (60 DIM) and 3.8 kg/d (305 DIM) more than cows in Q4. For multiparous cows, the differences were 3.0 kg/d (60 DIM) and 1.9 kg/d (305 DIM) in favor of Q1 cows. Overall, the associations between BCS and ∆BCS categories and milk yield were not consistent across time and they depended on the parity category. Nonetheless, as the assessment of BCS and ∆BCS approached the nadir, the association between greater milk yield and lower BCS or greater reduction in BCS became more evident.

•The association between greater milk yield and use of energy reserves represented by loss in body condition score (BCS) became more evident at the time of the nadir BCS.

•The magnitude of the nadir BCS reflects the interrelations among energy balance, use of energy reserves, and potential for milk production.

Introduction

Cows transitioning into lactation face the challenge of adjusting nutrient requirements from maintenance and gestation to those of high milk production (Horst et al., 2021). As the demand for energy outpaces the increments in energy intake following parturition, a 30- to 100-d period of negative balance is common to most high-producing cows (Coffey et al., 2002; Drackley, 1999). Adding complexity to this transition, the early postpartum period is accompanied by inflammation and dysregulation of the immune system, metabolic imbalances, and environmental stressors (Horst et al., 2021; Trevisi and Minuti, 2018).

To approach a nutritional equilibrium, cows mobilize fat and labile protein from body energy reserves, which is evidenced by reductions in body condition scores (BCSs) (Lean et al., 2013). Nonetheless, BCS changes during the postpartum and early lactation are variable, with a proportion of cows that are able to maintain or gain body condition through this period (Britt, 1992; Middleton et al., 2019; Pinedo et al., 2022).

Multiple variables have been identified in association with the magnitude of the postpartum BCS changes. Transition cow disorders (Bedere et al., 2018; Pinedo et al., 2022; Roche, 2006) and genetic merit for milk production (Lucy et al., 2009; Mao et al., 2004) are two cow-related factors that have been identified to affect the dynamics of postpartum BCS. However, as reviewed by Broster and Broster (1998) and Roche et al. (2009), studies report variable associations between BCS and milk yield. Some authors have identified positive associations between milk production and intermediate BCS of 3.0 to 3.3 at calving (Grainger et al., 1982; Markusfeld et al., 1997; Roche et al., 2009; Waltner et al., 1993) or greater reductions in BCS postpartum (Bedere et al., 2018; Ruegg and Milton, 1995; Waltner et al., 1993). Nonetheless, other research has indicated no associations of BCS at calving with subsequent milk yield (Garnsworthy and Jones, 1987, 1993).

Genetic selection for increased milk production has resulted in changes in the underlying metabolic processes in early lactation cows (Chagas et al., 2009). Extensive research has linked increased milk production with elevated activity of lipolytic enzymes, accompanied by greater gene expression in pathways associated with body fat mobilization (McNamara, 1991; McNamara and Hillers, 1986a,1986b; Smith and McNamara, 1990; Sumner and McNamara, 2007). In consequence, this selection for high milk yield has derived in greater and extended loss of BCS postpartum, with a limited ability for redirecting energy toward body reserves until later in lactation (Buckley et al., 2000a, 2000b; Roche, 2006; Veerkamp and Brotherstone, 1997).

Changes in BCS during the dry period have also been associated with milk yield in the subsequent lactation. According to Chebel et al. (2018) and Domecq et al. (1997), cows that increased their BCS during the dry period had a greater milk yield in the subsequent lactation. Nonetheless, the mechanisms explaining this association are likely different from those in lactating cows.

As BCS and milk yield depend on multiple interrelated factors, it is difficult to isolate the associations between fat mobilization and the production of milk. Moreover, most previous studies on BCS and milk yield include body condition scoring at few time points and cumulative values for milk yield, which may hide changes in lactation curves.

In recent years, a variety of solutions for automated body condition scoring have been developed and validated in dairy farms (Mullins et al., 2019; Zin et al., 2020). The availability of daily individual milk yield coupled with daily BCS, originating from automated camera devices within commercial dairies, makes it possible to perform more detailed analyses on the associations between these two variables. Moreover, daily data provides opportunities to identify time points and time periods where these associations may be more evident, as well as potential confounding variables, such as the occurrence of disease, that could affect these associations (Pinedo et al., 2022a, 2022b; Truman et al., 2022). Monitoring the dynamics of BCS from dry-off to calving and into early lactation and awareness of the associations between BCS and fertility (Carvalho et al., 2014; Pinedo et al., 2022a, 2022b; Roche et al., 2009) and milk yield may improve management decisions related to culling, insemination, grouping, and nutrition. Moreover, as proposed in a review by Giordano et al. (2022), high-frequency individual data would allow for the identification of cows with different reproductive and performance potentials. Subgroups of cows sharing biological and performance features could be allocated to specific nutritional and reproductive management strategies optimizing cow performance and herd profitability.

We hypothesized that the analysis of high-sfrequency data will support the previous research indicating that the greatest reductions in BCS through early lactation will be identified in high-producing cows (Buckley et al., 2000a, 2000b; McCarthy et al., 2007; Roche, 2006), who are devoting more energy resources to milk production. Moreover, it was envisioned that the differences in BCS in high and low producing cows would be variable over time. Furthermore, it was anticipated that these associations will be disrupted in cows affected by early lactation health disorders, as greater BCS loss would be driven by suboptimal health and not by superiority in milk production. In consequence, the objective of this study was to analyze the associations of BCS and BCS change (∆BCS) during the dry period and the first 100 d of lactation with daily milk yield. Examining the involvement of health status in the associations between BCS and milk yield was a secondary objective of this research.

