The proportion of older Thai people is growing rapidly as a result of increased life expectancy. Since older people tend to suffer from various vascular and neurodegenerative disorders as well as cancers, preventive measures are needed to keep them fit as they are living longer. Nutrition and physical activity are two main lifestyle factors for which there is strong evidence of a positive effect in the prevention and management of such disabling conditionsReference Johnson, Bernard and Funderburg1, Reference Drewnowski and Shultz2. Although many nutritional surveys have previously been done in developed countries, surveys recruiting large numbers of older subjects in developing countries have been scarce and would be expected to yield different results due to socio-economic and cultural differences. However, the vitamin status of older people varies even within developed countries. For example, only 5% of blood vitamin concentrations analysed in the German VERA study (Verbundstudie Ernährungserhebung und Risikofaktorenanalytik) were found to be below the physiological range, while the prevalence of vitamin B12 deficiency was found to be 3–44% in the USAReference Volkert and Stehle3, Reference Lindenbaum, Rosenberg, Wilson, Stabler and Allen4. Furthermore, according to the Euronut SENECA study (Survey in Europe on Nutrition and the Elderly; a Concerted Action) involving 12 European countries, biochemical vitamin E deficiency was found in seven centres with prevalence ranging from 0.5 to 25%5. Therefore, the specific purpose of this mega-project funded by the Thai government was to gain the necessary health data for national policy planning.
Materials and methods
The present cross-sectional study was part of a comprehensive survey of older people in Thailand approved by the Research Committee of Mahidol University. The whole project finished in the year 2000. One district in each of the four regions of the country was randomly selected. For the nutritional survey, around 500 older people aged 60 years and over in each region were randomly recruited, leading to a total number of 2336 subjects. After background characteristics were assessed, the subjects underwent anthropometric measurements as well as blood tests to measure biochemical data and various vitamin levels, i.e. vitamin A (retinol), β-carotene, vitamin B1 (thiamin), vitamin B12 (cyanocobalamin), folic acid, vitamin C (ascorbic acid) and vitamin E (α-tocopherol).
Whole blood and serum were kept at − 20°C before biochemical assays to guarantee the accuracy of measurement. Retinol, β-carotene and α-tocopherol were measured by high-performance liquid chromatography (HPLC). Since normal levels of retinol6, β-carotene7 and α-tocopherolReference Packer and Fuchs8 are respectively 20–50, 25–125 and 600–1200 μg dl− 1, the cut-off points for vitamin A, β-carotene and vitamin E deficiency were set at < 20, < 25 and < 600 μg dl− 1, respectively. Erythrocyte transketolase activity and the thiamin pyrophosphate (TPP) effect were used to identify thiamin deficiency. Those who had a TPP effect of >15% were classified as vitamin B1-deficient. A microbiological assay using Lactobacillus casei ATCC 7469 was used to measure the folic acid level. A laboratory technique employing radioisotope dilution and coated charcoal was used in the measurement of cyanocobalamin. Since normal levels of cyanocobalamin, serum folate and erythrocyte folate among the Thai population are 200–800 pg dl− 1, 5–21 ng dl− 1 and 221–1113 ng dl− 1, the cut-off points indicating biochemical deficiency of vitamin B12, serum folate and erythrocyte folate were 200 pg dl− 1, 5 ng dl− 1 and 221 ng dl− 1, respectivelyReference Devakul, Sundharagiati and Pooprasert9–Reference Areekul and Kitkornphan12. Serum ascorbic acid was measured by HPLC. As normal ascorbic acid level is in the range 0.4–1.5 mg dl− 1, levels < 0.4 mg dl− 1 indicate vitamin C deficiencyReference Jacob, Shils, Olson and Shike13.
The software package SPSS version 10.0 (SPSS Inc., Chicago, IL, USA) was for statistical analyses with the level of significance set at P < 0.05. Descriptive statistics (mean ± standard deviation, SD) were calculated for each parameter of demographic characteristics, underlying medical conditions, lifestyle patterns and biochemical data and analysed by univariate analysis (chi-square and Student t-tests) to determine the factors associated with the examined vitamin deficiencies. The statistically significant factors were sequentially entered in a multiple logistic regression analysis to find the independent risk factors for each biochemical vitamin deficiency.
Results
The percentage of participants from the central, northeastern, northern and southern regions was 26.2, 25.3, 25.4 and 23.0%, respectively. The proportion of males to females was 889:1447 (1:1.63). The mean ( ± SD) age of subjects was 68.94 ± 6.76 years (range 60–97 years). The mean ( ± SD) body mass index was 22.62 ± 4.34 kg m− 2 (range 10.74–47.94 kg m− 2). Regarding the educational background of the subjects, 24.6% had never been to formal school and 65.8% had achieved only a primary education. About 57% still lived with their spouse while 6.3% lived alone. Although 14.4% had satisfactory financial status, 18.7% admitted that they had inadequate income for daily life. Regarding health-risk behaviours, 31.8% were either current smokers or had quit smoking for 10 years or less and 25.0% were either current alcohol drinkers or had quit drinking for 10 years or less.
