Glucose intolerance in dairy goats with pregnancy toxemia: Lack of correlation between blood pH and beta hydroxybutyric acid values

Abstract

This study assessed the response to a glucose tolerance test in dairy goats with pregnancy toxemia (PT), in healthy, pregnant, non-lactating dairy goats in the last month of gestation (HP), and in healthy, lactating, non-pregnant, dairy goats in mid-lactation (HL). A 500 mL volume of a 5% glucose solution was administered by the IV route. Blood glucose concentrations returned to pre-infusion levels by 90 min in all 8 HL goats, and by 180 min in all 8 HP goats. In contrast, concentrations of blood glucose were still significantly above pre-infusion levels at 180 min post-infusion in all 8 PT goats. Thus, marked glucose intolerance was demonstrated in the PT goats, and mild intolerance was noted in the HP goats. In 25 goats diagnosed with PT and having blood beta hydroxybutyric acid (BHBA) values ≥ 2.9 mmol/L, the correlation coefficient for BHBA with blood pH was non-significant.

Introduction

Several authors have recommended that dairy goats that are acutely ill, pregnant, non-lactating and diagnosed with pregnancy toxemia (PT), which is always accompanied by hepatic lipidosis (1,2), should be treated with glucose by the IV route (1,2). However, in our experience with obese PT does, IV administration of glucose has provided little benefit, as evidenced by either reduced severity of clinical signs or reduced case fatality rates, whether the patients were hypoglycemic or normoglycemic at the time of treatment (3). Because we had previously demonstrated glucose intolerance in a few clinical cases of PT in does that were not included in any formal study, we decided to explore the possibility that glucose intolerance might explain why our PT does fail to respond to glucose therapy.

A glucose tolerance test (GTT) is used to assess the ability of an animal to process a large glucose load that is administered by either the IV or oral route. The IV route has been the route utilized most often by veterinary clinicians in domestic animals (4–6). In most species in which glucose tolerance testing has been performed in normal, healthy individuals, blood glucose values return to pre-treatment levels by 90 min following glucose administration (4).

Glucose intolerance can result from an insufficient insulin secretion response (5), insufficient effectiveness of insulin (insulin resistance) (5), increased amounts of antagonistic (counter-regulatory) hormones such as glucocorticoids (5), and from liver dysfunction (resulting in reduced glucose uptake and storage in the liver) (5,7). The authors postulate that glucose intolerance observed in PT goats and healthy, pregnant, non-lactating dairy goats, in the last month of gestation (HP goats) was caused (at least in part) by insulin resistance, which has been consistently demonstrated in pregnant sheep (8) and cattle (9–13). Reduced insulin secretion could also have been a contributing factor (14), and impaired hepatic function might also have played a role (15), in view of the marked hepatic lipidosis that was observed in 6 of the fatal PT cases.

Obesity in sheep is associated with a decrease in insulin sensitivity (16,17) which was suggested to be consistent with decreased insulin receptors in peripheral tissues (17). Insulin resistance during late gestation might possibly be an etiologic factor in PT in ewes (8).

Pregnancy toxemia appears to be related to the glucose demands of the fetal-placental unity, due to the rapid growth of large or multiple fetuses during the last month of gestation (18). The pregnant female must cope with this negative energy balance through metabolic adaptations (19). The consequences of this negative energy balance for the dam are hypoglycemia, lipid mobilization, and accumulation of ketone bodies (20). With lipid mobilization, large amounts of non-esterified fatty acids are delivered to the liver. A certain percentage of these fatty acids are esterified in the cytosol, forming triglycerides and favoring fat accumulation in the liver (19). At least in ewes, however, it is still uncertain whether the increase in energy demands of late gestation can fully explain the appearance of PT (21).