Materials and Methods

Study population

This retrospective observational study analyzed data extracted from on-farm software provided by a commercial dairy. As the research did not include any intervention and the authors did not have any interaction with the animals, IACUC approval was not required.

This research follows a series of studies exploring potential associations of automated BCS during the dry period and the first 100 d of lactation and cow fertility and health (Hernandez-Gotelli et al., 2023; Pinedo et al., 2022a, 2022b). The current analyses were focused on milk yield as the main outcome of interest and included 12,042 lactations (primiparous = 4,361; multiparous = 7,681) in 7,626 Holstein cows calving between April 2019 and January 2022 in a commercial dairy farm located in Windsor, Colorado, USA. To have sufficient individual information on both milk yield and BCS, cows were required to have at least one artificial insemination (AI) postpartum to be considered in the analyses. Combined death and live culling before 80 DIM (primiparous) and 60 DIM (multiparous) were 13.7% and 12.6%, respectively, and these cows were not considered in the analysis.

Housing consisted of a cross-ventilated barn, and cows were subjected to 3 milkings per day (6:00 a.m.; 2:00 p.m.; 10:00 p.m.) in a 90-units rotary parlor. Close-up cows were fed a calcium and phosphorus binder (Xzelit, Vilofoss, Fredericia, Denmark) to reduce the absorption of calcium. Feed was delivered once per day and the cows were maintained at a stocking density of <80%. Subsequently, cows were transitioned to the fresh pen and fed three times per day with a TMR residual of 5%. Primiparous and multiparous cows were housed in two different groups with < 80% stocking density. Cows were moved from the fresh group within 21 DIM. The nutrient composition of the pre- and postpartum diets used during this period at the study farm is presented in Table 1.

Table 1.

Nutrient content of the prepartum and postpartum diets during the study

Nutrient content	Prepartum	Postpartum
Dry matter intake (kg/d)	14.68	15.88
Dry matter (%)	48.35	41.56
Crude protein (%DM)	15.83	16.48
Metabolizable protein (MP, g/d)	1,485.47	1,844.60
Neutral detergent fiber (%DM)	36.69	30.02
Forage NDF (%DM)	33.2	24.1
Starch (%DM)	19.85	25.1
Sugars (%DM)	4.18	4.26
Soluble fiber (%DM)	6.65	6.24
Ash (%DM)	10.15	9.19
Ca (%DM)	0.57	0.8
P (%DM)	0.29	0.33
Mg (%DM)	0.22	0.33
K (%DM)	1.3	1.12
S (%DM)	0.21	0.2
Na (%DM)	0.45	0.47
Cl (%DM)	0.4	0.44
Zn (ppm)	30.42	87.29
Mn (ppm)	40.95	72.1
Cu (ppm)	8.48	15.14
Co (ppm)	0.05	0.96
I (ppm)	0	0.88
Se (ppm)	0.19	0.37
NEl (Mcal/kg)	1.496	1.716
Met (%MP)	2.3	2.2
Lys (%MP)	7.2	7.27
DCAD (mEq/kg)1	288.23	241.9

¹Close-up cows were fed a calcium and phosphorus binder (Xzelit, Vilofoss, Fredericia, Denmark).

Reproductive management was based on AI preceded by a double OvSynch protocol. First AI occurred at about 80 DIM and 60 DIM in primiparous and multiparous cows, respectively. Cows determined nonpregnant at 32 ± 3 after the first AI, were resubmitted to AI based on estrus detection using the DeLaval activity meter system (DelPro Farm Manager software).

Individual cow information was collected from dry-off (multiparous) or calving (primiparous) and through the lactation until subsequent dry-off, live culling, or death. Demographic, calving-related events and health data were extracted from on-farm software (Dairy Comp 305; Valley Ag Software, Tulare, CA), while daily milk yield and BCS were extracted from DelPro Farm Manager software (DeLaval International AB, Tumba, Sweden). All the collected information was subsequently combined in a single file with records organized by cow ID and lactation number. Records started with the cow’s unique identification number, followed by the calving date, and ended with the date of dry-off, live culling, or death. The date of and BCS at the previous dry-off were also included in multiparous cows. Lactation number, calving-related and disease events, daily milk yield up to 305 DIM, and daily BCS were also incorporated in each record.

Body condition scoring and BCS categorization

Scores were generated by BCS video cameras (DeLaval International AB) previously validated by Mullins et al. (2019) that were mounted on the sorting gate at each exit (n = 2) of the milking parlor (Pinedo et al., 2022a, 2022b). As the cow passed under the mounted camera, a continuous video (30 FPS, 32,000 captured reference points) was taken and a 3D image from the video was automatically created and saved by the BCS camera software (Mullins et al., 2019). Subsequently, the 3D images were processed through an algorithm and analyzed to locate the physical features (pelvic bones, tail head ligaments, sacral ligaments, and vertebrae) of the cow to calculate the BCS that were stored in DelPro Farm Manager Software. The three daily BCS were averaged into one daily score for the analysis. The scoring of body condition was based on the methodology proposed by Ferguson et al. (1994) but modified to report 0.1 increments.

