Is the β thalassaemia trait of clinical importance?

Summary

Although the β thalassaemia trait affects millions of people worldwide, there have been no controlled studies to determine whether it is associated with any clinical disability or abnormal physical signs. To address this question, 402 individuals were studied: 217 with β thalassaemia trait, of whom 154 were aware of the diagnosis and 63 were unaware until after the completion of the study; 89 normal controls; and 96 controls with mild hypochromic anaemia. There was a significant increase in symptoms ascribable to anaemia and episodes of pyrexia in those with the β thalassaemia trait that were not influenced by prior knowledge that they had this condition. There was no difference in physical findings, notably splenomegaly, between those with β thalassaemia trait and either control group.

The inherited disorders of haemoglobin are the commonest human monogenic diseases. It has been estimated that approximately 7% of the world population are carriers, and that 300 000–500 000 babies with severe forms of these conditions are born each year (WHO, 2002; Christianson et al, 2006). Because these conditions occur at their highest frequency across the tropical regions of the world, they pose a particular problem for the developing countries. In sub-Saharan Africa, the main problem stems from the sickle cell disorders, while in Asia, the different forms of thalassaemia are the principle form of haemoglobinopathy. As the developing countries pass through the demographic transition towards improved nutrition and healthcare, and as childhood mortality, i.e., death in the first five years, declines, these disorders pose an increasing health burden on these countries (Weatherall et al, 2006).

The main approach to the control of the thalassaemias in Asia is based on population screening and genetic counselling followed by premarital advice and, where acceptable, prenatal diagnosis. Counselling of thalassaemia carriers is usually restricted to an explanation of the implications of the carrier state for the risk of producing a severely affected child. Very little is discussed about the effect of being a thalassaemia carrier, largely because very little is known; research in this field has tended to focus on the more severe forms of the disease.

While the importance of identifying the β thalassaemia trait for counselling is well established, as is its association with anaemia of pregnancy (White et al, 1985), the literature on associated clinical findings is conflicting; some studies have reported a high frequency of symptoms of anaemia together with splenomegaly (Fessas, 1959; Gardikas, 1968; Mazza et al, 1976), while others have concluded that it is symptomless, with no physical findings (Pootrakul et al, 1973). None of the clinical studies of β thalassaemia trait to date has been controlled however.

In Sri Lanka, a long-term study of the clinical status of older patients with different forms of thalassaemia is under way. The present report describes the findings in a group of β thalassaemia heterozygotes and the clinical accompaniments of this condition and its potential complications are reviewed.

Patients and methods

Subjects were recruited, strictly on a voluntary basis, from two main sources: parents of patients with β thalassaemia major; and volunteers who responded to an advertisement to participate in the study. All subjects were Sri Lankan nationals from the districts of Kurunegala and Anuradhapura and were from a similar socio-economic background. They were divided into four groups (Fig 1). Of the 217 individuals with β thalassaemia trait, 154 were aware of the diagnosis before their clinical assessment, whereas 63 were only told of the diagnosis after the results of their blood tests had been obtained and after their clinical assessment; neither the assessors nor the subjects were aware of the diagnosis. Of the 185 controls, 89 had completely normal haematological findings, whereas 96 showed mild anaemia and red-cell hypochromia, reflecting the high incidence of mild iron deficiency in the Sri Lankan population. All those with β thalassaemia trait received a full explanation and counselling regarding the significance of this finding.

Figure 1.  Characteristics of comparison groups.

Ethical approval was obtained from the Ethics Committee of the University of Colombo.

For each individual, a medical officer completed a questionnaire covering the issues listed in Table I. The same person was responsible for interviewing and examining all the subjects. A validated linear analogue scale was used for self-assessment of quality of life, grading from 1 to 10; those who scored less than 6 were categorised as having adverse health (Streiner & Norman, 1995). All patients underwent a complete clinical examination with particular emphasis on the presence or absence of hepato-splenomegaly. The methods for haematological and haemoglobin analysis have been reported previously (Fisher et al, 2003). Statistical data were examined using an Epi 6 package. P values <0·05 were considered significant.

Table I.   Comparative analysis of symptoms and signs between the study groups. An MCV < 80 fl and MCH < 27 pg without β thalsaaemia trait are labelled as “anaemic controls”.

