Our study identified dry-period and periparturient-period nutritional strategy relationships with metabolic- (NEFA and BHB) and inflammation-related (Hp) analytes, DI, milk yield, and reproductive performance. Evaluating nutritional strategies provides dairy nutritionists with guidelines for the best strategy to implement on the farm for cow performance, while allowing for the flexibility to adjust the nutrients to suit the needs of the cows.
To our knowledge, there have not been any large, observational studies evaluating nutritional strategies on transition-period health and performance outcomes on dairy farms; therefore, our discussion will focus on results reported in controlled research trials evaluating the plane of energy or overall nutrient supply. It is important to note major differences between our study and previous controlled research trials. First, one of the biggest differences we observed between the present study and previous controlled research trials was the vast difference in the energy concentration of the treatment strategies in many of the controlled trials. There are moderate differences in the mean nutrient concentration in our defined nutritional strategies (
Table 1). Based on the authors’ extensive field involvement and experience with commercial dairies in the northeastern United States, the nutrient concentrations are within the typical use range. It is of note that many controlled studies have tested the limits or have utilized treatments that have discernable differences (e.g., 14% vs. 26% starch) as a way of investigating biological functions; therefore, some of these diets may not be representative of diets typically fed on a farm [
17,
18]. Secondly, some of these controlled research trials have tested treatments that involved feed restriction, which is not advised on farms as it can alter feeding behavior [
19]. For the purpose of this discussion, results from restricted-fed treatments within studies will not be included. Thirdly, our analysis was conducted at the herd-level, whereas, in controlled research trials, it was analyzed at the cow-level; therefore, our outcomes may differ slightly compared to those in the literature. In addition, we are not able to infer possible cow-level biological mechanisms due to the analysis being conducted at the herd-level and missing critical cow-level data, such as individual dry matter intake (DMI); however, herd-level associations between the proportion of cows with elevated analytes and postpartum performance outcomes have been evaluated and discussed previously for this dataset [
15]. Lastly, the controlled research trials discussed herein evaluated cows in tiestalls and not freestalls; therefore, the cows observed in this study were subjected to additional environmental factors, such as negative social interactions and increased competition, which may alter the cows’ feeding and lying behavior. This may provide one possible explanation as to why differences were observed in this study compared to controlled research trials. For these reasons, this study provides external validity for controlled research trials.
4.1. Dry-Period Nutritional Strategies
In our study, the dry-period nutritional strategy was only associated with the prevalence of elevated prepartum NEFA, postpartum NEFA, BHB and Hp concentrations, 21-d PR and PP outcomes. Since most controlled research trials evaluating nutritional strategies have not evaluated reproduction, reproductive outcomes will be reviewed later in the discussion. Overall, results from our dry-period nutritional strategy models were similar between multiparous and primiparous cows.
Although an interaction between far-off and close-up nutritional strategies was observed for the prevalence of elevated prepartum NEFA concentrations, significant differences between common nutritional strategies were not observed; however, multiparous cows in HF-fed herds had a higher prevalence of elevated prepartum NEFA concentrations compared to LF-fed herds. Similarly, Janovick et al. [
4] and Mann et al. [
5] reported higher prepartum NEFA concentrations in multiparous cows fed a controlled-energy dry-period diet compared to a high-energy dry-period diet. In another study observing primiparous and multiparous cows fed a controlled-energy far-off diet and either a controlled-energy or high-energy close-up diet, Vasquez, et al. [
20] also reported high prepartum NEFA concentrations in cows fed the controlled-energy close-up diet.
Contrary to our hypothesis, an interaction between far-off and close-up nutritional strategies was not observed for the prevalence of elevated postpartum NEFA concentrations for multiparous cows. In addition, an interaction between far-off and close-up nutritional strategies on the prevalence of elevated postpartum NEFA concentrations was observed for primiparous cows, although significant differences between common nutritional strategies were not observed; however, HF-fed herds had a higher prevalence of elevated postpartum NEFA concentrations than LF-fed herds. Contrary to our results, previous studies reported lower postpartum NEFA concentrations for cows fed a controlled-energy diet compared to a high-energy diet [
4,
5,
6,
20]. Our analysis was at the herd-level and not the cow-level; therefore, it is plausible that if the unit of observation was the cow, we may have detected differences in postpartum NEFA concentrations for multiparous cows. Our herd-level analysis evaluated if there was a difference in the proportion of cows above the identified threshold.
