Triglyceride response following an oral fat tolerance test in Burmese cats, other pedigree cats and domestic crossbred cats

Abstract

Primary lipid disorders causing fasting triglyceridaemia have been documented infrequently in Burmese cats. Due to the known increased risk of diabetes mellitus and sporadic reports of lipid aqueous in this breed, the aim of this study was to determine whether healthy Burmese cats displayed a more pronounced pre- or post-prandial triglyceridaemia compared to other cats. Serum triglyceride (TG) concentrations were determined at baseline and variably at 2, 4 and 6 h after ingestion of a high-fat meal (ie, an oral fat tolerance test) in a representative sample of Burmese and non-Burmese cats. The median 4 and 6 h serum TG concentrations were significantly higher in Burmese cats (4 h–2.8 mmol/l; 6 h–8.2 mmol/l) than in other pedigree and domestic crossbred cats (4 h–1.5 mmol/l; 6 h–1.0 mmol/l). The non-Burmese group had post-prandial TG concentrations ranging from 0.6 to 3.9 mmol/l. Seven Burmese cats had post-prandial TG concentrations between 6.6 and 19.0 mmol/l, five had concentrations between 4.2 and 4.7 mmol/l, while the remaining 15 had post-prandial concentrations between 0.5 and 2.8 mmol/l. None of these Burmese cats had fasting triglyceridaemia. Most Burmese cats with a 4 h TG > 6.0 mmol/l had elevated fasting very low density lipoprotein (VLDL) concentrations. This study demonstrates that a proportion of Burmese cats in Australia have delayed TG clearance compared to other cats. The potential repercussions of this observation with reference to lipid aqueous, pancreatitis and diabetes mellitus in Burmese cats are discussed.

Lipaemia in cats is attributed to an increase in triglyceride (TG) concentrations in plasma. As chylomicrons (CM) and VLDL transport the majority of exogenous and endogenous TGs, respectively, hypertriglyceridaemia results from an increase in one or both of these lipoproteins in the circulation. Lipoprotein lipase (LPL) is involved in reducing the circulating TG concentration by hydrolysing TG from CM and VLDL to produce non-esterified fatty acids and glycerol. ¹ This process is activated by the co-factor apolipoprotein C2 (apoCII) and modulated by apolipoprotein E (apoE). ² Synthesis of LPL occurs mainly in adipose tissue, cardiac and skeletal muscle and it later attaches to the luminal surface of capillary endothelial cells via chains of heparin sulphate-proteoglycans. ¹ Heparin administration is known to decrease LPL binding from these sites, resulting in an increased circulating LPL concentration which accelerates the rate of plasma TG clearance. ³

Over 60% of the variability in fasting serum lipid concentrations in humans is due to the effects of a variety of genes. ⁴ Environmental factors such as dietary composition and obesity also influence fasting lipid concentrations. ⁴ In humans, inherited primary disorders resulting in hypertriglyceridaemia include deficiencies in LPL, hepatic lipase and apoCII. Varying degrees of fasting hypertriglyceridaemia and/or hypercholesterolaemia are associated with these genetic disorders, although heterozygous carriers with a LPL deficiency may only show signs of disease when secondary factors such as obesity are present concurrently. ⁵

Primary lipid disorders are not commonly observed in cats. Lipid disorders secondary to diabetes mellitus, pancreatitis, hyperadrenocorticism and administration of corticosteroids or progestagens are more likely to account for fasting lipaemia in this species. ^6,7 The best characterised primary lipid disorder in cats is inherited fasting hyperchylomicronaemia, an autosomal recessive disorder resulting from reduced LPL activity due to a point mutation in exon 8 of the LPL gene. ^8,9 Cats homozygous for this condition develop severe fasting and post-prandial hypertriglyceridaemia, comprising a marked increase in CM and a mild to moderate increase in VLDL. Heterozygous cats have normal fasting TG concentrations but prolonged post-prandial lipaemia after an oral fat challenge. ¹⁰ Persistently elevated lipid concentrations in homozygous cats result in the development of one or more of the following: lipaemia retinalis, peripheral neuropathy, cutaneous xanthomatosis and, less commonly, anaemia. ^11–13 Abdominal pain and pancreatitis have not been reported as features of hyperchylomicronaemia in these cats, in contrast to similarly affected canine or human patients. ^14–16 These clinical manifestations can be successfully managed by feeding a low fat diet.