	Normal	Traits	Anaemic controls	P-values Normal vs. traits	P-values Traits vs. anaemic controls
Values with parentheses are percentages.
Hypertension	0 (0·00)	8 (3·69)	1 (1·04)	0·06	0·206
Diabetes mellitus	0 (0·00)	6 (2·76)	0 (0·00)	0·11	0·104
Ischaemic heart disease	1 (1·12)	3 (1·38)	1 (1·04)	0·85	0·821
Bronchial asthma	6 (6·74)	22 (10·14)	10 (10·42)	0·34	0·887
Frequent episodes of
Headache	17 (19·10)	82 (37·79)	31 (32·29)	0·0018	0·49
Lethargy	5 (5·62)	37 (17·05)	21 (21·88)	0·009	0·23
Fatigue	2 (2·25)	37 (17·05)	21 (21·88)	0·00046	0·23
Dizziness	0 (0·00)	33 (15·21)	14 (14·58)	0·0001	0·99
Sleepiness	3 (3·37)	36 (16·59)	19 (19·79)	0·0015	0·42
Exercise tolerance – frequent symptoms
Climbing uphill	5 (5·62)	33 (15·21)	14 (14·58)	0·02	0·89
Walking or flat	2 (2·25)	19 (8·76)	15 (15·63)	0·04	0·071
Attacks of left upper quadrant abdominal pain	0 (0·00)	4 (1·84)	2 (2·08)	0·202	0·83
Jaundice	0 (0·00)	3 (1·38)	1 (1·04)	0·267	0·83
Pyrexial episodes	23 (25·84)	104 (47·93)	37 (39)	0·0006	0·213
Anaemia in pregnancy	1 (1·89)	22 (14·56)	10 (14)	0·012	0·89
Work absenteeism >2 weeks during past year	0 (0·00)	6 (2·76)	3 (3)	0·113	0·886
>6 visits to a doctor during past year	0 (0·00)	18 (8·29)	9 (9)	0·005	0·57
Self rank of poor health (<6 on a scale of 1–10)	7 (7·87)	40 (18·43)	17 (18)	0·02	0·99
Palpable hepatomegaly	2 (2·25)	16 (7·37)	2 (2·08)	0·069	0·076
Palpable splenomegaly	2 (2·25)	5 (2·30)	3 (3)	0·949	0·63

Results

The results of comparisons between the different groups are summarised in Table I and the haematological and haemoglobin data for each group in Fig 1. The most important finding was the highly significant difference between the normal subjects and those with β thalassaemia trait with respect to symptoms suggestive of anaemia, including lethargy, fatigue, and dizziness. Those with the trait also had significantly more visits to their practitioners within the previous year and, surprisingly, had significantly higher numbers of attacks of fever necessitating medical attention. There was no significant increase in the frequency of palpable spleens or other physical signs between the normal or anaemic controls and those with β thalassaemia trait.

When the β thalassaemia trait data was analysed further by symptomatology, approximately 60% were found to have frequent symptoms. In the symptomatic group the haemoglobin values ranged from 75 g/l to 115 g/l, (mean 107 g/l), whereas the asymptomatic patients had haemoglobin values that ranged from 81 g/l to 150 g/l, with a mean of 113 g/l.

Comparisons between those with β thalassaemia trait who knew they were affected and those who did not showed no significant differences between any of the parameters listed in Table I. The fact that there was no significant difference between the parameters of those with β thalassaemia trait and those with mild anaemia (Table I), adds further weight to the probability that the former is associated with genuine symptoms of anaemia.

Since the high frequency of mild iron deficiency anaemia among the Sri Lankan population is well documented (de Silva et al, 2003) it seems likely that the bulk of those who were found to be mildly anaemic with hypochromic red cell indices were suffering from this condition. Although the α⁺ thalassaemia trait occurs in about 15% of the Sri Lankan population (Fisher et al, 2003), affected individuals are not anaemic and have normal red cell indices.

Discussion

Because there are many millions of carriers for β thalassaemia worldwide, it is important to know whether this condition is the cause of symptoms or complications and, particularly, whether these vary between different populations and environments.