During late gestation, ewes have reduced ability to metabolize BHBA, thereby promoting hyperketonemia (21) which exerts adverse effects on energy balance and glucose metabolism (21). It appears that impaired ketone body utilization in late pregnancy facilitates development of PT, especially in ewes carrying twins (21). Insulin resistance has also been documented in dairy cows with induced or spontaneous hepatic lipidosis and ketosis (9–13), and in humans with spontaneous hepatic lipidosis and ketosis (22).

Pregnancy toxemia occurs during the last month of gestation in ewes and does (1,2). Obese goats that are pregnant with multiple fetuses are particularly vulnerable to PT (23). It is believed that large amounts of accumulated intra-abdominal fat and rapid expansion of the uterus during late gestation combine to reduce the space available for ruminal expansion. As a result, these animals are unable to consume the volumes of feedstuffs needed to meet their energy requirements (20,23,24). A decrease in the plane of nutrition during the latter half of pregnancy, coupled with a short period of food deprivation, are also very important risk factors, especially in ewes bearing twins and triplets (20). Excessively thin ewes are also at high risk (20).

The primary objective of this study was to assess the response to an IV glucose load in goats diagnosed with PT, and to compare that response to responses in HP goats, and in healthy, non-pregnant goats in mid lactation (more than 100 days in milk) (HL goats). A physical examination was performed and biochemical blood values were determined in all goats. A second objective was to determine the extent to which values for blood BHBA and blood pH might be correlated in 25 does diagnosed with PT.

Materials and methods

This study was done on a 2000-head dairy goat farm, located 48 km northeast of Lisbon, Portugal. The goats were of 2 breeds, Alpine (~1200 animals) and Saanen (~600 animals) with some animals being crosses of these 2 breeds (~200 animals). All goats were continually housed in confinement, and all adult lactating does had access to free stalls and were fed a complete ration, ad libitum. Their Total Mixed Ration (TMR) consisted of corn silage (30%), alfalfa hay (22%), brewer’s grains (22%), rye-grass hay (8%), and a concentrate mix (18%). The TMR was offered once per day. All the adult goats had free access to mineral blocks containing NaCl, magnesium, and trace amounts of iron, iodine, cobalt, manganese, zinc, and selenium.

There were 3 kidding seasons per year, beginning in January, April, and October. Each kidding season began on the first day of the month, and continued for 45 d. Daily milk production in this herd averaged approximately 3 L/doe. Machine milking was performed twice daily. One month before their estimated kidding dates, lactating pregnant does were moved to a separate dedicated pen, the TMR was withdrawn, milking ceased, and wheat straw was fed, ad libitum. In addition, 1 kg/head per day of a different concentrate mix was provided, split into 4 feedings, equally distributed throughout each 24-hour period. This ration change was intended to reduce the rate of occurrence of PT (which ranged from 2.0% to 6.9% during this study), by reducing overconsumption of calories, excessive weight gain, and abdominal fat accumulation, during the last 30 d of gestation.

A second study involved 25 PT goats, whose blood BHBA values were compared with their blood pH values. This study included 8 PT goats that were enrolled in the glucose tolerance test (GTT) study, plus 17 PT goats not previously involved in any formal study.

Experiment 1 involved 24 does, divided into 3 groups of 8 (PT, HP, and HL). Three PT does gave birth in April/May 2013, 2 in October/November 2013 and 3 in January/February 2014. The rates of PT occurrence (number of cases of PT divided by the number of does kidding) during each of these 3 kidding periods were 22/415 (5.3%), 7/356 (2.0%), and 29/418 (6.9%), respectively.

Any doe in the dry doe pen that did not show normal interest in eating when fresh concentrate mix was offered was considered to be a PT suspect, and was tested for urine ketone bodies (if a urine sample could be obtained). Ketonuria and aciduria were confirmed in 4 does. A blood sample was obtained from every PT suspect, and tested for BHBA on a day when the veterinarian was present on the farm. All 25 of the PT does enrolled in this study had individual blood BHBA values between 2.9 and 8.0 mmol/L. Additional (more definitive) clinical signs of PT, such as persistent recumbency, depression, swollen limbs, rapid respirations (polypnea), inability to rise, stand and walk, and/or neurological signs would also trigger a request for the patient to be examined by a veterinarian, and for the blood BHBA concentration to be determined.