For multiparous cows, an interaction between far-off and close-up nutritional strategies on the prevalence of elevated BHB concentrations was not observed; however, there was a greater prevalence of elevated BHB concentrations in herds where multiparous cows were fed an LF close-up diet compared to a HF diet. For primiparous cows, there was an interaction between the far-off and close-up strategy such that CE × HF- and NCE × LF-fed herds had a similar prevalence of elevated BHB concentrations; however, only NCE × LF-fed herds had a statistically lower prevalence of elevated BHB concentrations than CE × LF-fed herds. These results support controlled research findings for multiparous cows showing that feeding a controlled-energy close-up diet can minimize increases in postpartum BHB concentrations [
4,
5,
6]; however, results for primiparous cows are not consistent. Richards et al. [
6] observed greater postpartum BHB concentrations in primiparous and multiparous cows fed a high-energy diet throughout the dry period compared to cows fed a controlled-energy diet throughout the dry period or a controlled-energy far-off followed by a high-energy close-up diet. Similarly, Janovick et al. [
4] observed greater postpartum BHB concentrations in primiparous cows fed a high-energy dry-period diet compared to a controlled-energy dry-period diet.
Contrary to our hypothesis, an interaction between far-off and close-up nutritional strategies was not observed for the prevalence of elevated Hp concentrations for primiparous or multiparous cows, although cows in HF-fed herds had a higher prevalence of elevated Hp concentrations than LF-fed herds. To our knowledge, Hp concentrations have not been evaluated in controlled research trials evaluating prepartum nutritional strategies and should be evaluated in the future to explore possible biological mechanisms.
An association was not observed between far-off and close-up nutritional strategies and DI. Although caution should be used when interpreting health data in controlled research trials due to sample size, Mann et al. [
5] reported half as many hyperketonemia cases and no clinical ketosis cases within 3 wk postpartum for cows fed a controlled-energy diet during the entire prepartum period compared to cows fed an intermediate (controlled-energy far-off diet and intermediate-energy close-up diet) or high-energy diet (high energy during the entire prepartum period). Janovick et al. [
4] observed an increased incidence of displaced abomasum and ketosis, and had more cows with more than 1 negative health event, regardless of parity, for cows fed the high-energy prepartum diet for the entire dry period compared to cows fed the controlled-energy diet for the entire dry period. On the contrary, there were not any discernable differences in negative health events between cows that were fed a controlled-energy diet for the whole dry period, a step-up diet (controlled-energy at dry-off, switching to high-energy 21 d before expected parturition), or a high-energy diet for the whole dry period in the study by Richards et al. [
6].
As hypothesized, there was not an association between far-off and close-up nutritional strategies on milk yield (WK4MP or ME305). Our results are similar to previous studies [
4,
5,
6,
8], which did not report a difference in milk yield between different dry-period nutritional strategies.
In general, previous studies investigating transition cow nutritional strategies have not considered the fresh diet when investigating dry-period nutritional effects on postpartum performance, metabolic, and health outcomes. Most previous studies had cows transition onto a higher-energy fresh cow diet [
4,
6,
8,
20]; however, Mann et al. [
5] had cows transitioning onto a lower-energy fresh cow diet (21.2% starch, 35.4% NDF). In addition, these previous studies only considered the overall dry-period strategy (i.e., the interaction between the far-off and close-up strategy) as the treatment and did not investigate the far-off and close-up nutritional strategies in a 2 × 2 factorial study design, which would have been a more similar comparison to the current observational study design.
4.2. Periparturient-Period Nutritional Strategies
The periparturient-period nutritional strategy models identified close-up and fresh-period nutritional strategy associations with the prevalence of elevated postpartum NEFA and Hp concentrations for primiparous cows, prevalence of elevated postpartum BHB concentrations, DI, PP for primiparous cows, and PR for multiparous cows. This section of the discussion will focus on the few controlled research studies that have evaluated the effects of the interaction between close-up and fresh-period nutritional strategies. As stated previously, most controlled research trials evaluating nutritional strategies have not evaluated reproduction; therefore, we will review reproductive outcomes later in the discussion.