The prevalence of vitamin deficiencies was 6.1% for vitamin A, 83.0% for β-carotene, 30.1% for vitamin B1, 38.8% for erythrocyte folate, 9.9% for vitamin C and 55.5% for vitamin E. The prevalence of vitamin B12 deficiency was only 0.6%. Since the prevalence of vitamin B12 deficiency was very low among older Thai adults, we decided not to study the factors associated with this deficiency in detail.
Table 1 shows the prevalence of vitamin deficiencies according to selected demographic characteristics and underlying medical conditions, while Table 2 compares the factors associated with the various vitamin deficiencies in univariate analyses. The independent factors determining each vitamin deficiency derived from multiple logistic regression analyses are shown in Table 3.
Prevalence (%) of each qualitative factor found in various vitamin deficiencies
* Significant difference compared with those who did not have the vitamin deficiency (univariate analysis): P < 0.05.
Comparisons of quantitative factors associated with various vitamin deficiencies by univariate analysis
BMI – body mass index; TPP – thiamine pyrophosphate; NA – not applicable.
Values are expressed as mean ± standard deviation.
* Significant difference compared with those who did not have the vitamin deficiency: P < 0.05.
Independent factors determining each vitamin deficiency by multiple logistic regression analysis
OR – odds ratio; CI – confidence interval; DBP – diastolic blood pressure; HDL – high-density lipoprotein; BMI – body mass index; TPP – thiamine pyrophosphate.
Discussion
With the exception of vitamin B12 deficiency, the lowest prevalence was found for vitamin A deficiency (6.1%) which is consistent with surveys in other parts of the worldReference Panemangalore and Lee14. This universal finding may be due to large hepatic retinol stores and reduced renal clearance of retinyl esters in the elderly. It takes up to 9 months for severely retinol-depleted subjects to show a biochemical deficiencyReference Bamji and Briggs15. However, although our survey was done nearly 20 years after the NHANES I study (National Health and Nutrition Examination Survey), the prevalence among older Thai people was much higher than in that study (0.3%)Reference Bowman and Rosenberg16. The subjects who were at risk of vitamin A deficiency were those who had memory problems, high diastolic blood pressure and a low vitamin E level. This may be due to poor dietary habits among people with cognitive inability. The coexistence of vitamin A and E deficiency, both lipid-soluble vitamins, was previously described in patients with chronic leg ulcers and poor wound healingReference Rojas and Phillips17.
In the light of factors known to affect carotene levels such as gender and fat intakeReference Henderson, Mobarhan, Bowen, Stacewicz-Sapuntzakis, Langenberg and Kiani18, our results accordingly showed that male gender, higher triglycerides and high-density lipoprotein significantly determined the cases of β-carotene deficiency. Not only because of its highest prevalence (83.0%) among our survey of vitamin levels, but also because it is one of the antioxidant nutrients, older male adults and especially those who once worked as manual workers must be targeted by health personnel to increase the intake of β-carotene-rich foods in their daily diet. Other independent nutritional factors which were found more among β-carotene-deficient subjects were higher albumin and globulin and a higher folate level. This may be due to the lack of nutritional knowledge among older people and recommendations to increase diet variety should be implemented.
Subjects with vitamin B1 deficiency also had other poor general nutritional indices, e.g. body mass index. This association was much more striking than for other vitamin deficiencies. Since the main sources of vitamin B1 among Thai people are rice, pork and legumes, which are the main ingredients in Thai dishes, cases with vitamin B1 deficiency consequently suffer from protein–energy malnutrition. Our prevalence (30.1%) was quite high compared with previous studies, where a prevalence of 3–15% has been foundReference Iber, Blass, Brin and Leevy19. It was not surprising that independent factors of vitamin B1 deficiency were no formal education and being a manual worker. Poverty was found to be a significant factor for vitamin B1 deficiency in the NHANES I studyReference Bowman and Rosenberg16. However, subjects with diabetes were protected from vitamin B1 deficiency; this may be due to the common vitamin B complex prescribed by doctors for diabetic patients who usually complain of peripheral numbness due to diabetic neuropathy. It is a common medical myth that dry beriberi has a burning paresthesia, so vitamin B1 might do well in diabetic neuropathy also.
