The prevalence of thiamine deficiency among elderly nursing home residents: A cross‐sectional study

Abstract

Background

Although thiamine deficiency (TD) can lead to Wernicke encephalopathy, the characteristics associated with TD in the elderly have not yet been clarified. We sought to clarify the frequency of TD among an institutionalized elderly population with a controlled dietary intake and to identify possible factors related to TD.

Method

We undertook a cross‐sectional study of residents in three nursing homes for the elderly as of June 2020. Blood thiamine concentrations were measured using a high‐performance liquid chromatography method, with TD defined as a concentration of <21.3 ng/mL. Basic data (age, sex, height, weight, and BMI), dietary intake for the previous 3 weeks, degree of care (DOC), degree of independence in daily life for elderly with dementia (DIDLED), and comorbidities were analyzed using descriptive statistical methods.

Results

The mean age (±SD) was 86.9 years (±8.29), with 84 residents (70.0%) being female. The DIDLED varied from total independence to long‐term care level 5 (full assistance), with 89.2% suffering dementia. The mean whole blood thiamine value was 36.18 (±17.58) ng/ml, with TD confirmed in 7 (5.8%) of the 120 residents. All TD patients suffered from dementia. No TD was observed in patients with a near‐normal food intake, and no related factors were observed among the other items.

Conclusion

Reduced food intake may at increase the risk of TD and symptoms of TD may be overlooked in those displaying symptoms of dementia; thus, it is important for clinicians working with the elderly to remain aware of the potential for TD.

We sought to clarify the frequency of TD among an institutionalized elderly population with a controlled dietary intake and to identify possible factors related to TD. 5.8% of them had thiamine deficiency. Reduced food intake may increase the risk of TD and symptoms of TD may be overlooked in those displaying symptoms of dementia; thus, it is important for clinicians working with the elderly to remain aware of the potential for TD.

1. INTRODUCTION

Thiamine (vitamin B1), in its biologically active form, thiamine pyrophosphate, is an essential coenzyme for glucose metabolism. However, as it cannot be synthesized in the human body, it must be taken in from external sources. The physiological store of thiamine is depleted within about 18 days, so a continued reduction in food intake because of loss of appetite can easily lead to thiamine deficiency (TD). ¹ Wernicke encephalopathy (WE) is a typical neuropsychiatric disorder caused by TD. The classic triad of signs of WE are disturbed consciousness, ataxic gait, and nystagmus, but less than 20% of cases show all three signs. ² Further, TD is often overlooked because the symptoms may be nonspecific, and a number of apparently asymptomatic cases have also been reported. ³ , ⁴

Risk factors for TD include alcoholism, hyperemesis gravidarum, anorexia, and heart failure, as well as advanced age and dementia, ¹ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ with cancer and depression also reported in recent years. ¹⁰ , ¹¹ , ¹² , ¹³ , ¹⁴ It has also been reported that the use of proton pump inhibitors (PPIs) and diuretics may reduce blood thiamine levels. ⁸ , ¹⁵ , ¹⁶ , ¹⁷

However, as mentioned above, asymptomatic TD is also common, ¹² , ¹⁸ and a number cases in which TD was overlooked have been reported. ¹⁹ In particular, it is possible that TD is overlooked in the elderly as it is not uncommon for the elderly to be unable to communicate as a result of suffering dementia or being bedridden. In fact, TD has been shown to exacerbate the behavioral and psychological symptoms of dementia (BPSD) in some cases. ²⁰ In addition, it has been reported that community‐dwelling elderly may not be consuming adequate levels of thiamine. ⁵ In other words, although there is a risk of TD in the elderly, the actual situation regarding the associated factors has not been clarified. Therefore, we undertook a cross‐sectional study focused on the elderly in nursing homes where dietary intake is properly managed, and investigated dietary intake, degree of care (DOC), degree of independence in daily life of elderly with dementia (DIDLED), and Body Mass Index (BMI) to clarify what factors were associated with TD.

2. METHODS

2.1. Study design

This is a cross‐sectional study of elderly nursing home residents.

2.2. Setting

Subjects consisted of those, aged 65 years or older, who had been resident in facilities for the elderly (Facility A, B, C) in X Town, Y Prefecture for 3 months or more continuously, who were capable of oral intake and from whom written consent was obtained (or for whom consent was obtained from the family or equivalent when judged unable to communicate). Facility A and B are special nursing homes for the elderly, where elderly people requiring a high degree of nursing care live with physical care and living support provided, the purpose of these facilities being to care for the elderly. On the other hand, facility C is a nursing home for the elderly who are physically independent but have difficulty living at home because of environmental and economic problems. As the main purpose of this facility is to care for the elderly and help them reintegrate into society, those who need less care than those in special elderly nursing homes are admitted. However, with the recent aging of residents, at present, residents also include those who are highly dependent on nursing care.