In the current study, an interaction between the close-up and fresh nutritional strategies was observed for the prevalence of elevated postpartum NEFA concentrations such that primiparous cows in herds fed a HF × HS diet had a higher prevalence of elevated NEFA concentrations than LF × HS- or HF × LS-fed herds and LF × LS-fed herds had a higher prevalence than LF × HS-fed herds. In a 2 × 2 factorial design, Rabelo et al. [
17] fed multiparous and primiparous cows a low-energy (1.58 Mcal/kg NE
L, 39.7% NDF, and 38.2% non-fiber carbohydrate (NFC)) or high-energy (1.70 Mcal/kg NE
L, 32.2% NDF, 44.6% NFC) close-up dry-period diet and a low-energy (1.57 Mcal/kg NE
L, 29.9% NDF, and 41.1% NFC) or high-energy (1.63 Mcal/kg NE
L, 24.9% NDF, 47.2% NFC) fresh-period diet. The authors observed a main effect of close-up nutritional strategy on postpartum NEFA concentrations: cows fed the high-energy prepartum diet had lower NEFA concentrations than cows fed a low-energy prepartum diet. The authors did not observe an interaction between the close-up and fresh strategy for postpartum NEFA concentrations, unlike the present study; however, both prepartum treatments in the Rabelo et al. [
12,
17] study were much higher in NE
L and NFC concentrations than in our study. In another 2 × 2 factorial design, Haisan et al. [
18] fed multiparous and primiparous cows a control (14.0% starch, 47.7% NDF) or high-starch (26.1% starch, 37.8% NDF) close-up diet and a high-fiber (25.1% starch, 33.8% NDF) or high-starch (32.8% starch, 27.2% NDF) fresh diet. The authors did not observe an interaction between close-up and fresh-period nutritional strategies; however, there was a main effect of close-up and fresh-period nutritional strategy such that cows fed the high-starch prepartum or postpartum diet had higher postpartum NEFA concentrations than the cows on the control prepartum or high-fiber postpartum treatment.
In our periparturient model, an association between the main effect of close-up strategy and the prevalence of elevated BHB concentrations was observed, such that cows in LF-fed herds had a higher prevalence than HF-fed herds. Similar to our results, Haisan et al. [
18] observed an association between the close-up strategy and BHB concentrations at 10 DIM such that cows fed the high-starch close-up diet had greater BHB concentrations than cows fed the control close-up diet; however, Rabelo et al. [
17] did not observe an effect of close-up nutritional strategy on postpartum BHB concentrations.
We also observed an interaction between the close-up and fresh nutritional strategies for DI such that herds that fed cows a HF × HS or LF × LS diet had a lower DI compared to herds that fed cows a HF × LS or LF × HS diet. Contrary to our findings, Haisan et al. [
18] reported a greater incidence of negative health events in cows fed the high-starch prepartum and postpartum diets, but there was not a discernable difference amongst the other nutritional strategy combinations. It is important to note that there are vast differences in the starch concentration of the high-starch close-up and fresh diets reported by Haisan et al. [
18] compared to the current study; therefore, the high-starch close-up and high-starch fresh strategy does not adequately reflect any of the strategies observed in our study. As stated before, caution needs to be used when interpreting health event data from controlled research trials due to a limited sample size. The study by Haisan et al. [
18] is the only study that reported health event data when investigating periparturient nutritional strategy effects.
Contrary to our hypothesis, an association was not observed between the interaction of close-up and fresh nutritional strategies and milk yield outcomes (WK4MP or ME305). Rabelo et al. [
12] did not observe any treatment effects on milk yield through 20 DIM; however, Haisan et al. [
18] reported greater milk yield through 20 DIM in cows fed the control prepartum diet and high-starch postpartum compared to any other treatment. We also hypothesized that cows fed HF close-up and HS fresh diets would have an increased prevalence of elevated Hp concentrations compared to cows fed HF close-up and LS fresh diets; however, an association between periparturient nutritional strategies and the prevalence of elevated Hp concentrations was not observed. Haisan et al. [
18] did not observe an interaction between prepartum and postpartum nutritional strategies on Hp concentrations; however, cows fed the high-starch postpartum diet had lower Hp concentrations than cows fed the high-fiber postpartum diet.
Our results and others [
12,
18] indicate that the interaction between the close-up and fresh diet should be considered when evaluating certain outcomes.