The following clinical signs were observed in the 8 PT does in this study: anorexia (n = 8 does), absence of ruminal motility (n = 7 does), sternal recumbency, but able to rise upon stimulation (n = 6 does), swollen limbs (n = 5 does), neurological abnormalities, such as “star gazing” (opisthotonus) (n = 2 does), and abnormally drooped ears (n = 2 does). Six of the 8 PT does died, even though kidding was induced using dexamethasone (Vetacort; Vétoquinol, Barcarena, Portugal), 1 mg/10 kg body weight (BW), IM, and dexcloprostenol (Gestavet-Prost^®; Hipra, Lisbon, Portugal), 125 μL, IM), (n = 5 does) or a cesarean section was performed (n = 3 does).

Each HP (control) goat was picked from the same pen from which its cohort PT goat came, on the same day that the PT case was enrolled in the study, and selected to mirror, as closely as possible, its cohort PT goat with respect to age, body condition score (BCS), body weight, and breed (in that approximate order of importance).

The 5 criteria used in selection of the HL does are: > 100 days in milk (DIM), milk production > 1.5 L/d, BCS, age, and breed. Most of the HL does were thin, so BCS were weighted more heavily in the selection process than were ages or breeds. An attempt was made to select HL goats whose BCS most closely approximated those of the PT goats.

At the time that each doe was being selected and enrolled in this study, its weight and BCS were determined. A blood sample was obtained from the jugular vein of each of the 24 does, on the side opposite that subsequently used for glucose infusion. The following devices were used on-farm: a validated electronic on-farm test (Precision Xceed; Abbott, UK) to quantify BHBA (25) and a portable analyzer (i-Stat; Sensor Devices, Waukesha, Wisconsin, USA) to measure Na⁺, K⁺, Cl⁻, HCO₃⁻, glucose, pH, base excess, pCO₂, anion gap and urea nitrogen.

Subsequently, an 18-gauge 2-inch catheter (Introcan; W.B. Braun, Rohrdorf, Germany) was introduced into the appropriate jugular vein, and 500 mL of a 5% glucose solution was administered to each of the 24 experimental does, over an average time of 10 min (about 0.37g of glucose/kg BW). Blood samples were then collected from all PT, HP, and HL does at 30 and 90 min after completion of glucose infusion, from the same vein previously used for this purpose, and blood glucose values were determined. A third blood glucose value was determined at 180 min post-infusion, in does in which the 90-min blood glucose concentration exceeded that doe’s pre-infusion level.

In Experiment 2, the objective was to determine the extent to which the blood values for BHBA and pH in individual goats with PT might be correlated. The criteria for inclusion of PT goats in this experiment were that i) there was a compatible case history, ii) characteristic clinical signs of PT were present, and iii) blood concentrations of BHBA were ≥ 2.9 mmol/L.

Blood samples were obtained from the jugular vein of 17 additional PT does making a total of 25 does with clinical signs of PT in which blood BHBA values were determined with the electronic on-farm test (Precision Xceed) and blood pH values were measured with a portable analyzer (iStat). Of these 25 PT goats, 5 were Saanen, 9 were Alpine and 11 were Saanen-Alpine cross-breds.

Statistical analysis

Data from the three 8-goat groups were analyzed with statistical software (IBM SPSS Statistics V.19, Armonk, New York, USA) for descriptive statistics and for hypothesis testing. Due to the small sample size, the non-parametric Mann Whitney U-Test (26) was used to compare the distributions of 2 independent samples, so as to obtain a 95% confidence interval level. Pearson’s correlation (26) was used to investigate the association between the blood values for pH and BHBA in the 25 PT does.