Another manifestation of lipid disorders in cats is lipid aqueous, a sporadic condition whereby lipid accumulates in the aqueous humour of the eye. This condition is thought to result from a transient breakdown in the blood-aqueous barrier, possibly due to primary anterior uveitis, in the setting of concurrently increased plasma chylomicron and/or VLDL concentrations. ¹⁷ Large lipoproteins ‘leak’ across this barrier, causing a previously clear anterior chamber to become hazy or even opaque, ¹⁸ as illustrated in Fig 1. In preliminary reports, Hardman and collaborators observed a primary lipid disorder in otherwise healthy adolescent Burmese cats. ^19,20 These patients had initially presented with one or more episodes of lipid aqueous, with or without concurrent uveitis, and mild to moderately elevated fasting TG referable to increased VLDL lipoproteins. Although post-heparin LPL activity was not measured, plasma TG concentrations were unchanged after intravenous (IV) administration of heparin in three cats (Hardman, unpublished data). A similar condition has also been described in young Burmese cats in the United Kingdom (UK). ^17,21

Fig 1.

Appearance of two young Burmese cats with lipid aqueous; note the lipid accumulation in the anterior chamber (A) in the right eye of a lilac Burmese cat (photo courtesy of Dr C Hardman) (B) and in the left eye of a brown Burmese cat (reproduced with permission from Hardman and Stanley. ¹⁹

In Australia and the UK, there is an over-representation of diabetic Burmese cats, ^22–25 with affected cats often having a greater requirement for exogenous insulin than domestic crossbred diabetic cats (DB Church and R Malik unpublished observation). It is possible that a defect in lipid metabolism may be linked to insulin resistance in this sub-group of Burmese cats and that these cats may be at risk for the development of lipid aqueous well before they develop diabetes.

On the basis of the existing knowledge on lipid disorders in cats and other species, the hypothesis for this study was that some Burmese cats have reduced TG clearance from plasma which may explain the aetiology of lipid aqueous. As measuring fasting TG concentrations has been demonstrated to be an insensitive method for detecting subtle or incomplete defects in lipid metabolism, ⁵ a modified oral fat tolerance test (OFTT) was utilised in this study.

Methods

Cats

Healthy cats (Burmese, other pedigree and domestic crossbred) were selected at random from three catteries (n=25 cats) and four veterinary hospitals (n=43 cats) in the Sydney metropolitan region. On all but two occasions, it was possible to collect blood specimens from Burmese cats and non-Burmese cats sharing a common diet, environment and likely having similar activity levels, ie, cats living in the same household or cattery. Cats in both groups displayed a wide range of ages, weights and body condition scores (BCSs) as might be expected from the broad inclusion criteria. BCS was determined using a nine point scale (1–3=thin, 4–5=ideal, and 6–9=overweight). ²⁶ Recent dietary history, in particular diet composition, consistency (dry/canned/fresh meat) and frequency of feeding (once or twice daily, or ad-libitum) were recorded. All cats were deemed healthy based on prior medical history and routine physical examination.

OFTT

Each cat was fasted for at least 15 h overnight prior to blood collection, at which time the baseline TG concentration was determined (t=0 h). For a small number of cats (n=4), a baseline sample was not collected as this was not feasible due to the patients' non-compliance. Each cat received 1–1.5 gfat/kg body weight using a standardised prescription canned diet formulated for pets recovering from major surgery or illness (Hill's Prescription Diet canine/feline a/d). This diet consisted of 30.4% fat, 44.2% protein and 15.4% carbohydrate on a dry-matter basis. This meal was selected in part because it is widely available worldwide and, therefore, suitable for inclusion in a standardised test protocol. The test meal was either eaten voluntarily (33 cats) or gently force-fed (35 cats) over a 10–15 min period. Blood was collected for determination of TG concentration at 2, 4 and 6 h post-prandially. Due to a variety of practical and technical limitations, not all cats had blood collected at each time point.