All automated BCS data were recorded in and downloaded from DelPro Farm Manager and BCS at dry-off (BCSdry), calving (BCS1), 7 DIM (BCS7), 14 DIM (BCS14), 21 DIM (BCS21), and nadir (nBCS; defined as the lowest daily BCS from calving to 100 DIM) were selected for the analyses and subsequently categorized considering quartiles (Q1 = 25% lowest BCS; Q4 = 25% greatest BCS), separately for primiparous and multiparous cows. Changes in BCS were calculated from dry-off to calving (multiparous cows); and from calving to 7 DIM, 14 DIM, 21 DIM, and nadir and assigned into quartile categories considering Q1 as the 25% of cows with the greatest decrease of BCS. Multiparous cows were assigned into categories of dry period length considering quartiles (short < 56 d; Q2 = 56 to 59 d; Q3 = 60 to 63 d; long > 63 d) to test the associations among duration of the dry period and BCS at calving and change in BCS from dry-off to calving. Finally, time to nadir was defined as the DIM when the cow reached nBCS. Values were categorized using the quartile distribution (Q1 = lower DIM), separately for primiparous and multiparous cows.

Other explanatory variables and study outcomes

Calving-related events and diseases were obtained from records stored in on-farm software. Diagnoses were completed on a daily basis by farm personnel trained by the attending veterinarian. Only health events diagnosed before the day of nadir BCS were considered in the analyses. When a cow had repeated events for the same disease, only the first event was considered.

Reproductive health events of interest included dystocia, retained fetal membranes (membranes not expelled after 24 h postcalving; Kelton et al., 1998), metritis (7 ± 4 DIM; watery, reddish/brownish fetid vaginal discharge, independent of fever; McDougall et al., 2007), and pyometra (25 ± 3 DIM; ultrasound examination evidencing accumulation of purulent material within the uterine lumen in the presence of a persistent corpus luteum and a closed cervix; Sheldon et al., 2006). Metabolic disorders included clinical hypocalcemia (down cow or cow unsteady prior to calving to 1 or 2 d after calving with no other abnormal physical exam findings and responsive to calcium administration), subclinical ketosis (6 ± 5 DIM; blood BHB > 1.3 mmol/L), and left displaced abomasum (off feed, scant pasty manure, ping in left flank, usually within 30 DIM). Other health disorders considered were lameness (assessed weekly; score > 2; Bicalho et al., 2007), clinical mastitis (abnormal milk or udder inflammation); digestive problems (off feed, altered feces), injury (visible body trauma including wounds, ulcerations, and swelling), and pneumonia (nasal discharge, respiratory distress, altered lung sounds; Pinedo et al., 2022a).

Health status within each lactation was classified based on the number of health events before nadir BCS as healthy, affected by one event, or having multiple events. Parity was created as a binary variable including primiparous (lactation number = 1) and multiparous (lactation number ≥ 2) cows. Calvings were grouped by season (spring, summer, fall, or winter).

Daily milk yields up to 305 DIM were obtained from DelPro Farm Manager. Least square means (SE) for average milk yield up to 60 DIM and 305 DIM were calculated and considered as the main outcomes to test the association between milk yield and BCS dynamics. Milk yield up to 60 DIM was selected to represent the cow’s performance during the early lactation, while milk at 305 DIM was a representation of the entire lactation.

Statistical analyses

All the analyses were conducted separately for primiparous and multiparous cows using SAS 9.4 (SAS Institute Inc., Cary, NC). Initial data evaluation was completed by building frequency tables and through visual assessment of BCS and milk yield. Descriptive statistics for daily BCS at specific time points and time periods, separated by parity category were calculated using the PROC UNIVARIATE.

Least square means (SE) for daily average milk yield up to 60 DIM and 305 DIM by category of BCS and ∆BCS at multiple time points and time periods were calculated and compared using ANOVA for repeated measures analysis, considering ID as the REPEATED statement (PROC MIXED). Initial univariable models using only BCS or ∆BCS as explanatory variables were followed by multivariable models that considered calving season, BCS1, and occurrence of disease before nBCS as potential covariables.

Lactation curves from calving to 305 DIM were built from the resulting daily milk yield LSM by category of BCS and ∆BCS and by health status and compared through repeated measures analysis. For all outcome variables, significant predictors were determined at P<0.05; interaction terms and controlling variables remained in the models at P ≤ 0.10.

Results

After edits, data from 12,042 lactations (primiparous = 4,361; multiparous = 7,681) were used in this study. A total of 4,661, 12,042, 11,880, 11,390, 11,146, and 12,042, lactations had BCS at dry-off, calving, 7 DIM, 14 DIM, 21 DIM, and nadir, respectively. Overall, the distribution of calvings across seasons was spring 23.2%, summer 30.9%, fall 29.2%, and winter 16.7%.

The mean (SD) duration of the dry period was 60.1 (6.9) d and the category of dry period length was not associated with BCS change from dry-off to calving. However, BCS at calving was greater in cows in the extended dry period length quartile (BCS = 3.33) compared with cows in Q3 (3.31), Q2 (3.30), and in the short dry period length category (3.30).

Descriptive statistics for BCS at specific time points and time periods are presented in Table 2. These data include the cutoff values used to allocate BCS and ∆BCS into the Q1 to Q4 categories.

Table 2.