Results

The ages, body condition scores, and body weights for all 3 groups of experimental does are summarized in Table 1. In addition, the following information was collected and recorded for only 1 or 2 of the 3 8-doe GTT groups: i) days in milk (in HL does only); ii) daily milk production (in HL does only); and iii) number of kids carried/delivered (in PT and HP does only).

Table 1.

Ages, body condition scores (BCS), body weights, and days in milk (DIM), daily milk production, and number of fetuses in pregnancy toxemia (PT) does, healthy pregnant non-lactating (HP) does and healthy lactating non-pregnant (HL) does

Parameter	8 PT Does median (range)	8 HP Does median (range)	8 HL Does median (range)
Age (years)	5.5 (2 to 8)	4.0 (2 to 6)	5.0 (2 to 6)
BCS	3.0 (3.0 to 3.5)	3.5 (2.5 to 5)	3.3 (2 to 4)
Body weight (kg)	68.5 (51 to 77)	64.5 (55 to 70)	60.0 (51 to 71)
DIM	NA	NA	132 (109 to 144)
Milk production (L)	NA	NA	2.1 (1.7 to 2.7)
Number of fetuses	2.5 (2 to 3)	2.0 (1 to 3)	NA

BCS — body condition score; NA — not applicable; DIM — days in milk.

Blood chemistry results for PT, HP, and HL does are summarized in Table 2 (27,28). Blood values were significantly lower in PT does than in HP does for K⁺, pH, HCO₃⁻, base excess, pCO₂ (P < 0.001) and glucose (P < 0.005). Blood values were significantly higher in PT does than in HP does for anion gap (P < 0.005), and BHBA (P < 0.001).

Table 2.

Blood chemistry values from 8 pregnancy toxemia (PT) does, healthy pregnant non-lactating (HP) does and healthy lactating, non-pregnant (HL) does

Parameter	8 PT Does median (range)	8 HP Does median (range)	8 HL Does median (range)	Reference range
Na+ (mmol/L)	141.5a,b (129 to144)	143.0a (139 to 147)	138.5b (132 to 143)	142 to 155*
K+ (mmol/L)	3.0a (2.1 to 3.7)	3.9b (3.7 to 4.3)	3.7b (3.4 to 4.2)	3.5 to 6.7*
Cl− (mmol/L)	111.5a,b (100 to 121)	109.5b (100 to 112)	103.5a (100 to 108)	99 to 110*
Glucose (mmol/L)	1.6a (1.1 to 2.7)	2.8b (2.0 to 3.6)	2.8b (2.3 to 3.3)	50 to 75*
pH	7.18a (6.95 to 7.33)	7.40b (7.36 to 7.53)	7.39b (7.36 to 7.42)	7.32 to 7.5*
HCO3− (mmol/L)	9.0a (4 to 15)	23.7b (20 to 30)	24.2b (20 to 29)	20 to 29**
Base excess (mmol/L)	−19.5a (−28 to −11)	−0.6b (−6 to +5)	0b (−5 to +5)	−5 to +4**
Anion gap (mmol/L)	21.5a (17 to 26)	14.5b (13 to 19)	13.5 ± 2.3b (11 to 17)	12 to 24**
pCO2 (mmHg)	23.5a (17 to 29)	36.7b (28 to 52)	41.7b (33 to 45)	38 to 45*
BUN (mmol/L)	5.2a,b (3.6 to 8.7)	5.3a (3.2 to 7.1)	7.9b (6.1 to 10.4)	10 to 20*
BHBA (mmol/L)	6.8a (4.3 to 8.0)	0.3b (0.1 to 0.7)	0.4b (0.2 to 0.5)	< 1*

^*Christian and Pugh (27).

^**Stevens et al (28).

^a,bWithin a row, values with different superscript letters are significantly different among groups.

PT ≠ HL, P < 0.001 (Glucose, pH, HCO₃⁻, pCO₂, BE, Anion gap, BHBA), P < 0.005 (K⁺), P < 0. 05 (Cl⁻).