Blood collection and handling

Blood was collected from the external jugular vein or cephalic vein using a 23 gauge needle and syringe or a 23 gauge butterfly needle and syringe, respectively. Blood was placed in plain serum and/or ethylene-diamine-tetra-acetic acid (EDTA) tubes. Following centrifugation for 10 min at 2500×g, serum or plasma was inspected for the presence of lipaemia (Fig 2), separated and processed within 12 h of collection. Serum TG and cholesterol concentrations were determined and the remaining serum stored at 4°C overnight to determine the presence or absence of CM. Fasting and 4 h post-prandial serum TG concentrations were determined in 61 and 51 cats, respectively. Fasting serum cholesterol was determined in 48 cats. Fasting and 4 h post-prandial serum glucose concentrations were determined in 30 cats; serum was separated within 1 h of blood collection. Point-of-care devices to measure TG (Accutrend GCT, Roche Diagnostics, Auckland, New Zealand and PTS. Cardiochek, Pursuit Performance, Adelaide, SA, Australia) ²⁷ and glucose concentrations in whole blood were used concurrently. Lipoprotein electrophoresis was performed on a number of representative fasting and 4 h post-prandial plasma or serum samples from both Burmese (n=12) and non-Burmese cats (n=9).

Fig 2.

Appearance of t=4 h serum after centrifugation. Serum on the left is from a domestic crossbred cat (TG 1.5 mmol/l). Serum on the right is from an ‘affected’ Burmese cat (TG 9.3 mmol/l).

Analytical methods

Serum TG and cholesterol concentrations were performed by a commercial NATA-accredited laboratory (Symbion Vetnostics Laboratory, North Ryde, NSW, Australia). The measurement of all serum analytes was performed on an automated Hitachi-Roche analyser using standard methods. Serum TG and cholesterol concentrations were determined using glycerol-phosphate oxidase and peroxidase (GPO-PAP) and cholesterol esterase, cholesterol oxidase and peroxidase (CHOD-PAP) test kits, respectively (Boehringer Mannheim). Serum glucose concentrations were determined using a hexokinase method with a Gluco-quant enzyme kit (Roche Diagnostics/Hitachi). Both point-of-care devices used to measure TG used the GPO-PAP method and the blood glucose device utilised the hexokinase method incorporated into reagent strips.

Lipoprotein electrophoresis was performed by the Protein and Lipid Department, Royal Prince Alfred Hospital, Sydney. Lipoproteins were separated using the CIBA Corning ACI Electrophoresis System using standard methods. Briefly, 5 μl of plasma or serum was applied to a 0.6% agarose gel. Electrophoresis was performed for 45 min in a Tris–barbital buffer (pH 8.6) at 100 V. Gels were stained with fat red 7B (Helena Laboratories). Completed gels were analysed quantitatively using a Beckman scanning densitometer, results of which were recorded as a percentage.

Statistical analysis

Serum TG and cholesterol concentrations were compared at each time point using Kruskall–Wallis analyses for four groups: (1) domestic crossbred cats, (2) other pedigree cats (3) Burmese cats and (4) pooled groups 1 and 2. Independent effects of age, gender and diet on serum TG concentrations were determined by analysis of co-variance. Correlations were determined using Spearman's correlation coefficient. Significant associations between categorical variables were determined using Fisher's exact tests. Densitometer percentage readings were compared using a two sample t-test. Values are expressed as mean±standard deviation (SD) and/or median and interquartile range (IQR). P values of <0.05 were considered significant. Reference intervals for fasting and 4 h serum TG concentration were created using mean±2SD for non-Burmese cats. All data analyses were performed using a statistical software package (Minitab 15.1.1, State College, PA, USA).

Results

Cats

Group 1

A total of 17 crossbred cats were recruited, comprising 14 domestic shorthair, two domestic medium-hair cats and one domestic longhair cat.