Descriptive statistics for BCS at the time points and time periods in analysis, presented by parity category. A total of 4,661, 12,042, 11,880, 11,390, 11,146, and 12,042, lactations had BCS at dry-off, calving, 7 DIM, 14 DIM, 21 DIM, and nadir, respectively

	Lower quartile	Median	Upper quartile	Mean	SD
Time point (BCS)
Primiparous
Calving	3.30	3.50	3.60	3.40	0.15
7 DIM	3.30	3.40	3.50	3.37	0.14
14 DIM	3.20	3.30	3.40	3.29	0.16
21 DIM	3.10	3.20	3.30	3.22	0.17
Nadir	2.90	3.07	3.20	3.03	0.20
Multiparous
Dry-off	3.30	3.40	3.50	3.39	0.49
Calving	3.28	3.40	3.50	3.37	0.15
7 DIM	3.30	3.40	3.50	3.30	0.21
14 DIM	3.20	3.30	3.40	3.22	0.22
21 DIM	3.10	3.20	3.40	3.13	0.22
Nadir	2.70	2.90	3.00	2.86	0.25
Time period (ΔBCS)
Primiparous
Calving to 7 DIM	−0.070	0.000	0.000	−0.007	0.107
Calving to 14 DIM	−0.150	−0.100	0.000	−0.077	0.134
Calving to 21 DIM	−0.250	−0.170	−0.080	−0.157	0.147
Calving to nadir	−0.444	−0.303	−0.200	−0.340	0.180
Multiparous
Dry-off to calving	−0.20	−0.10	0.00	−0.10	0.49
Calving to 7 DIM	−0.10	0.00	0.03	−0.01	0.13
Calving to 14 DIM	−0.20	−0.10	0.00	−0.10	0.17
Calving to 21 DIM	−0.30	−0.20	−0.10	−0.19	0.18
Calving to nadir	−0.60	−0.43	−0.30	−0.46	0.22

In primiparous cows, the differences in average milk yield (kg/d) at 60 DIM by category of BCS were significant for scores determined at 14 DIM, 21 DIM, and at nadir BCS (Figure 1). The most extreme differences were calculated at nadir, where cows in Q1 (lowest BCS category) produced 3.3 kg/d (P < 0.001) more than cows in Q4 (greatest BCS). This trend was similar for ∆ BCS categories where cows in Q1 for the period calving to nadir (greatest loss of BCS) produced 4.3 kg/d (P < 0.001) more than cows in Q4 (smallest loss of BCS).

Figure 1.

Average daily milk yield up to 60 DIM in primiparous cows by category of BCS at calving, 7 DIM, 14 DIM, 21 DIM, and at nadir (top panel) and by category of ∆BCS from calving to 7 DIM, calving to 14 DIM, calving to 21 DIM, and calving to nadir (bottom panel). Values for BCS were categorized using the quartile distribution [Q1 = lowest BCS (blue), Q2 = orange; Q3 = gray; Q4 = greatest BCS (yellow)].Values for ∆BCS were categorized using the quartile distribution [Q1 = greatest loss of BCS (blue), Q2 = orange; Q3 = gray; Q4 = smallest loss of BCS (yellow)].The full model included calving season and BCS1 as covariables.

For multiparous cows, the differences in average milk yield (kg/d) at 60 DIM were most significant for categories of BCS at the previous dry-off and the nadir (Figure 2). Cows in Q1 at dry-off subsequently produced 1.3 kg/d (P < 0.001) more than cows in Q4, while cows in Q1 at nadir produced 2.4 kg/d (P < 0.001) more than cows in Q4. For ∆BCS from calving to nadir, cows in Q1 produced 3.0 kg/d (P < 0.001) more than cows in Q4. Interestingly, the order in average milk yield was reversed for the ∆BCS from dry-off to calving, with cows in BCS Q4 producing 2.1 kg/d more than cows in Q1 (P < 0.001).

Figure 2.

Average daily milk yield up to 60 DIM in multiparous cows by category of BCS at dry-off, calving, 7 DIM, 14 DIM, 21 DIM, and at nadir (top panel) and by category of ∆BCS from dry-off to calving, calving to 7 DIM, calving to 14 DIM, calving to 21 DIM, and calving to nadir (bottom panel). Values for BCS were categorized using the quartile distribution [Q1 = lowest BCS (blue), Q2 = orange; Q3 = gray; Q4 = greatest BCS (yellow)]. Values for ∆BCS were categorized using the quartile distribution [Q1 = greatest loss of BCS (blue), Q2 = orange; Q3 = gray; Q4 = smallest loss of BCS (yellow)].The full model included calving season and BCS1 as covariables.

Lactation curves calculated by BCS and ∆BCS categories are presented by parity category in Figures 3 to 6, while LSM for average daily milk yield up to 305 DIM by BCS category are reported in Tables 3 and 4. Consistently, BCS and ∆BCS categories from scores obtained as the lactation progresses from calving to nadir (and at dry-off) resulted in more distinct lactation curves among groups. As expected, cows in lower BCS categories and cows with more pronounced BCS loss after calving had milk yield above cows that evidenced greater BCS or had smaller reductions in BCS.

Figure 3.

Lactation curves for daily milk yield (kg) up to 305 DIM in primiparous cows by category of BCS at calving (top left panel), 14 DIM (top right), 21 DIM (bottom left), and at nadir (bottom right). Values for BCS were categorized using the quartile distribution [Q1 = lowest BCS (blue), Q2 = orange; Q3 = gray; Q4 = greatest BCS (yellow)]. The full model included calving season and BCS1 as covariables.

Figure 6.