PT ≠ HP, P < 0.001 (K⁺, pH, HCO₃⁻, pCO₂, BE, BHBA), P < 0.005 (Glucose, Anion gap).

HP ≠ HL, P < 0.05 (BUN, Na⁺ and Cl⁻).

BUN — blood urea nitrogen; BHBA — beta hydroxybutyric acid.

Blood values were significantly lower in PT does than in HL does for K⁺ (P < 0.005), and for glucose, pH, HCO₃⁻, base excess, and pCO₂ (P < 0.001). Blood values were significantly higher in PT does than in HL does for Cl⁻ (P < 0.05), anion gap and BHBA (P < 0.001).

When the blood values of the HP and HL groups were compared, Na⁺ and Cl⁻ were significantly higher in the HP group (P < 0.05) and urea nitrogen was significantly higher in the HL group (P < 0.05).

The results of glucose tolerance tests are shown (Figure 1). Blood glucose values were significantly lower in the PT group than in the HP and HL groups, before glucose infusion was begun (P < 0.01). Compared with pre-infusion values, at 30 min post-infusion, blood glucose values were significantly higher in all 3 groups (P < 0.05), but glucose values were not significantly different between the 3 groups. At 90 min post-infusion, the blood glucose values in the HP and HL groups were significantly lower than in the PT group (P < 0.001 and P < 0.05, respectively) and blood glucose values in the HL group were similar to pre-infusion levels. In contrast, values in both the PT group and the HP group were significantly higher than pre-infusion levels (P < 0.05). At 180 min post-infusion, the blood glucose values were not significantly different between the PT and HP groups, but PT group blood glucose values were still significantly higher than the pre-infusion values (P < 0.05), which was clearly indicative of glucose intolerance.

Figure 1.

Blood glucose concentrations (mmol/L) in 8 does with pregnancy toxemia (PT), 8 healthy pregnant non-lactating (HP) does, and 8 healthy lactating non-pregnant (HL) does, before and after IV administration of 0.5 L of a 5% glucose solution.

* Differences between groups P < 0.01 (0 min).

^†Differences between groups P < 0.05 (90 min, PT versus HP), P < 0.001 (90 min, PT versus HL).

^‡Differences within the groups compared with the baseline P < 0.05 (0 min).

Blood pH values were compared with the blood BHBA values in 25 does with clinical signs of pregnancy toxemia (Figure 2). The Pearson’s correlation between these 2 variables was not statistically significant (P = 0.161). The blood BHBA values from all 25 individual PT cases ranged between 2.9 to 8.0 mmol/L (Figure 2).

Figure 2.

Blood pH and BHBA values in 25 does with clinical signs of pregnancy toxemia (PT) and beta hydroxybutyric acid (BHBA) values > 2.9 mmol/L.

Discussion

Glucose tolerance tests evaluate the body’s ability to normalize a glucose load, whether administered by the IV or oral route (5,6). The results of this study indicate that PT does do not respond well to a glucose challenge, even though their pre-infusion blood values were lower than those of the HP (P < 0.005) or HL does (P < 0.001). The response of the HP does to the glucose challenge was intermediate between the PT and HL groups.