2.3. Thiamine measurement

In this study, venous blood samples (approx. 2 mL) collected at a health checkup in June 2020 were used for thiamine measurement. All residents of the facilities in this study underwent a physical examination and blood testing every June.

The measurement of whole blood thiamine values was entrusted to BML Corporation, who used high‐performance liquid chromatography (HPLC) (reference range: 21.3–81.9 ng/mL). The current study measured thiamine levels in whole blood. Measurement of thiamine concentration in whole blood provides the sum of free thiamine present in serum or plasma, and mono‐ and di‐phosphorylated thiamine present in erythrocytes and leukocytes, and this value is said to reflect the amount of thiamine in tissues. ²¹

2.4. Background data (for nursing home residents)

Date of birth, sex, height, weight, BMI, and comorbidities from the Charlson Co‐morbidity Index (CCI) ²² obtained from nursing care and medical records were examined.

2.5. Nursing care data

The average daily food intake for the previous 3 weeks as of June 2020 was calculated from nursing care records and classified into five levels according to the classification of Barbato ⁶ ; (1) normal (90% or more), (2) about 2 meals a day (50–89%), (3) about 1 meal a day (10–49%), (4) a few bites a day (1–9%), (5) no intake (0%).

DOC (ranging from “independent,” where no nursing care is required in daily life, to a level at which daily life is possible with supervision and a little help, Requiring help 1 and 2, and Long‐term care 1–5) (Table S1) and DIDLED (Independent, I, IIa, IIb, IIIa, IIIb, IV, M in order of the severity of dementia) (Table S2) were also measured.

2.6. Statistical analysis

Patient background factors were summarized using descriptive statistical methods (Table 1), and the prevalence of TD was calculated by dietary intake, DOC, DIDLED, and BMI (< or >18.5) (Table 2). The distribution of thiamine concentration was plotted on histograms by total and dietary intake (Figure 2, Figure 3).

TABLE 1.

Patient characteristics (n = 120).

Variables		n	(%)
Female		84	(70.00)
Age
Min–max	66–102
Mean (SD)	86.90 (8.29)
Median	89
BMI
Min–max	14–32.7
Mean (SD)	21.32 (3.14)
Median	21.2
Charlson index score
Min–max	0–6
Mean (SD)	2.32 (1.30)
Median	2
0		1	(0.83)
1		46	(38.33)
2		24	(20.00)
≥3		49	(40.83)
Myocardial infarction		3	(2.50)
Congestive heart failure		21	(17.50)
Peripheral vascular disease		10	(8.33)
Cerebrovascular disease		44	(36.67)
Dementia		107	(89.17)
Chronic pulmonary disease		1	(0.83)
Rheumatologic disease		3	(2.50)
Peptic ulcer disease		4	(3.33)
Mild liver disease		2	(1.67)
Diabetes without chronic complications		4	(3.33)
Hemiplegia or paraplegia		31	(25.83)
Renal disease		2	(1.67)
Diabetes with chronic complications		1	(0.83)
Any malignancy, including leukemia and lymphoma, except malignant neoplasm of skin		5	(4.17)
Moderate or severe liver disease		0	(0.00)
Metastatic solid tumor		0	(0.00)
AIDS/HIV		0	(0.00)
Barbato
(1)		57	(47.50)
(2)		27	(22.50)
(3)		36	(30.00)
(4)		0	(0.00)
(5)		0	(0.00)
DOC
Independent		10	(8.33)
Requiring help 1		2	(1.67)
2		4	(3.33)
Long‐term care 1		6	(5.00)
2		6	(5.00)
3		22	(18.33)
4		38	(31.67)
5		32	(26.67)
DIDLED
Independent		3	(2.50)
I		7	(5.83)
IIa		18	(15.00)
IIb		21	(17.50)
IIIa		34	(28.33)
IIIb		16	(13.33)
IV		20	(16.67)
M		1	(0.83)

Note: Barbato: (Mean daily food intake over the past 3 weeks) (1) normal (90% or more), (2) about 2 meals a day (50–89%), (3) about 1 meal a day (10–49%), (4) a few bites a day (1–9%), (5) no intake (0%).

Abbreviations: BMI, body mass index; DOC, Degree of care; DIDLED, Degree of independence in daily life of elderly with dementia.