Previous and uncontrolled studies of the symptomatology and the frequency of splenomegaly in individuals with β thalassaemia trait have given variable and inconsistent results (reviewed by Weatherall & Clegg, 2001). The findings in the present study suggest that β thalassaemia trait may cause symptoms that can be ascribed to mild anaemia, and that these are genuine and are not the result of prior knowledge or concern about having this condition. Interestingly, there was no significant difference between these symptoms between the individuals with β thalassaemia trait and those controls that were mildly anaemic. There is now an extensive literature on the effects of variable degrees of iron deficiency anaemia on development and function (Caulfield et al, 2006). Whether these effects are mediated through mild anaemia per se or whether they reflect a more general role of iron deficiency in growth and development is still not clear. Evidently, it will be important to investigate these parameters in children with the β thalassaemia trait.

Although malaria transmission is now relatively low in Sri Lanka (Briet et al, 2005) a few individuals with palpable splenomegaly are still likely to be encountered due to chronic malaria. However, there was no significant difference in the frequency of splenomegaly between those with thalassaemia trait and the control population. It appears therefore that, if an individual with β thalassaemia trait has a palpable spleen, other causes should be sought.

Perhaps the most surprising finding in this study was the significantly increased frequency of infectious episodes in the β thalassaemia group. An increased susceptibility of patients with severe forms of thalassaemia to infection is, of course, well documented. But it has not been reported previously in those with the trait (Weatherall & Clegg, 2001). There is considerable evidence that the β thalassaemia trait confers relative resistance to P. falciparum malaria (Weatherall & Clegg, 2001, 2002). However, the data regarding malaria resistance is confined to infection with P. falciparum. Nothing is known about relative susceptibility to P. vivax malaria which is endemic in parts of Sri Lanka. This parasite only invades young reticulocytes and young red cells and therefore individuals with increased rates of red cell turnover may be more susceptible. Indeed, this has been observed in the case of the mild forms of α thalassaemia in very young children in Melanesia (Williams et al, 1996). Clearly, further studies are required regarding the pattern of infection in individuals with β thalassaemia trait and, in particular, their relative proneness to P. vivax malaria in Asian countries. In view of the suggestion by Williams et al (1996) that early P. vivax infection may have a later protective effect against P. falciparum infection, and because very little is known about the possible mechanisms of protection against P. falciparum by the β thalassaemia trait, studies of this kind will have a much more general interest.

Apart from the well known increased frequency of anaemia of pregnancy in those with β thalassaemia trait, are there any other clinical associations that might be of importance or require further study? In an extensive study aimed at defining the lifespan of individuals heterozygous for β thalassaemia in Italy, no significant differences were found between those with the trait and normals (Gallerani et al, 1990). Surprisingly, the same group, in a prospective study of 4401 subjects, concluded that male β thalassaemia carriers showed a significant degree of protection against myocardial infarction (Gallerani et al, 1991). Furthermore, the mean age at which myocardial infarction occurred in male heterozygotes was significantly higher than in normal males; no such differences were found among female subjects. This topic was reviewed again recently by Tassiopoulos et al (2005), who, after correcting for other risk factors, also showed a reduction in advanced coronary artery disease in β thalassaemia trait. They suggested that the lipoprotein and blood rheology profiles in those with β thalassaemia trait might be the underlying protective mechanism. Whether these observations can be extrapolated to populations in the developing countries remains to be seen.

In short, because the β thalassaemia trait is so common, more needs to be learnt about its functional effects and, in particular, whether in particular populations it is associated with increased proneness to infection and, if so, what type of infection. In particular, it would be very important to try to obtain further information about its relationship to susceptibility to malaria, not just P. falciparum malaria, but also malaria due to P. vivax infection, which is so common in many Asian countries.

Acknowledgements

This work was supported in part by grants to AP and DJW from the Wellcome Trust. We thank Nancy Olivieri for her help in the study, Tim Peto for statistical help and Liz Rose for her help in preparing this manuscript, D. Kottachchi, M. Weerasinghe, R. Samaranayake and V. Perera for help with sample collection at the National Thalassaemia Centre, and S. Mendis and C. Munasinghe for help with data analysis.