Lactation curves for daily milk yield (kg) up to 305 DIM in multiparous cows by category of ∆BCS from dry-off to calving (top left panel), calving to 14 DIM (top right), calving to 21 DIM (bottom left), and calving to nadir (bottom right). Values for ∆BCS were categorized using the quartile distribution [Q1 = greatest loss of BCS (blue), Q2 = orange; Q3 = gray; Q4 = smallest loss of BCS (yellow)]. The full model included calving season and BCS1 as covariables.

Table 3.

Least square means (SE) for average daily milk yield (kg) up to 305 DIM by BCS category at multiple time points (top) and by ∆BCS category at different periods (bottom) in primiparous cows. The full model for the repeated measures analysis included calving season and BCS1 as covariables. Different superscripts within columns indicate P < 0.05

	Category of BCS at multiple time points
	Calving		14 DIM		21 DIM		Nadir
BCS category	LSM	SEM	LSM	SEM	LSM	SEM	LSM	SEM
Q1	34.4ab	0.014	34.8a	0.014	35.2a	0.014	35.8a	0.014
Q2	34.2ab	0.014	34.2b	0.014	34.3b	0.014	35.0b	0.018
Q3	34.7a	0.014	33.8b	0.018	33.8b	0.018	33.8c	0.014
Q4	33.8b	0.014	33.4b	0.023	32.9c	0.018	32.4d	0.014

	Time period between BCS assessments
	Calving to 7 DIM		Calving to 14 DIM		Calving to 21 DIM		Calving to nadir
∆BCS category	LSM	SEM	LSM	SEM	LSM	SEM	LSM	SEM
Q1	34.2ab	0.014	34.7a	0.014	35.5a	0.014	36.2a	0.014
Q2	34.5ab	0.014	34.5ab	0.014	34.6b	0.014	34.7b	0.014
Q3	34.8a	0.023	34.0bc	0.018	33.7c	0.014	33.7c	0.018
Q4	33.9b	0.014	33.4c	0.014	33.0d	0.014	32.4d	0.014

Table 4.

Least square means (SE) for average daily milk yield (kg) up to 305 DIM by BCS category at multiple time points (top) and by ∆ BCS category at different periods (bottom) in multiparous cows. The full model for the repeated measures analysis included calving season and BCS1 as covariables. Different superscripts within columns indicate P<0.05

	Time of BCS assessment
	Dry-off		Calving		21 DIM		Nadir
BCS category	LSM	SEM	LSM	SEM	LSM	SEM	LSM	SEM
Q1	44.9a	0.014	44.0a	0.018	44.0	0.014	45.1a	0.014
Q2	43.4b	0.014	43.9a	0.018	44.3	0.018	44.2b	0.018
Q3	43.0b	0.023	44.2a	0.017	44.0	0.018	43.3b	0.014
Q4	42.1c	0.023	43.6b	0.018	43.8	0.018	43.0c	0.023

	Time period between BCS assessments
	Dry-off to calving		calving to 14 DIM		calving to 21 DIM		calving to nadir
∆BCS category	LSM	SEM	LSM	SEM	LSM	SEM	LSM	SEM
Q1	42.4c	0.014	44.6a	0.014	44.4	0.014	45.0a	0.014
Q2	43.5b	0.014	43.9ab	0.018	44.0	0.014	44.5a	0.018
Q3	43.9b	0.023	43.5b	0.018	44.0	0.023	43.5b	0.018
Q4	44.9a	0.014	44.4ab	0.014	43.8	0.014	43.1b	0.014

When average milk yield (kg/d) at 305 DIM was analyzed, the differences in average daily milk yield between cows in the lowest and the greatest nBCS category (Q1 vs. Q4) were 3.4 kg/d (P < 0.001) for primiparous cows and 2.1 kg/d (P < 0.001) for multiparous cows. During the period from calving to nadir, primiparous cows in Q1 (greatest decrease of BCS) produced 3.8 kg/d (P < 0.001) more than cows in Q4. For multiparous cows, the difference for this period was 1.9 kg/d (P < 0.001) in favor of Q1 cows (Tables 3 and 4).

BCS at previous dry-off was also associated with subsequent milk yield in multiparous cows. Cows in Q1 at dry-off subsequently produced 2.8 kg/d (P < 0.001) more than cows in Q4. Nonetheless, this order was reverted for the ∆BCS from dry-off to calving, where a smaller loss resulted in greater subsequent milk yield. For ∆BCS from calving to nadir, cows in Q1 produced 2.5 kg/d (P < 0.001) more than cows in Q4 (Table 4).

Mean ± SD DIM at nadir for primiparous and multiparous cows were 43.1 ± 26.8 DIM and 47.8 ± 25.1 DIM (P < 0.0001), respectively. In both primiparous and multiparous cows, greater days to nBCS were associated with greater milk yield at 60 DIM and 305 DIM (Table 5).

Table 5.