TABLE 2.

Comparison of dietary intake, DOC, DIDLED, and BMI between TD and non‐TD patients.

	TD n = 7		Non‐TD n = 113
	n	(%)	n	(%)
Barbato
(1)	–	–	57	(50.4)
(2)	3	(42.9)	24	(21.2)
(3)	4	(57.1)	32	(28.3)
(4)	–	–	–	–
(5)	–	–	–	–
DOC
Independent	–	–	10	(8.8)
Requiring help 1	–	–	2	(1.8)
2	–	–	4	(3.5)
Long‐term care 1	–	–	6	(5.3)
2	1	(14.3)	5	(4.4)
3	1	(14.3)	21	(18.6)
4	1	(14.3)	37	(32.7)
5	4	(57.1)	28	(24.8)
DIDLED
Independent	–	–	3	(2.7)
I	–	–	7	(6.2)
II (IIa + IIb)	–	–	39	(34.5)
III(IIIa+IIIb)	6	(85.7)	44	(38.9)
IV	1	(14.3)	19	(16.8)
M	–	–	1	(0.9)
BMI
<18.5	1	(14.3)	21	(18.6)
≧18.5	6	(85.7)	92	(81.4)

FIGURE 2.

Thiamine concentration distribution (n = 120).

FIGURE 3.

Thiamine concentration distribution by dietary intake.

3. RESULTS

A total of 120 people who met the eligibility criteria were enrolled in the study (Figure 1). The mean age (±SD) was 86.9 years (±8.29), with 70.0% (84) of those enrolled being female. BMI was 21.32 (±3.14), mean CCI score was 2.32 (±1.30), and 89.2% of the subjects had dementia. The average daily food intake for the previous 3 weeks (based on the 5‐level classification of Barbato) was (1) normal (90% or more), 47.5%, (2) about 2 meals a day (50–89%) 22.5%, (3) about 1 meal a day (10–49%) 30.3%, and there were no subjects in category (4) or (5). With regard to the DOC, 8.3% were Independent, 1.7% were Requiring help 1, 3.3% were Requiring help 2, 5.0% each were Long‐term care 1 and 2, 18.3% were Long‐term care 3, 31.7% were Long‐term care 4, and 26.7% were Long‐term care 5. In terms of DIDLED, 2.5% were independent, 5.8% were level I, 15.0% were level IIa, 17.5% were level IIb, 28.3% were level IIIa, 13.3% were level IIIb, 16.7% were level IV, and 0.8% were level M (Table 1).

FIGURE 1.

Patient selection.

The mean (±SD) whole blood thiamine concentration was 36.2 (±17.6) ng/ml with a median of 34.5 ng/mL (Figure 2). TD was confirmed in 7 (5.8%) of the 120 residents. All residents with TD suffered from severe dementia (Table S2) with a degree of dementia of IIIa or higher that required constant nursing care.

Table 2 presents the relationships between dietary intake, DOC, DIDLED, and BMI in the thiamine‐deficient and non‐thiamine‐deficient groups. No TD was observed in those with near‐normal food intake (Barbato level 1) or in those with a DOC of less than Long‐term care 1.

4. DISCUSSION

In this study, we measured thiamine concentration in the residents of elderly nursing homes to clarify the prevalence of TD and its related background characteristics. The results confirmed TD (thiamine <21.3 ng/mL) in 7 of 120 residents (5.8%). In addition, all of the patients diagnosed with TD were suffering from advanced dementia and had a DIDLED of IIIa or higher.

The three facilities surveyed this time are run by the same corporation, and the food content provided and the food intake of the residents are managed uniformly at the three facilities, so it can be said that the survey was conducted under almost the same conditions for all residents enrolled. The fact that as many as 5.8% of the subjects were actually thiamine deficient is a surprising result. Elderly populations have been identified as being at high‐risk for TD. ¹ , ²³ In addition, it has been reported that community‐dwelling elderly may not receive an adequate thiamine intake. ⁵ Simply being an older adult confers a high risk of TD, so even under conditions in which meals are uniformly managed, the elderly must always be conscious of the potential for TD.

As the elderly population often suffer comorbidities, we investigated comorbidities using the Charlson Co‐morbidity Index (CCI). ²² Results revealed that all the thiamine‐deficient residents in this study had severe dementia. A relationship between dementia and TD has been noted previously. ¹ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ Of the 120 residents enrolled in this study, 107 (89.2%) also had dementia, and it can be suggested that the majority of the participants in this study were at risk for TD at that time.