Some conflicting results showed that those who had folate deficiency had significantly higher vitamin B1 (lower TPP effect) and β-carotene levels. Because both folic acid and β-carotene are mostly found in green leafy vegetables and fruits, the explanation cannot be from poor intake; however, 50–95% of the natural folate content of foods is lost in the cooking and canning processes. Edentulous older people could only consume overcooked boiled rice as their main food, which is a good source of vitamin B1. Edentulism is not associated with total energy or food intake, but rather with the food groups consumedReference Lee, Weyant, Corby, Kritchevsky, Harris and Rooks20. Since folate deficiency is associated with the pathogenesis of atherosclerotic lesions, our results accordingly revealed that male gender, history of heart disease, lower high-density lipoprotein and haematocrit were independent risk factors of folate deficiency. Furthermore, our prevalence of erythrocyte folate deficiency was 38.8%, much higher than that found in developed countries; it was only 3.3% in New Zealand and hardly found in the Euronut SENECA study5, Reference Hanger, Sainsbury, Gilchrist, Beard and Duncan21. Folate supplements, which are relatively inexpensive and safe, should be routinely prescribed to those older people with atherosclerotic risk. Folate status appears to be of greater concern than vitamin B12 status for older peopleReference Quinn and Basu22. This opinion is consistent with our very low prevalence of vitamin B12 deficiency, i.e. only 0.6%. The main explanations for this finding are the nationwide consumption of fish sauce, which is not only inexpensive but also rich in vitamin B12, and the low incidence of atrophic gastritis (11.6%) in ThailandReference Atisook, Kachinthorn, Luengrojanakul, Tanwandee, Pakdirat and Puapairoj23.
As a result of geographic differences, our prevalence of vitamin C deficiency was much lower than the general figure, i.e. 9.9% versus 25%Reference Cheng, Cohen, Bhagavan and Watson24. Various kinds of oranges can easily be found all year round and are affordable throughout Thailand due to its tropical climate. This may also be the main reason why smoking was not an independent factor for vitamin C deficiency in our study, as seen in a study from The Netherlands in which the vitamin C level was not affected by smoking but was strongly associated with the daily intakeReference Lowik, Hulshof, Schneijder, Schrijver, Colen and van Houten25. Interestingly, vitamin C deficiency was the only deficiency found in which increasing age independently determined cases. This highlights the poor access to food and the higher requirements of the older adult26. Those who had vitamin C deficiency had a higher erythrocyte folate level. Since both vitamin C and folic acid are found in large amounts in leafy green vegetables, poor intake could not explain such a dilemma. However, as vitamin C is essential in the biochemical effect of folate reductase, which changes folate to tetrahydrofolate resulting in the synthesis of nucleic acids and DNA, lack of vitamin C might lead to the accumulation of folic acid in tissue and not in serum. Serum folic acid reflects daily intake, which should be low in subjects with vitamin C deficiency because it is commonly found in the same food sources. This explanation is supported by our evidence that those who had vitamin C deficiency also had lower serum folic acid (P = 0.024) but a higher erythrocyte folate level than those who had a normal vitamin C level.
The prevalence of vitamin E deficiency was quite high compared with that found in developed countries, i.e. 55.5% versus 2.5%Reference Baker, Frank, Thind, Jaslow and Louria27. The same explanation as in folate deficiency may be applied in this case, since vitamin E could easily be degraded during the cooking process as well. The scenario that a majority of those with vitamin E deficiency were of male gender, ex-office workers and had higher diastolic blood pressure giving rise to more heart disease may be related to the antioxidant role of vitamin E in atherosclerosis. Those who had vitamin C deficiency and lower β-carotene, serum folate and vitamin A levels were also more likely to suffer from vitamin E deficiency. These relationships are simply explained by inadequate intake of vegetables and fruits.
For national policy planning in primary prevention of vitamin deficiencies among older Thai people, those who are at risk of each vitamin deficiency can be targeted in the community as follows.
• Those who consume their daily diet inappropriately, e.g. have a monotonous daily diet. This group involves older people who have one kind of vitamin deficiency but have other better nutritional parameters.
○ For edentulous older people, easy access to dentures should be promoted nationwide.
○ For those who live alone or are unaware of the importance of diet variety in maintaining health, repeated heath education should be implemented in the community.
• Those who are unable to access adequate nutrition. This group has multiple vitamin deficiencies which are derived from the common food sources.
○ For the poor (poor financial status, low education, manual workers) who are more likely to have vitamin B1 deficiency, folate deficiency and protein–energy malnutrition, social input should be the first priority.
○ For the very elderly who are more likely to have vitamin C deficiency, repeated health education to caregivers should be highlighted.
• Those who are at high risk of atherosclerosis (male gender, high blood pressure, ex-office worker, history of heart disease, low high-density lipoprotein, haemocrit).
○ Folic acid and vitamin E status should be routinely considered by the physicians who look after older patients.
• The three most common vitamin deficiencies among older Thai people are vitamin E, folic acid and vitamin B1. Vitamin B12 deficiency is not a primary health problem in the country.
Acknowledgements
The authors thank Professor Emeritus Adulya Viriyavejakula, Head of the Comprehensive Survey of Older Thai People Project and Mrs Ratana Petchurai and her team from the Department of Research Administration, Mahidol University for project management. Statistical analysis was supervised by Mr Suthipol Udompunturak, Department of Research Promotion, Faculty of Medicine, Siriraj Hospital.