Group 2

A total of 18 non-Burmese pedigree cats were recruited, consisting of the following breeds: Tonkinese (six cats – one of these cats was subsequently excluded as an outlier), British Shorthair (two), Siamese (two), Oriental (two), Devon Rex (two), Snowshoe (one), Abyssinian (one), Burmilla (one) and Russian Blue (one).

Group 3

Of the 33 Burmese cats, the distribution of coat colour was as follows: brown (11), chocolate (six), lilac (four), red (four), lilac tortoiseshell (four), cream (two) and blue (two).

Age, weight and gender distributions for each group are presented in Table 1. Although there were more female cats in the Burmese group due to low numbers of male cats at catteries, this was not significantly different from the non-Burmese cohort (P=0.1). There were no other obvious or statistically significant differences in signalment between the three groups.

Table 1.

Characteristics of Burmese, other pedigree and domestic crossbred cats used in this study, (median [IQR] shown)

	Domestic crossbred (group 1)	Other pedigree (group 2) ∥	Non-Burmese (groups 1 and 2 pooled)	Burmese (group 3)
Number	17	18	35	33
Age (years)	3.5 (1.5–7.0)	5.0 (2.0–7.0)	5.0 (1.8–7.0)	3.8 (1.0–8.6)
Gender	12 MN	10 M (7 N, 3 E)	22 M (19 N, 3 E)	14 M (8 N, 6 E)
	5 FN	7 F (4 N, 3 E)	12 F (9 N, 3 E)	19 F (13 N, 6 E)
Weight (kg)	5.1 (4.5–6.5)	4.3 (3.7–5.6)	4.8 (4.0–5.6)	4.0 (3.6–4.8)
Triglyceride † (mmol/l)
0 h #	0.4 (0.3–0.5)	0.4 (0.3–0.5)	0.4 (0.3–0.5)	0.4 (0.3–0.5)
2 h	1.3 (0.4–2.0)	1.4 (0.9–2.4)	1.3 (0.7–2.0)	1.9 (1.1–2.9)
4 h	1.6 (0.8–2.3)	1.4 (1.0–1.6)	1.5 (1.0–2.0)	2.8 * (1.5–7.0)
6 h	1.0 (0.7–1.3)	1.3 (0.7–1.6)	1.0 (0.7–1.4)	8.2 ** (2.6–14.6)
0 h cholesterol ‡ (mmol/l) #	3.5 (2.8–4.6)	4.0 (3.0–5.0)	3.7 (2.9–4.8)	3.7 (2.9–4.0)
0 h glucose § (mmol/l) #	4.4 (4.1–4.8)	5.1 (4.5–5.8)	4.7 (4.2–5.6)	4.7 (4.1–5.0)

^*P<0.05 (significantly different to all other groups);

^**P<0.01 (significantly different to all other groups). ME=male entire, MN=male neutered, FE=female entire, FN=female neutered. Values in bold indicates significantly different values (P< 0.05, P< 0.01, respectively) in Burmese compared to non-Burmese cats.

^#Reference ranges Symbion Vetnostics Laboratories: serum triglyceride 0.1–0.8 mmol/l, serum cholesterol 2.4–5.2 mmol/l, serum glucose 3.9–8.3 mmol/l.

^†Not all cats were tested at each time point.

^‡48 Cats tested.

^§30 Cats tested.

^∥One Tonkinese cat with marked post-prandial hypertriglyceridaemia was considered an outlier and excluded from statistical analyses.

OFTT

Groups 1 and 2

Changes in TG concentrations in domestic crossbred and non-Burmese pedigree cats during an OFTT were similar (Table 1). As differences between the groups were not statistically significant, results for groups 1 and 2 cats were pooled, as illustrated in Table 1 and Figs 3 and 4.

Fig 3.

Serum TG concentrations in all cats, comparing Burmese and non-Burmese cats at 0, 2, 4 and 6 h post-oral fat challenge.

Fig 4.