Least square means (SE) for average daily milk yield (kg) up to 60 DIM and 305 DIM by time to nadir BCS category in primiparous (top) and multiparous cows (bottom). The full model for the repeated measures analysis included calving season and BCS1 as covariables. Different superscripts within columns indicate P < 0.05

	Primiparous
	Milk 60		Milk 305
Time to nBCS	LSM	SEM	LSM	SEM
Q1 (least DIM)	28.7a	0.163	32.3a	0.177
Q2	30.3b	0.172	34.1b	0.172
Q3	31.5c	0.168	35.4c	0.177
Q4	31.5c	0.168	35.1c	0.177

	Multiparous
Time to nBCS	LSM	SEM	LSM	SEM
Q1 (least DIM)	43.7a	0.159	43.3a	0.227
Q2	45.5b	0.154	43.9a	0.195
Q3	46.4c	0.154	44.2ab	0.204
Q4	47.3d	0.159	44.7b	0.172

Overall, 6,541, 3,218, and 2,283 cows were classified as healthy, affected by one, and affected by more than one health event, respectively. When cows were presented by occurrence of disease (healthy: one health event; or multiple events), the associations between nBCS and milk yield described for the lactation curves in the overall population were marginally altered in primiparous cows affected by postpartum health events. In the case of healthy cows, the lactation curves were differentiated by nBCS quartile categories, with greater production for cows in the lowest nBCS category (Q1), followed by cows in Q2, Q3, and Q4 (Table 6). This distinction among curves was less evident in the subpopulation of cows with one or multiple health events (Figure 7). Contrary to what was anticipated, these trends were not clear in multiparous cows, where cows with multiple health events categorized in the lowest BCS (Q1) maintained the greatest milk yield, compared with the other BCS categories in this group (Figure 8).

Table 6.

Least square means (SE) for average daily milk yield (kg) up to 305 DIM by nadir BCS category in primiparous (top) and multiparous cows (bottom). The full model for the repeated measures analysis included calving season and BCS1 as covariables. Different superscripts within columns indicate P < 0.05

	Health status
	No health disorder		One health disorder		Multiple health disorders
	Primiparous

BCS category	LSM	SE	LSM	SE	LSM	SE
Q1 (lowest BCS)	36.2a	0.018	35.3a	0.027	35.3a	0.036
Q2	35.5a	0.023	34.3ab	0.036	34.3ab	0.041
Q3	34.1b	0.018	33.5bc	0.032	33.4b	0.045
Q4	32.6c	0.018	32.2c	0.045	31.4c	0.050

	Multiparous
∆BCS category	LSM	SE	LSM	SE	LSM	SE
Q1 (greatest loss in BCS)	45.9a	0.014	44.8ab	0.014	44.7a	0.014
Q2	44.6b	0.014	45.1a	0.018	42.5b	0.014
Q3	44.0b	0.023	44.2ab	0.018	42.4b	0.023
Q4	43.5b	0.014	42.8b	0.014	41.8b	0.014

Figure 7.

Lactation curves for daily milk yield (kg) up to 305 DIM in primiparous cows by category of BCS at nadir, in cows with no health disorder (top panel), one health disorder (middle panel), and multiple health disorders (bottom panel) before nadir BCS. Values for nadir BCS were categorized using the quartile distribution [Q1 = lowest BCS (blue), Q2 = orange; Q3 = gray; Q4 = greatest BCS (yellow)]. The full model included calving season and BCS1 as covariables.

Figure 8.

Lactation curves for daily milk yield (kg) up to 305 DIM in multiparous cows by category of BCS at nadir, in cows with no health disorder (top panel), one health disorder (middle panel), and multiple health disorders (bottom panel) before nadir BCS. Values for nadir BCS were categorized using the quartile distribution [Q1 = lowest BCS (blue), Q2 = orange; Q3 = gray; Q4 = greatest BCS (yellow)]. The full model included calving season and BCS1 as covariables.

Figure 4.

Lactation curves for daily milk yield (kg) up to 305 DIM in primiparous cows by category of ∆BCS from calving to 7 DIM (top left panel), calving to 14 DIM (top right), calving to 21 DIM (bottom left), and calving to nadir (bottom right). Values for ∆BCS were categorized using the quartile distribution [Q1 = greatest loss of BCS (blue), Q2 = orange; Q3 = gray; Q4 = smallest loss of BCS (yellow)]. The full model included calving season and BCS1 as covariables.

Figure 5.

Lactation curves for daily milk yield (kg) up to 305 DIM in multiparous cows by category of BCS at dry-off (top left panel), calving (top right), 21 DIM (bottom left), and at nadir (bottom right). Values for BCS were categorized using the quartile distribution [Q1 = lowest BCS (blue), Q2 = orange; Q3 = gray; Q4 = greatest BCS (yellow)]. The full model included calving season and BCS1 as covariables.

Discussion

In line with our hypothesis, this study identified some significant associations between BCS and ∆BCS at specific times and average daily milk yield at 60 DIM and 305 DIM. Variation in the shape of the lactation curves calculated by BCS and ∆BCS categories was also evident at specific time points and time periods. Although these were anticipated findings that have been reported elsewhere (Broster and Broster, 1998; Roche et al., 2009), the current analyses indicated that these associations were variable and dependent on the time of BC assessment. Moreover, the magnitude of the differences in milk yield among categories of BCS and ∆BCS were conditional to the parity category.

The current study faces some limitations. To have sufficient individual information on both milk yield and BCS, our inclusion criteria required cows to have at least one AI postpartum (occurring at 80 DIM and 60 DIM in primiparous and multiparous cows, respectively). Moreover, this was intended to avoid the potential confounding effect of severe injury or disease. Nevertheless, as a result, 13.7% and 12.6% of the primiparous and multiparous cows were not considered in the results reported here. The use of only one dairy is a second limitation of this study that reduces the external validity of our findings.