4.3. Fresh-Period Nutritional Strategies
In the periparturient models, the main effect of fresh-period nutritional strategy was associated with the prevalence of elevated BHB concentrations for all cows and the prevalence of elevated Hp concentrations for primiparous cows. In agreement with our hypothesis, herds with cows fed a higher-starch fresh diet had a lower prevalence of elevated BHB concentrations than herds with cows fed a lower-starch fresh diet. McCarthy et al. [
13,
21] evaluated dietary starch in the fresh period and fed a high-starch (26.2% starch, 34.3% NDF, 1.64 Mcal/kg NE
L) or low-starch (21.5% starch, 36.9% starch, 1.56 Mcal/kg NE
L) diet through 21 DIM before all cows were fed the high-starch diet through 63 DIM. Supporting our results, the authors observed lower BHB concentrations through 21 DIM, regardless of parity, in cows fed the high-starch diet compared to the low-starch diet. Rabelo et al. [
17] also observed a postpartum treatment by time interaction such that BHB concentrations were lower for cows fed the high-starch postpartum diet at 7 and 21 DIM. Our results and others [
17,
21] support the notion that increasing dietary starch immediately after parturition, rather than delaying, as in a step-up fresh approach, is effective in minimizing the increase in BHB concentrations; however, these results are not consistent in the literature. Dann and Nelson [
22] did not observe a difference in BHB concentrations for the first 3 wk of lactation when feeding a low-starch (21.0% starch through 91 DIM), medium-to-high-starch (23.2% starch for 21 DIM then 25.5% until 91 DIM) or high-starch (25.5% starch through 91 DIM) postpartum diet. Sun and Oba [
23] also did not observe treatment or a treatment by week interaction for BHB concentrations when feeding high-starch (29.2% starch) or low-starch (19.1%) diets from parturition through 12 wk of lactation.
In agreement with our hypothesis, herds that fed primiparous cows an HS fresh diet had a higher prevalence of elevated Hp concentrations than herds that fed primiparous cows an LS fresh diet. Supporting our results, McCarthy et al. [
21] observed higher Hp concentrations through 15 DIM in cows fed a high-starch compared to a low-starch diet, regardless of parity; however, Haisan et al. [
18] observed the opposite results. Regardless of parity, Haisan et al. [
18] reported lower serum Hp concentrations in cows fed the high-starch fresh diet compared to the high-fiber fresh diet; however, it is important to note that the high-fiber diet was very similar to the high-starch diet in the study by McCarthy et al. [
21].
Contrary to our hypothesis, we did not observe an association between fresh-period nutritional strategy and the prevalence of elevated postpartum NEFA concentrations, milk yield (WK4MK or ME305), or DI. Dann and Nelson [
22] reported higher NEFA concentrations in cows fed the medium-starch diet through 21 DIM compared to the high-starch diet, yet greater milk yield in cows fed the low-starch diet compared to the high-starch diet. McCarthy et al. [
21] reported lower NEFA concentrations, no difference in the frequency of health events, and greater early-lactation milk yield [
13] for cows fed a high-starch fresh diet compared to a low-starch diet. Sun and Oba [
23] reported higher NEFA concentrations for primiparous (wk 2, 6, and 8 postpartum) and multiparous cows (wk 3 and 4) fed the low-starch postpartum diet compared to the high-starch diet. It is important to note that cows fed the low-starch postpartum diet stayed on that diet through 12 wk of lactation instead of transitioning onto a higher-energy diet earlier in lactation to maintain milk yield, as typically observed in a step-up fresh approach. Interestingly, Sun and Oba [
23] did not observe a difference in milk yield between treatments. Although we did not observe any associations with these outcomes, discrepancies amongst the literature when evaluating the starch concentration during the fresh period may be due to differences in the prepartum or postpartum diets. It has been proposed that hypophagic effects may be observed when providing highly fermentable starch sources in fresh diets [
10,
24]. However, research within the last decade would indicate there might be an interaction between starch or starch digestibility with forage or forage NDF levels on postpartum performance and health [
25]. Favorable responses have been observed when feeding a higher-starch or starch digestibility fresh diet with higher forage or forage NDF concentrations [
13,
21,
26] while neutral or negative responses have been observed when feeding a higher-starch fresh diet with lower forage or forage NDF concentrations [
18,
22,
23,
24]. Results from a study by Tebbe and Weiss [
27] suggest that primiparous cows may benefit from a fresh diet with higher starch and lower forage NDF; however, this study lacked a higher-starch and forage NDF fresh diet treatment, prompting the need for further investigation.