Serum TG concentrations at t=4 h during an OFTT in Burmese (n=25) and non-Burmese cats (n=26). Y-axis reference line at 3.2 mmol/l demonstrates upper reference interval for 4 h post-prandial TG peak. Age groupings are adolescent (<1 year), adult (1–6 years) and senior (≥7 years).

The TG concentration peaked at 4 h in all non-Burmese cats in which serial blood specimens were collected at 0, 2, 4 and 6 h (n=7 cats) and the mean, median and IQR for serum TG concentrations at t=4 h were 1.6, 1.5 and 1.0–2.0 mmol/l, respectively. Based on the mean±2SD from cats in groups 1 and 2, the reference interval for 4 h post-prandial serum TG was 0.1–3.2 mmol/l; therefore, TG concentrations in excess of 3.2 mmol/l were considered to be elevated. Two Tonkinese and one domestic shorthair cat had 4 h TG concentrations above this cut-off value (3.4, 6.4 and 3.9 mmol/l, respectively). The Tonkinese cat with a post-prandial serum TG concentration of 6.4 mmol/l was excluded as an outlier. Serum TG concentrations returned to baseline or were reduced substantially 6 h after eating.

Group 3

For 25 Burmese cats tested at 4 h the mean, median and IQR for TG were 4.2, 2.8 and 1.5–7.0 mmol/l; significantly higher than non-Burmese cats (in groups 1 and 2 combined; P=0.05). Of these Burmese cats, 12/25 (48%) had 4 h post-prandial TG concentrations above the reference interval of 3.2 mmol/l; seven of these 12 cats (28%) had markedly elevated TG concentrations ranging from 6.6 to 11.5 mmol/l, while the remaining five (20%) had moderately elevated concentrations of 4.2–4.7 mmol/l (Fig 3). In those cats tested at 6 h (n=5), in three the serum TG concentrations were lower at 6 h (2.1, 3.0 and 8.2 mmol/l) than 4 h post-prandially (4.3, 4.3 and 9.5 mmol/l), while in the other two, serum TG was higher at 6 h (10.1 and 19.0 mmol/l) than at 4 h (8.2 and 11.5 mmol/l).

The 12 cats with moderate (>4.0 mmol/l) and marked (>6.0 mmol/l) 4 h post-prandial hypertriglyceridaemia were compared to the remaining 13 cats (‘non-affected’) in group 3. The mean±SD ages and weights of ‘affected’ cats (5.5±3.5 years and 4.5±0.6 kg) were not significantly different compared to ‘unaffected’ Burmese cats (4.2±4.6 years and 3.7±0.8 kg). Two of the ‘affected’ Burmese cats were <1 year old, six were between 1 and 6 years-of-age and four were older than 7 years. There was no significant difference in post-prandial TG concentrations between young and older cats, as illustrated in Fig 4. Eight of the 12 ‘affected’ cats had a BCS of 4 or 5, the remaining four cats were moderately overweight (BCS of 6 or 7). Only two of the 13 ‘non-affected’ Burmese had a BCS of 6; the rest had a BCS of 4 or 5. Differences in BCS between ‘affected’ and ‘non-affected’ Burmese were not significant (P=0.4). Of the 12 ‘affected’ Burmese cats, males and females were equally represented, but females accounted for eight of the 13 ‘non-affected’ Burmese cats.

Dietary comparisons between cats

Five of the seven Burmese cats with marked post-prandial hypertriglyceridaemia shared the same environment and had been given the same diet as individual cats included in the non-Burmese cats in groups 1 and 2. An additional Burmese cat within this sub-group shared the same diet and environment as another Burmese cat with a ‘normal’ OFTT response.

Of the 25 Burmese cats tested at t=4 h, only 14 cats ate the test diet willingly; interestingly nine of these cats were ‘affected’ Burmese that developed moderate to marked post-prandial TG elevations. This relationship, however, was not statistically significant (P=0.4).

