Carotenemia

Abstract We performed a family study to investigate the heritability of reduced serum retinol levels observed in type 1 diabetes cases. Diet and serum factors, including retinol, total carotene, malondialdehyde, and retinol binding protein levels, were measured in 11 multiple-case families. The mean serum retinol level of the diabetics (46 ug/dl) was significantly less than the mean serum retinol level of the nondiabetics (60.9 ug/dl). The level of retinol binding protein was also significantly lower in diabetics (6.2 mg/dl) than in nondiabetics (7.6 mg/dl). The serum values of retinol binding protein were closely related within families, including both diabetic and nondiabetic family members. A characteristic shared between diabetics and one-third of their family members was a low ratio of serum retinol to total carotene, suggesting a low conversion of dietary carotene into retinol. Analysis of food frequency reports showed no difference between dietary retinol or total carotene level in diabetics or their relatives. This study offers evidence that heritability and the reduced conversion of carotene may play a role in the level of serum retinol in type 1 diabetes cases.

KEY WORDS: RETINOL, RETINOL BINDING PROTEIN, CAROTENE, MALONDIALDEHYDE, INSULIN-DEPENDENT DIABETES MELLITUS, GENETICS, DIET

The destruction of the beta cells of the pancreas, which defines type 1 diabetes, is considered an autoimmune process that shows multifactorial inheritance (Davies et al. 1994; Cornall 1993). The destruction of the beta cells begins with their infiltration by macrophages, T lymphocytes, and B lymphocytes (Hanenberg et al. 1989; Ihm and Yoon 1990). We became interested in the relationship between type 1 diabetes and retinol for two reasons: (1) because retinol has been shown to stimulate the cytotoxic activity of macrophages (Turpin et al. 1990) and T cells (Friedman and Sklan 1989) and to regulate cell growth, differentiation, and cytokine production in B lymphocytes (Blomhoff et al. 1992) and (2) because retinol has been shown to be decreased in the sera of type 1 diabetics (Basu et al. 1989; Wako et al. 1986).

Vitamin A, a collective term for compounds with retinol activity, is an essential nutrient prevalent in fortified milk. It can be obtained by absorption of retinol from the diet or it can be produced from the conversion of beta carotene in the intestinal mucosa and liver. Steady-state levels of retinol are maintained in the blood by the release of stored retinol from the liver (Blomhoff et al. 1990). Most of the retinol is transported from the liver to the tissues by retinol binding protein (RBP) in a 1:1 ratio of retinol to RBP. A previous study performed with streptozotocin-induced diabetic rats suggested that retinol levels were low because of insufficient RBP (Basu et al. 1990).

No previous studies with type 1 patients have investigated RBP levels in families to determine whether there is a genetic component to the reduced level of expression. Using a family design, we evaluated the levels of serum retinol, RBP, and total carotene for each diabetic and first-degree family member. Dietary information was collected from all participants to rule out inadequate intake as the cause of the low retinol level. To exclude the possibility that oxidative stress incurred by diabetes was depressing the retinol level, we also included the levels of malondialdehyde and serum vitamin E in the analysis.

Methods and Materials

Study Population. The participants in this investigation were selected from multiple type 1 sibling case families who were part of the Familial Diabetes Complications Study. Type 1 diabetics were identified from the Children's Hospital of Pittsburgh (CHP) and Allegheny County registries. Eligibility criteria for the CHP Registry include diagnosis of type 1 diabetes at less than 17 years of age, on insulin therapy at the time of discharge, and diagnosed or seen at Children's Hospital within one year of diagnosis. Eligibility requirements for the Allegheny County Registry include diagnosis of type 1 diabetes at less than 20 years of age, on insulin therapy at the time of hospital discharge, and a resident of Allegheny County at the time of diagnosis.

The sample group for the current analysis consisted of 11 multiple-case families, including the proband, diabetic and nondiabetic siblings, and the parents. Analyses of serum factors were performed in two ways: as a paired design with diabetics and siblings matched for age and sex and as independent comparisons of diabetics and relatives, adjusted for age.

Fasting blood samples from each family member were drawn and tested for retinol, alpha-tocopherol, total carotene, malondialdehyde, and RBP levels. Food frequency questionnaires were collected from all the study participants (Willett et al. 1985). In addition, a history of smoking, alcohol use, vitamin supplements, and medications was obtained. Use of vitamins excluded the participant from all analyses. The use of hormonal medications excluded six of the participants from the study. There were no significant differences between diabetics and nondiabetic family members with regard to smoking, alcohol use, body weight, or ethnic group. The investigator did not know the diabetic status of the individual during the serum analyses.