The availability of daily BCS and milk yield data in the study population provided the opportunity for selecting specific time points and time periods to evaluate these potential associations and assess cows’ future performance early in the lactation. Interestingly, the differentials in milk yield for cows in various BCS and ∆BCS categories were more notorious when cows were scored through the progressing lactation. Greater milk yield was consistently associated with lower BCS and greater loss of BCS. In consequence, the largest volumes in milk yield in favor of cows in the lowest BCS or ∆BCS quartile category were mostly identified for cows scored at nadir and for the BCS change between calving and nadir.

Changes in BCS during the transition period are common to most cows, and fat mobilization from body reserves in response to energy unbalances occurring in early lactation is likely exacerbated in high-producing cows (Lean et al., 2013). As genetic selection supported by genomic evaluations has become highly efficient, the rate of progress for increased milk and solid production continues to increase (Council on Dairy Cattle Breeding, 2022; Wiggans and Carrillo, 2022). This increment in yield has resulted in homothetic changes in early lactation cows, resulting in adaptations in the nutrient partitioning and utilization to support high milk production (Bauman and Currie, 1980; Bell, 1995; Chagas et al., 2009).

The ability to properly direct nutrient partitioning, together with the level of energy stores and their mobilization dynamics are key components of milk production. As for milk yield, the genetic component determining variation in fat reserves seems to be significant. Early estimates of heritability for BCS ranged from 0.24 to 0.45 (Buttchereit et al., 2012; Gallo et al., 1999; Veerkamp, 1998; Veerkamp and Brotherstone, 1997) and the heritability of BCS change from week 1 to week 10 of lactation was reported to be 0.09 in an experimental herd (Pryce et al., 2000).

Regarding the association between energy reserves and milk yield, Dechow et al. (2001) reported genetic correlations in first lactation between BCS at calving and mature equivalent (ME) milk, ME fat, and ME protein ranging from −0.02 to 0.22. Negative genetic correlations were calculated through the remainder of the lactation and were strongest between BCS at pregnancy check and milk yield (−0.22 to −0.49). In the second lactation, genetic correlations were negative for ME milk and all BCS traits, including BCS at calving. Body condition score at dry-off also was negatively correlated to production in the subsequent lactation for the first- and second-parity cows.

Notably, in a more recent study completed in France, genetic correlations between loss of weight between weeks 1 and 5 after calving and daily milk production were moderate and ranged from −0.26 to 0.05 in the first lactation and from −0.11 to 0.10 in the second lactation, suggesting potential for selection for milk production without increasing early body mobilization (Tribout et al., 2023).

Related to the availability of energy stores, many studies have identified a positive effect of intermediate BCS of 3.0 to 3.3 at calving (Grainger et al., 1982; Markusfeld et al., 1997; Roche et al., 2009; Waltner et al., 1993) on milk production. This greater availability of energy for the cow allows for sparing glucose production for lactose synthesis. In contrast, if the optimum calving BCS is surpassed, milk yield is reduced, mostly as a result of lower DMI in over-conditioned cows (Garnsworthy and Topps, 1982). Moreover, excessive BCS mobilization to overcome the excessive BCS-mediated hypophagia will aggravate the risk of subclinical metabolic disorders, such as ketosis and fatty liver (Roche et el., 2009).

Interestingly, and contrary to earlier research supporting this concept (Grainger et al., 1982; Markusfeld et al., 1997; Waltner et al., 1993), no clear differences in milk yield were established in this study among BCS categories at calving for both primiparous and multiparous cows, as the only category with greater yield was Q3 in multiparous cows.

Notably, differences in milk yield among categories of BCS were more evident when the lactations progressed, starting at 14 DIM in primiparous and better identifiable at 21 DIM in multiparous cows. This delayed differentiation in milk production agrees with the idea that the rate of postcalving BCS gain or loss associated with milk production is a better reflection of energy balance than BCS at one specific point (Roche et al., 2009). Moreover, in support of this concept, in this study, the magnitudes of the differences among BCS/∆BCS categories were greater when changes in BCS, rather than BCS at fixed time points, were considered.

The lactation curves presented in this study suggest that the magnitude of the differences in milk yield among categories of BCS and ∆BCS were conditional to the parity category. As cows approaching their first calving require nutrients for their own continued growth in addition to that of their developing calf, they are in a differing metabolic state to that experienced by multiparous cows. Accordingly, parity influences the pattern of changes in metabolic hormones and metabolites following calving (Meikle et al., 2004; Wathes et al., 2007). Other differences previously identified in the first-parity cows included higher concentrations of IGF-I and leptin throughout the peripartum and early lactation, which would affect the level of partitioning of nutrients into milk (Wathes et al., 2007). Moreover, in a recent report from Barragan et al. (2020), primiparous cows also had higher postpartum circulating concentrations of substance P, haptoglobin, and cortisol compared with multiparous cows.

A significant finding in multiparous cows was the clear association of BCS at dry-off and milk yield in the following lactation. The difference between the two extreme BCS categories was 1.3 kg/d (60 DIM) in favor of cows in the lowest BCS quartile category. This finding is supported by Domecq et al. (1997) who reported that each additional point of body condition at dry-off was associated with 300 kg less milk at 120 d of lactation.