Decreased dietary intake can easily lead to TD. ¹ We investigated dietary intake using the classification of Barbato, ⁶ and found that none of the subjects in this study showed a dietary intake limited to only a few bites a day. All of the subjects were eating at least one meal a day. Among them, all of the thiamine‐deficient subjects in our study ate only one or two meals a day (Barbato classification 2 and 3) (Table 2, Figure 3). In other words, it appears that people who are only able to take in approx. half of their normal diet may experience TD.

There were no thiamine‐deficient persons among those requiring a low level of care (Long‐term care level 1 or less). Although it may be seen as unreasonable to discuss thiamine deficiency based only on DOC scores, the higher the DOC required, the lower the level of consciousness and the more difficult it is to communicate; thus, changes in conditions associated with TD, such as disturbance of consciousness, may go unnoticed. In fact, there are reports that TD is overlooked unless it is suspected by clinicians. ¹⁹ Asymptomatic TD has also been reported, ¹² , ¹⁸ and it is reported that 60–80% of cases of WE are overlooked. ¹ Blood tests are the only way to diagnose asymptomatic TD. ¹ It is important for medical staff to always be aware of this fact and to conduct regular blood thiamine measurements for people in special situations, such as those in institutions. Measurement of suspected TD has the benefit of providing information on the physical condition of the residents; however, in consideration of the high cost, the time needed to obtain test results, ³ and the burden of venipuncture for patients, supplementary thiamine administration may afford a method of preventing TD without the need for blood thiamine measurement.

This study has a number of limitations. First of all, the study population was small at 120 people. Further study based on a larger number of subjects may reveal more characteristics associated with TD. For example, some patients have been reported to develop TD despite adequate dietary intake. ¹³ , ²⁴ , ²⁵ In this study, no TD was observed among those who were able to consume almost all of their meals; however, by increasing the sample size, it may be possible to detect TD among those in whom dietary intake is not limited.

Second, dietary intake was only investigated for a 3‐week period, so it may be possible to find potential TD by observing intake over a longer period.

There are also reports that changes in body weight are associated with TD ¹⁴ ; however, only body weight was measured at the time of blood sampling in this study, and as we did not observe changes in body weight over the 3‐week period in which food intake was reviewed, we were unable to respond to changes over time.

Third, the meals provided by institutions also varied. The recommended daily intake of thiamine in the elderly is 1.2 mg/day for men and 1.0 mg/day for women. ⁸ However, it is unknown whether the recommended thiamine was actually provided. It is known that cooking methods reduce thiamine, ²⁶ thus it is difficult to accurately grasp the thiamine intake at each facility. It has been reported that 56.8% of nursing home residents were actually deficient in thiamine. ²⁷ There is also a report that TD was more common in those hospitalized from institutions than in those hospitalized from home. ⁸ For this reason, it cannot be denied that there are many biases such as the content of meals provided at facilities and the resident characteristics, but the results should sound a warning bell that residents of nursing homes in various regions are at risk of TD.

Fourth, the three facilities that participated were operated by the same corporation, and that the content of meals was judged to be basically uniform. However, it is conceivable that differences in the quality of nursing care provided to residents may result in differences not only in dietary intake but also in the management of general conditions. The rate of TD in nursing homes in Japan is as high as 58.8% ²⁷ and the prevalence of TD can vary widely depending on the quality of care in an institution and the caregiving skills of each caregiver. Therefore, further study based in other nursing facilities is required to clarify these factors.

5. CONCLUSION

Seven of 120 (5.8%) residents of elderly nursing homes with dementia were thiamine deficient. This prevalence is by no means low, and we consider elderly people with dementia who have a limited food intake, in particular, are more likely to be deficient in thiamine. Symptoms associated with TD can be improved without sequelae if treated early. It is important for clinicians working with the elderly to always suspect and measure TD.

CONFLICT OF INTEREST STATEMENT

The authors have stated explicitly that there are no conflicts of interest in connection with this article.

ETHICS APPROVAL STATEMENT

This study was approved by the Ethics Committee of Ogano Town Central Hospital (Ogano 2020 No. 9).

PATIENT CONSENT STATEMENT

None.

CLINICAL TRIAL REGISTRATION

None.

Supporting information

Table S1

Table S2

Uchida N, Ishida M, Sato I, Yoshioka A, Takahashi T, Furuya D, et al. The prevalence of thiamine deficiency among elderly nursing home residents: A cross‐sectional study. J Gen Fam Med. 2023;24:148–153. 10.1002/jgf2.609