Fasting lipid and glucose concentrations

Fasting serum TG and cholesterol concentrations for groups 1, 2 and 3 are illustrated in Table 1. There was no significant difference in median fasting TG concentrations between Burmese, other pedigree and domestic crossbred cats. The median fasting serum TG concentration was higher in ‘affected’ Burmese cats (0.5 mmol/l) than ‘non-affected’ Burmese cats (0.3 mmol/l) and non-Burmese cats (0.4 mmol/l), although this was not statistically significant (P=0.5). Two cats, both domestic shorthairs, had fasting TG concentrations above the reference interval (0.9 mmol/l), however, their 4 h TG concentrations were below the 4 h reference cut-off of 3.2 mmol/l. All three groups had similar fasting serum cholesterol and glucose concentrations. Both Burmese and non-Burmese had similar or lower serum glucose concentrations 4 h post-prandially compared to baseline. No correlation was found between weight, age, gender or diet and fasting serum TG or cholesterol concentrations.

Lipid electrophoresis

Groups 1 and 2

Lipoprotein electrophoresis using serum or plasma collected at baseline and 4 h post-prandially was performed in nine ‘normal’ cats. Two cats had a faint chylomicron band despite a normal fasting TG concentration. A 10-month-old, female neutered domestic shorthair cat had a more prominent fasting and VLDL band compared to the rest of those tested. This cat had a 4 h post-prandial TG concentration of 3.9 mmol/l, the highest recorded TG concentration in the non-Burmese group.

Group 3

Lipoprotein electrophoresis was performed at 0 h (12 cats) and 4 h post-prandially (13 cats). Of these, 10 of the 12 ‘affected’ Burmese cats were included. None had chylomicron bands present at 0 h. Based on densitometer readings, Burmese cats had more prominent VLDL bands and less prominent high density lipoprotein (HDL) bands compared to groups 1 and 2 cats at 0 h (P=0.04 and 0.06, respectively) but not at 4 h (P=0.6 and 0.3, respectively). Although low density lipoprotein (LDL) bands were more prominent in Burmese cats at both 0 and 4 h, the difference was not significant (P=0.5 and 0.1, respectively). Post-prandial chylomicron band staining reflected the degree of triglyceridaemia.

Six of the seven Burmese cats with marked post-prandial hypertriglyceridaemia (TG>6.0 mmol/l) had 0 h plasma samples tested; four had prominent VLDL bands, one had a more prominent LDL band and one cat had similar LDL and VLDL staining. These cats also had more prominent VLDL bands and sometimes LDL bands at t=4 h compared to non-Burmese cats. An example, comparing an ‘affected’ Burmese to a normal domestic crossbred, is illustrated in Fig 5.

Fig 5.

Plasma lipoprotein electrophoretogram showing a t=0 h and 4 h sample from a normal domestic crossbred cat (top two rows) and an ‘affected’ Burmese cat (bottom two rows). Note the difference in staining of the VLDL bands (arrows) at t=0 h in both cats. Also note the marked trailing due to excessive CM in the affected Burmese cat, as marked with a star.

Discussion

The most important findings of this study were that seven of 25 Burmese and one Tonkinese cat had marked post-prandial hypertriglyceridaemia (6.6–19.0 mmol/l) 4–6 h after an oral fat challenge. A further five Burmese cats had moderately elevated post-prandial TG concentrations (4.2–4.7 mmol/l), compared to the remaining 13 Burmese cats (0.5–2.8 mmol/l). All other pedigree and domestic crossbred cats had post-prandial TG concentrations between 0.6 and 3.9 mmol/l. Of the 51 cats who had blood collected 4 h post-prandially, 24 cats had a TG concentration of ≤1.6 mmol/l (seven Burmese, 17 non-Burmese), 15 cats had a TG concentration of 1.7–4.0 mmol/l (six Burmese, nine non-Burmese) and 12 cats had a TG concentration >4.0 mmol/l (12 Burmese). This shows that a sub-group of at least 12 Burmese cats had a reduced ability to clear TGs following a high-fat meal. This was thought to most likely reflect an inborn error of lipid metabolism, as unaffected cats of similar BCS and age, sharing the same environment and lifestyle, had normal TG clearance. Most, but not all, Burmese cats with a marked post-prandial TG response had more prominent VLDL bands in fasting serum samples compared to ‘normal’ cats. Importantly, delayed TG clearance occurred mainly in adult Burmese cats over 3 years-of-age.