Contrary to the trends identified in lactating cows, future milk yield was greater in cows that gained BCS (Q4) during the dry period, with 2.1 kg/d (60 DIM) more milk in Q4 than in Q1 cows. Interestingly, data reported by Daros et al. (2021) indicates that the BCS at dry-off is associated with changes in feeding behavior, with over-conditioned cows having lesser daily DMI and feeding time during the early and late dry periods compared with not over-conditioned animals. In addition, the authors identified an effect of the previous 305-d milk yield on DMI; cows that produced more milk had greater DMI throughout the dry period. It is also plausible to suggest that high-producing cows would maintain a moderate BCS through the whole lactation, whereas low producing cows would tend to gain more weight, increasing their BCS as dry-off approaches. In consequence, this group of cows with higher milk yield would coincide with the cows that had lower BCS during their previous dry-off. Moreover, connecting these concepts with our results in milk yield and ∆BCS during the dry period, a strong positive correlation has been described between prepartum and postpartum DMI (Grummer et al., 2004), suggesting that greater milk yield and improved performance should be obtained by minimizing the decline in DMI prepartum.

The associations determined during the dry period in this study also align with data presented by Chebel et al. (2018), where cows that gained BCS during the dry period had greater yield of milk in the subsequent lactation. Mean daily milk yield over 305 d was highest for cows gaining BCS during the dry period (40.6 kg/d), followed by no change in BCS (39.6 kg/d), moderate loss of BCS (39.2 kg/d) and excessive loss of BCS (39.3 kg/d). Similarly, Domecq et al. (1997) reported that a one-point increase in BCS between dry-off and parturition was associated with 4.5 kg/d more milk in the first 120 d of lactation. Nonetheless, it is reasonable to assume that the homothetic mechanisms involved in lipid mobilization during the precalving period are probably different from those in cows that are starting their lactation (Ghaffari et al., 2019; Schuh et al., 2019).

The greater discriminatory value of BC scoring farther from calving was evident when comparing lactation curves, with the best separation among BCS and ∆BCS categories identified at nadir (Figures 3 to 6). This separation in milk yield was better represented in primiparous than multiparous cows, where milk curves overlapped at multiple points across time. Notably, in this group, the greatest separation in lactation curves was determined for categories of BCS at dry-off and for the period dry-off to calving, although with an inverse order. Cows that were thinner at dry-off subsequently produced more milk than cows that had greater BCS. On the contrary, cows gaining BCS during the dry period were the individuals with the greatest milk yield.

Since the nadir BCS represents the lowest level in BC, this point provides a single measurement that could be used as a reference for decision-making at the farm. Previous research in our group identified clear associations between nBCS and pregnancy at AI1, with longer DIM at nadir, lower nBCS, and greater reductions in BCS between calving and nadir characterizing cows that failed to conceive at AI1 (Hernandez-Gotelli et al., 2023). In consequence, since a lower nBCS was associated with greater milk yield, the opposite directions in the associations of nBCS with milk yield and fertility suggest a compromise when making management and selection decisions involving these two key areas of cow performance.

Time to nadir BCS was associated with milk yield. In a previous study from our group (Hernandez-Gotelli et al., 2023), sick cows and high-producing cows had increased DIM at nadir. In addition, cows that conceived at AI1 had their nBCS earlier (45.1 ± 0.32 d) than cows that remained open after the first AI (46.4 ± 0.35 d). It is plausible to hypothesize that high-producing cows that are utilizing their energy reserves for lactogenesis would take longer to revert the descending trend in BCS. Likewise, cows with longer time losing BCS would be more likely to exhibit worsened reproductive performance.

A secondary aim of this study was to test the effect of disease on the anticipated associations of BCS and milk yield. Health status has been identified as a factor affecting the magnitude and direction of BCS changes postpartum (Chebel et al., 2018; Melendez et al., 2020) and in a recent report, cows with multiple disease events had significantly smaller BCS at that point and took longer to reach nBCS (Hernandez-Gotelli et al., 2023).

The assumption in the current study was that the association of high milk yield and greatest BCS loss would only hold in the group of healthy cows, while sick cows with large BCS loss will not evidence high production. The change in BCS between calving and nadir was used for this purpose.

Nevertheless, the proposed hypothesis was not sustained by the results. In primiparous cows, the separation in milk curves among categories of ∆BCS between calving and nadir partially remained for cows with one or multiple health events. Interestingly, for multiparous cows, the trend for greater milk yield in cows losing more BCS (Q4) was more evident in animals with multiple disease events, as compared with healthy cows or with cows with one health event. A possible reason for this unexpected finding is that high-producing cows are more prone to metabolic disorders such as ketosis (Benedet et al., 2019), and therefore, this group losing more BCS would include the top-producing individuals. In addition, as our inclusion criteria required survival until the first AI postpartum, cows affected with the most severe health issues that likely died or left the herd early in their lactation were not considered in our analyses.

Conclusions

The associations between BCS and ∆BCS categories and milk yield were not consistent and depended on the time of assessment and the parity of the cow. Nonetheless, as the assessment of BCS and ∆BCS approached to the nadir, the association between greater milk yield and use of energy reserves became more evident. These findings suggest that the magnitude of the nadir BCS is a significant factor involved in the interrelations among energy balance, use of energy reserves, and potential for milk production.

Glossary

Abbreviations

AI

artificial insemination

BCS

body condition score

BCS1

body condition score at calving

BCS7

body condition score at 7 d in milk

BCS14

body condition score at 14 d in milk

BCS21

body condition score at 21 d in milk

BCSdry

body condition score at dry-off

∆BCS

BCS change

DIM

days in milk

LSM

least square means

ME

mature equivalent

nBCS

nadir BCS

Q

quartile

SD

Standard deviation

SE

Standard error