Laboratory Methods. Serum retinol (vitamin A) and alpha-tocopherol (vitamin E) were measured by high-pressure liquid chromatography (HPLC) (Tietz 1986). Before extraction, 0.125% butylated hydroxytoluene (BHT) and 4,mu g/ml of retinyl acetate were added to each sample. The extraction of 200 mu l of serum was performed with heptane. The extracts were dried with nitrogen and resuspended in methanol before being applied to a 10-mu m C18 uBondapak column (3.9 X 30.0 cm) and eluted using a methanol:water (92:8) mobile phase. The flow rate was 0.7 ml/min for the first 20 min, then 1.0 ml/ min for the next 35 min. Absorbance was measured at 292 nm.

The carotenoids, including beta-carotene, were extracted from serum samples with ethanol and petroleum ether. The samples were protected from light, and the absorbance was read on a spectrophotometer at 450 nm within a halfhour of preparation (Tietz 1976).

Malondialdehyde (MDA), an oxidized lipid by-product, was detected in serum samples using thiobarbituric acid reactivity (Yagi 1976). Duplicate samples were first mixed with EtOH, 20% trichloroacetic acid, and thiobarbituric acid reagent and then heated in a water bath at 95 deg C. The samples were then extracted with butanol and centrifuged, and the supernatant was measured with a fluorometer at excitation 515 nm and emission 550 nm.

Serum RBP was measured by the single radial diffusion method (Mancini et al. 1965). The protein antigen solution was applied to a well containing monospecific antibodies to RBP. Antigen placed in the well diffused radially, producing a precipitin ring. The diameter of the ring was measured after 72 hr and was compared with known standards.

Statistical Analysis. Statistical analyses were performed using the sPss and BMDP statistical software packages. The dietary factors were analyzed using a Student's t test. The serum factors for diabetics and nondiabetic relatives were compared using analysis of covariance for the adjustment of age. The regression of RBP with vitamin A in the serum was performed using SPSS.

Results

Dietary Analyses. The average daily consumption of retinol and total carotene is shown in Table 1. No significant differences between diabetic and nondiabetic individuals were found with respect to average intake of retinol or carotene based on the food frequency questionnaire. However, the diabetics were found to eat significantly more protein than nondiabetic individuals.

Serum Factors and Diabetic Status. The mean serum retinol, RBP, alpha-tocopherol, total carotene, and MDA levels of diabetic and nondiabetic individuals are listed in Table 2. Serum retinol and RBP levels were significantly lower among the diabetics compared with their nondiabetic relatives. To correct for a possible lack of independence between observations of family members, similar analyses were performed comparing the diabetic proband with one nondiabetic sibling from each family using a paired t test (Table 3). Again, the diabetics had significantly lower serum retinol levels than their siblings. Because these results were the same as those reported in Table 2, we used the total cohort for the remaining analysis, adjusting for age.

The distribution of serum vitamin A level divided by serum total carotene level is shown in Figure 1 for diabetics and nondiabetic relatives. The greatest number of diabetics had low vitamin A/total carotene ratios, with decreasing proportions in the medium and high categories. The distribution of the relatives was more normal, although a greater number of relatives were in the low vitamin A/total carotene group than in the high ratio group.

Regression Analyses. The linear regression of RBP level, dietary retinol level, diabetic status, and age on serum retinol level showed that RBP level was independently related to retinol level (Table 4). RBP level was more closely related to serum retinol level in diabetic than in nondiabetic subjects. Age was a significant factor in nondiabetic relatives but was not predictive of serum retinol level in diabetics.

Next, the variation in RBP value by family, including the diabetic and nondiabetic sibs and parents, was examined. An analysis of covariance of RBP level was performed by family identification number, adjusting for age and sex (Table 5). The difference in RBP level between families was greater than the differences within families (p

Discussion

This study confirms that the retinol and RBP levels are significantly lower in patients with insulin-dependent diabetes. We have examined the metabolism of retinol in a family study and have shown that (1) RBP level is closely associated with serum retinol level and (2) genetic patterns exist in the serum level of RBP in the multiple-case families. The data also show a striking relationship between insulin-dependent diabetes and a low retinol/ carotene ratio. Carotenemia has been reported in patients with diabetes mellitus (Gouterman and Sibrack 1980). Thus the data presented here offer evidence that the retinol level is determined in part by the lower conversion rate of beta-carotene to retinol, within a range of RBP expression that is characteristic of the individual's genetic background.

Deficiencies in other nutrients, such as protein, iron, or alphatocopherol, can adversely affect retinol transport, storage, and utilization (Underwood 1984). The dietary questionnaires indicate, however, that these nutrients are in adequate supply in the type 1 diabetics we studied. The diabetics reported eating equivalent amounts of vitamin E and iron and significantly more protein than control subjects. These data, however, are averages of a self-reporting system that reflects dietary patterns rather than dietary intake. RBP has a short half-life, approximately 10 days. Validation studies, however, for vitamin A measured with food frequency versus food record showed reliable results (Willett et al. 1985).