Healthy pregnant ruminants are insulin resistant at the end of gestation and in early lactation (6,29,30). These homeorhetic adaptations are necessary to ensure an adequate supply of glucose for the gravid uterus and lactating mammary gland, in support of the growing offspring, both prenatally and postnatally (6,29). Reduced insulin sensitivity of peripheral tissues during late gestation assures adequate transfer of glucose from dam to fetus. If insulin-stimulated glucose utilization by insulin-sensitive tissues is not limited, the fetus might not survive due to fetal hypoglycemia (30). A decline in blood insulin concentration occurs in ewes in the early stages of pregnancy and it has been suggested that this represents a homeorhetic control mechanism for sparing glucose for the fetal brains and fetoplacental units of the dams (31). Insulin concentrations in maternal blood decrease considerably as the number of fetuses increases (31). Increases in plasma concentrations of BHBA and non-esterified fatty acids were observed in normal ewes, as the number of fetuses increased, and also as lambing approached (31). In female goats in early lactation, glucose utilization appears to become less responsive to administration of insulin (and, presumably, to the effects of endogenously produced insulin). These adaptations to the usual effects of insulin during early lactation were not maintained during mid-lactation (32). Some factors associated with insulin resistance in human females may mirror some of those involved in the development of ruminant hepatic lipidosis and ketosis. These factors include advancing pregnancy, obesity, hyperinsulinemia, fat feeding, hyperlipidemia, malnutrition, and other hormones (growth hormone and thyroxine) (30).

To our knowledge there is no prior study showing that obese goats with PT have reduced tolerance to administered glucose. In fact, the IV infusion of glucose to ewes and does affected with PT has been recommended as a therapeutic strategy by several authors (1,2,18,33). Others, however, have questioned its efficacy in both ewes and does with PT, because the survival rate in both ewes (14) and does (3) with PT was independent of the plasma concentration of glucose, and because the survival rate was poor in PT does (3 of a total of 22 cases) even after removal of the fetuses (by cesarian section or by pharmaceutical induction of parturition) and treatment with several medical regimens that provided supplemental glucose (3). Furthermore, administration of glucose to PT ewes was followed by a decrease in blood glucose values to below pre-treatment values by 6 h post-treatment, which may have resulted from a depressing effect of glucose infusion on hepatic glucose production and release into the blood stream (14).

Because the rate of occurrence of PT was essentially the same in Saanen and Alpine does in this herd, the age of the HL does was given greater value in the selection process than was breed. An effort was made to mirror the breeds represented in the PT group, but not at the expense of BCS and age.

The present study suggests that failure of glucose therapy, at least in obese PT does, might be related to the presence of marked hepatic lipidosis in such cases, as has been shown in humans (22) and dairy cows (9–13). The apparent inability of the livers in such cases to process glucose and avoid prolonged hyperglycemia might make the use of glucose therapy in such cases unhelpful, or even detrimental.

Malnutrition and/or feed restriction reduces the gluco-regulatory actions of insulin (30). In sheep, the effect of fasting (12 and 24 h) on insulin response was more marked in obese than in lean individuals (16), but in dairy cows, the magnitude and duration of malnutrition required to develop lipid-related metabolic disorders are largely unknown (30). We did not attempt to determine the impact that decline in food intake by the PT does in our study might have had on the response to the IV glucose load, because the dry does were fed as a group, and it was not possible to determine the feed consumption of individual does. Obese pregnant does in the last month of gestation and not eating would be at increased risk for developing hyperketonemia and hepatic lipidosis, because they would be aggravating their energy deficit and mobilizing body fat reserves (2).

Acidosis occurs frequently in ovine pregnancy toxemia (18). According to Marteniuk and Herdt (18), ketone bodies are metabolic acids, and the development of acidosis (ketoacidosis) in association with hyperketonemia is common in most species. However, in this study of 25 goats with pregnancy toxemia (all having the characteristic clinical signs and blood BHBA values ≥ 2.9 mmol/L) the Pearson’s correlation was non-significant, which means that there are probably other factors contributing to the severity of the metabolic acidosis besides BHBA (Figure 2). In humans, a moderate degree of lactic acidosis may be seen in some patients with diabetic ketoacidosis (34). How this occurs is not clear, although marked hypovolemia is likely to play an important role (34). Further studies are necessary to clarify this finding.

In summary, the results of this study clearly demonstrate that PT goats are glucose intolerant. There was little correlation between blood BHBA and pH. In our experience, blood pH values are a fairly reliable indicator of which PT goats are likely to survive and which are certain to die (3).