Crispin ¹⁷ briefly mentioned a presumptive familial hypertriglyceridaemia in Burmese cats, in which plasma TG concentrations were twice that of age-matched controls. Although this was attributed to an elevation in both CM and VLDL concentrations, methodological details were lacking. In Hardman's studies, Burmese cats that developed lipid aqueous with concurrent fasting hypertriglyceridaemia had elevated fasting VLDL concentrations with no CM on electrophoresis. ²⁰ Apart from ocular signs, these cats were clinically healthy with unremarkable biochemical analytes. Furthermore, their clinical signs were transient, often resolving spontaneously or following topical treatment for uveitis, with or without instituting a low fat diet. As all these patients were ‘adolescent’ cats and some cats had common ancestry, an inherited familial disorder was thought most likely (C Hardman, unpublished data).

On average, serum TG concentrations peak 3–4 h after a meal, usually returning to baseline within 8 h in healthy feline, human and canine patients. ^10,28–30 In this study, a proportion of Burmese cats developed post-prandial hypertriglyceridaemia with their peak TG concentration delayed for up to 6 h or longer. Two Burmese cats with lipid aqueous and two matched normal Burmese cats in Hardman's study displayed a moderate post-prandial TG response at 2 h compared to domestic crossbred cats, suggesting this lipid disorder can also occur in Burmese cats without a history of lipid aqueous (C Hardman unpublished data), a finding expanded upon by the present work. No Burmese cat with moderate or marked post-prandial hypertriglyceridaemia in the current study had a history of lipid aqueous, although this may merely reflect an absence of an episode of uveitis required to ‘unmask’ any underlying lipid metabolism defect.

The Burmese condition identified in this study appears to be quite different to the fasting hypertriglyceridaemia of cats described by Jones et al, ⁸ where mild elevations in VLDL were accompanied by marked elevations of CM in fasting plasma. Marked post-prandial hypertriglyceridaemia has also been described in cats with the same inherited LPL deficiency. ¹⁰ Fasting TG concentrations in normal cats and those heterozygous for the LPL gene defect are indistinguishable, although oral fat loading reveals a higher, delayed TG peak (mean 2.35 mmol/l) 5 h after feeding in heterozygous cats, compared to a 1.1 mmol/l peak 3 h post-prandially in normal cats. ¹⁰ Cats homozygous for the defective LPL allele had elevated fasting TG concentrations (2.38 mmol/l) which increased fourfold post-prandially to 9.36 mmol/l, 7 h after feeding a high-fat meal. People who are heterozygous carriers for this condition do not invariably develop clinical signs, however, they may or may not develop fasting hypertriglyceridaemia depending on additional factors such as diabetes mellitus, obesity or hyperinsulinaemia. ^31,32 An OFTT is mandatory to distinguish these carriers from normal patients, describing the inability to clear post-prandial TGs as ‘impaired TG tolerance’. ⁵

The present study has uncovered a five- to 10-fold post-prandial elevation, and subsequent delayed clearance of serum TG from normal fasting concentrations, in a substantial proportion of apparently normal Burmese cats. It is possible that a proportion of Burmese cats in Australia are heterozygous for a defective LPL allele, whereby normal fasting TG concentrations are accompanied by an impaired TG tolerance after an oral fat challenge.