4.4. Nutritional Strategies on Reproductive Outcomes
Very limited research has evaluated the effects of nutritional strategies pertaining to energy on reproductive outcomes. This may be due to the limited sample size within the study or the need for an increase in follow-up time; however, Cardoso et al. [
28] pooled results within their group from seven studies evaluating prepartum nutrition to evaluate the effects of nutrition on reproductive performance (
n = 354 multiparous cows, 54 primiparous cows). Far-off- and close-up-period nutritional strategies were assigned as controlled-energy (≤100% of NE
L requirements) or high-energy (>100% of NE
L requirements). All the lactating diets supplied to the cows in the Cardoso et al. [
28] study appeared to be similar and were higher-energy diets (≥1.67 Mcal/kg NE
L). Cardoso et al. [
28] reported less days to pregnancy when cows were fed a controlled-energy diet during the close-up period compared to the high-energy close-up diet; however, there were no differences in days to pregnancy when evaluating the effects of far-off-period nutrition. Similar to Cardoso et al. [
28], Vickers et al. [
7] reported greater odds of pregnancy at 120 and 150 DIM for cows fed a controlled-energy far-off and close-up diet (13.6% starch, 48.4% NDF, 1.41 Mcal/kg NE
L) compared to cows fed the controlled-energy far-off and a higher energy close-up diet (16.3% starch, 41.3% NDF, 1.45 Mcal/kg NE
L). These results correspond to our data for primiparous cows, but not for multiparous cows. In our study, herds fed a CE × HF or NCE × LF dry-period strategy had a higher 21-d PR for primiparous cows than CE × LF-fed herds. For multiparous cows, an interaction was not observed between far-off and close-up nutritional strategies; however, LF close-up-fed herds had a higher 21-d PR than HF-fed herds. In addition, an interaction between close-up and fresh nutritional strategies was observed for primiparous cows such that herds that fed a HF × HS diet had a higher PP than all other nutritional strategy combinations. It is important to note that we observed an interaction between far-off and close-up nutritional strategy for primiparous cows, while Cardoso et al. [
28] did not evaluate the interaction between far-off and close-up nutrition.
It has been suggested that feeding controlled-energy diets during the prepartum period will prevent the over-consumption of nutrients during the prepartum period and promote DMI during the early postpartum period, thus reducing the degree of negative energy balance compared to feeding high-energy prepartum diets [
9,
29]. The severity of negative energy balance in early lactation has been associated with pre- and postovulatory reproductive failure as it coincides with follicular development and uterine involution [
30]. Follicles and oocytes will ovulate from 50 to 60 d after development in the early lactation period [
29]; therefore, nutrition during the early postpartum period can play a critical role in determining if ovulation occurs after the herd voluntary waiting period. Although research on optimizing reproductive performance through postpartum nutritional strategies is limited, glucose is required for oocyte maturation and the development of the blastocyst and it has been suggested to feed more starch (more glucogenic diet) to increase blood insulin and reestablish the growth hormone-insulin-like growth factor 1 axis to resume ovarian activity [
9,
29].
4.5. Limitations and Strengths
Although we intended to adequately classify these herds into different nutritional strategies, there are many limitations to this study. In this study, we focused on starch and forage inclusion as a proxy for energy intake but we did not account for other nutrients, such as amino acids, that may provide substrates for the tricarboxylic acid cycle, or vitamins and minerals that may have an impact on immune function and reproductive performance. Our results may have been confounded if the dietary cation–anion difference varied greatly across nutritional strategies; however, we do not believe this to be an issue as commercial anionic supplements were included in the majority of close-up diets for multiparous cows (77.8% of farms [
14]) and there was not a difference in the distribution of herds implementing commercial anionic supplements between the nutritional strategies within each period. Individual or pen-level DMI, which would have provided additional information with regard to energy balance, could also be a confounder that was not accounted for due to herd-recording limitations. Forage quality was not accounted for, which may influence DMI, such as the forage chop length and TMR particle size distribution, moisture, or if the forage was spoiled or moldy. The time spent on each of these diets was also not accounted for. Particularly in the fresh period, cows may have been on the fresh cow diet for 14 to 30 d, which may limit intake depending on the concentration of forage NDF [
31]. We also did not assess the overall nutritional strategy of the herd as we were not powered for that many comparisons. Despite these limitations, this observational study was performed using a large number of freestall, commercial dairy herds versus tiestall herds, allowing for the influence of other environmental factors. We were able to evaluate if current nutritional strategy recommendations are supported in the field, despite the variation observed with other management factors.
Nutritional strategies are multifactorial and are likely driven by a wide range of factors, such as the body condition score of the cow, the interaction between nutrients in the diet, and social and environmental factors. These factors may explain why one strategy may work well on one farm but lead to an increase in fresh cow health disorders or decreased milk yield in other herds. As stated by Van Saun and Sniffen [
32], “In the current transition cow system, a range of feeding program and grouping strategies are observed, with no one approach consistently resulting in the desired outcome”.