The lack of a difference in serum alpha-tocopherol level between diabetics and their nondiabetic relatives excludes the possibility that the lower serum retinol level in diabetics is due to fat malabsorption or that the loss of vitamin E increases the demand for vitamin A (Herbert and Morgan 1953). The measurement of MDA also did not differ between diabetic and nondiabetic subjects, which does not support increased catabolism or stress from oxidant injury as a reason for the reduced retinol and RBP levels. No relationship was observed between vitamin A level and glycosylated hemoglobin. Therefore we do not believe that control of the blood sugar is related to the difference in retinol between diabetics and control subjects.

It is important to establish the cause of the decrease in serum retinol level observed in diabetics because many studies have shown that even marginal retinol deficiency is linked to an increased susceptibility to disease (Bloem et al. 1990; Arrieta et al. 1992). If the changes in retinol metabolism are heritable, then it may be possible to identify individuals with low retinol levels before the development of infectious disease.

1 Department of Epidemiology, University of Pittsburgh, Pittsburgh, PA 15260. 2 Department of Pathology, University of Pittsburgh, Pittsburgh, PA 15260.

Literature Cited

Arrieta, A.C., M. Zaleska, H.R. Stutman et al. 1992. Vitamin A levels in children with measles in Long Beach, California. J. Pediatr. 121:75-78.

Basu, T.K., J. Leichter, and J.H. McNeill. 1990. Plasma and liver vitamin A concentrations in streptozotocin diabetic rats. Nutr. Res. 10:421-427.

Basu, T.K., W.J. Tze, and J. Leichter. 1989. Serum vitamin A and retinol binding protein in patients with insulin-dependent diabetes mellitus. Am. J. Clin. Nutr. 50:329-331. Bloem, M.W., M. Wedel, R.J. Egger et al. 1990. Mild vitamin A deficiency and risk of respiratory tract diseases and diarrhea in preschool and school children in northeastern Thailand. Am. J. Epidemiol. 131:332-339.

Blomhoff, H.K., E.B. Smeland, B. Erikstein et al. 1992. Vitamin A is a key regulator for cell growth, cytokine production, and differentiation in normal B cells. J. Biol. Chem. 267:23,988-23,992.

Blomhoff, R., M. Green, T. Berg et al. 1990. Transport and storage of vitamin A. Science 250:399-404.

Cornall, R.J. 1993. Genetics of a multifactorial disease: Autoimmune type 1 diabetes mellitus. Clin. Sci. 84:257-262.

Davies, J.L., Y. Kawaguchi, S.T. Bennett et al. 1994. A genome-wide search for human type 1 diabetes susceptibility genes. Nature 371:130-136.

Friedman, A., and D. Sklan. 1989. Antigen-specific immune response impairment in the chick as influenced by dietary vitamin A. J. Nutr. 119:790-795. Gouterman, I.H., and L.A. Sibrack. 1980. Cutaneous manifestations of diabetes. Cutis 25:4554.

Hanenberg, H., V. Kolb-Bachofen, G. Kantwerk-Funke et al. 1989. Macrophage infiltration precedes and is a prerequisite for lymphocytic insulititis in pancreatic islets of prediabetic BB rats. Diabetologia 32:126-134.

Herbert, J.W., and A.F. Morgan. 1953. The influence of alpha-tocopherol upon the utilization of carotene and vitamin A. J. Nutr. 50:175-190.

Ihm, S.-H., and J.-W. Yoon. 1990. Studies on autoimmunity for initiation of B-cell destruction. Diabetes 39:1273-1278.

Mancini, G., A.O. Carbonara, and J.F. Heremans. 1965. Immunochemical quantitation of antigens by single radial inunodiffusion. Immunochemistry 2:235-254. Tietz, N. 1976. Fundamentals of Clinical Chemistry. Philadelphia, PA: W.B. Saunders, 93 931.

Tietz, N. 1986. Textbook of Clinical Chemistry. Philadelphia, PA: W.B. Saunders, 928-938. Turpin, J., K. Mehta, M. Blick et al. 1990. Effect of retinoids on the release and gene expression of tumor necrosis factor-alpha in human peripheral blood monocytes. J. Leuk. Biol. 48:4450.

Underwood, B.A. 1984. Vitamin A in animal and human nutrition. In The Retinoids, M.B. Spom, A.B. Roberts, and D.S. Goodman, eds. Orlando, FL: Academic Press, v. 1, 281392.

Wako, Y., K. Suzuki, Y. Goto et al. 1986. Vitamin A transport in plasma of diabetic patients. Tohoku J. Exp. Med. 149:133-143.

Willett, W.C., L. Sampson, M.J. Stampfer et al. 1985. Reproducibility and validity of a semi

quantitative food frequency questionnaire. Am. J. Epidemiol.122:51-65. Yagi, K. 1976. A simple fluorometric assay for lipoperoxide in blood plasma. Biochem. Med. 15:212-216.

Copyright Wayne State University Press Feb 1997
Provided by ProQuest Information and Learning Company. All rights Reserved