Lipid metabolism in cats is influenced by prior dietary history, ie, proportion of fat, protein and carbohydrate in the diet, as well as their adiposity and reproductive status. ³³ Prior to testing, all cats were fed varying types of commercial dry food and most were also given commercial canned food or fresh meat. As the cats in the current study did not consume a standardised diet for the period immediately prior to testing, it is possible that the observed TG response may in part be a reflection of their recent dietary history. However, five of the seven Burmese cats with marked post-prandial hypertriglyceridaemia were housed with non-Burmese cats, while an additional ‘affected’ Burmese cat was housed with an ‘unaffected’ Burmese cat; this provided a convenient and cogent ‘control’ for each affected Burmese as each ‘pair’ shared the same environment and consumed the same diet. Because of the effect of diet and other confounding variables, it is important to emphasise that the percentage of Burmese cats with impaired TG clearance may be an underestimate, as feeding a higher fat meal for a longer period prior to testing may help unmask this phenotype. Differences in post-prandial TG concentrations between studies may in part also reflect the composition of the test meal. An increase in fat and carbohydrate load during an OFTT in humans has been shown to accentuate post-prandial lipaemia, ³⁴ which may explain, in part, why non-affected cats in the current study had four-fold elevations in TG concentrations on average after feeding compared to previous studies demonstrating two-fold elevations in serum TG. ^10,28

In healthy individuals, post-prandial hepatic VLDL production is suppressed by high insulin concentrations, allowing LPL to hydrolyse chylomicron-TG. Insulin also acts directly on LPL, with a relative insulin deficiency contributing to reduced clearance of CM and VLDL from the circulation. Overweight, glucose intolerant or diabetic human patients with normal fasting serum TG concentrations may display prolonged post-prandial lipaemia due to a reduced insulin effect on LPL activity. ³⁰ Obesity in cats also produces insulin resistance and a subsequent increase in VLDL production. ³⁵ Although this may explain the post-prandial lipaemia demonstrated in some of the overweight cats in this study, eight ‘affected’ Burmese cats were of healthy BCS. A previous study has shown a trend towards Burmese cats over 6 years-of-age having glucose intolerance and higher insulin concentrations compared to age-matched non-Burmese cats. ³⁶ It is conceivable that Burmese cats in this study had glucose intolerance or higher fasting insulin concentrations. However, of the cats tested, fasting and post-prandial glucose concentrations were within the normal range and no significant difference was found between Burmese and non-Burmese cats. Age appears to have had no effect on fasting serum TG concentrations despite senior cats (over 7 years) having lower plasma LPL and hepatic lipase activities. ³⁷ A lowered enzyme activity with maturity does not explain why only four of the 12 ‘affected’ Burmese cats were over 7 years old.

Most cases of post-prandial hypertriglyceridaemia in people are multifactorial, as numerous factors influence TG clearance such as insulin concentration, LPL activity and apoE concentration. ³⁸ In the current study, 28% of healthy Burmese cats were found to develop marked (>6.0 mmol/l) post-prandial hypertriglyceridaemia after a high-fat meal. The significance of this finding currently remains unknown. It is possible that these cats have a lipid disorder that is similar to the group of cats with lipid aqueous. In the absence of uveitis, which results in increased permeability of the blood-aqueous barrier and, therefore, exposes the phenotype, the condition was only detected with provocative testing. Alternatively, lipid aqueous in young Burmese cats may be due in part to a transient suppression of LPL activity, a known effect of viral or protozoal infections in humans. ³⁹ This seems less likely, however, as many young non-Burmese cats develop infectious uveitis without the presence of lipid aqueous. As younger animals have greater permeability of iris micro-vasculature, ¹⁷ this may also help to explain why lipid aqueous occurred transiently in some Burmese cats with hypertriglyceridaemia without the presence of concurrent uveitis.

Further studies, such as pre- and post-heparin LPL activity, apoCII activity, LPL and apoE genotyping in both ‘affected’ and ‘non-affected’ age and diet-matched Burmese cats, will be required to determine whether the underlying lipid metabolism defect has a genetic basis, which seems likely. As a Tonkinese cat (Burmese/Siamese cross) was found to be ‘affected’, breeds originating from Burmese blood lines (eg, the Australian Mist) should be investigated also. Longitudinal studies of affected individuals will be required to ascertain whether these cats are at greater risk for developing diseases such as pancreatitis or diabetes mellitus. The use of an oral fat challenge may aid in detecting ‘at-risk’ cats at a younger age, well before adverse sequelae develop.