OUP user menu


David H. Overstreet, Amir H. Rezvani, Abbas Parsian
DOI: http://dx.doi.org/10.1093/alcalc/34.3.378 378-385 First published online: 1 May 1999


A recent study conducted a factor analysis on 18 behavioural measures obtained from four alcohol-preferring and five alcohol-non-preferring rat lines/strains. It was concluded that variables such as saccharin intake, ultrasonic vocalizations following an air puff, and defaecation in an open field were associated with voluntary and forced alcohol consumption. In contrast, measures such as time immobile in the forced swim test and time spent in the open arms of the elevated plus maze were not consistently associated with voluntary alcohol intake. The present study focuses on alcohol intake and related measures in four inbred strains of Fawn-Hooded (FH) rats that differ in voluntary alcohol intake and the ACI/N inbred rat strain, which voluntarily consumes very little alcohol. FH rats inbred by Jean Dodds (FH/Wjd) drank significantly more alcohol than FH rats inbred by Gordon Harrington (FH/Har) or selectively inbred by Abraham Provoost (FHH/Eur and FHL/EUR). In contrast, only the FH/Har strain was active in the forced swim test, suggesting that immobility and voluntary alcohol intake may be influenced by different genetic factors. The FH/Wjd rats were also much more immobile than the ACI/N rats in the forced swim test and drank almost 10 times as much alcohol voluntarily. Comparing the two parental lines with reciprocal F1 crosses revealed that alcohol consumption was influenced largely by additive genetic factors (F1 progeny had intermediate scores), whereas immobility was also influenced by dominance genetic factors (F1 progeny resembled the FH/Wjd parent). Preliminary analysis of 43 F2 progeny indicated that alcohol intake and immobility were not correlated. Thus, immobility in the forced swim test and high voluntary consumption of alcohol, two prominent features of the FH/Wjd rat strain which may be related to its serotonergic dysfunction, appear to be mediated by different genetic factors.


Selective breeding for high and low alcohol consumption in rats has resulted in a number of different alcohol-preferring lines about which much information has been collected. Sometimes behavioural features claimed by one laboratory to be characteristic of its alcohol-preferring rats have not been replicated. Consequently, our laboratory recently conducted a factor analytical study comparing three of the major selected lines with the inbred Fawn-Hooded strain (FH/Wjd) and their non-preferring counterparts and the alcoholnon-preferring inbred ACI strain (Overstreet et al., 1997a). This study confirmed the association between excessive saccharin intake and high voluntary alcohol intake, but indicated that several other measures which had been claimed in isolated reports to be associated with voluntary alcohol intake did not load on the alcohol factor. In particular, immobility in the forced swim test and time spent on the open arms of the elevated plus maze were not consistently associated with high voluntary alcohol intake (Overstreet et al., 1997a).

Unlike the studies with mice, much less attention has been devoted to differences in alcohol intake in inbred rat strains. Li and Lumeng (1984) reported large differences in alcohol intake in the eight inbred strains maintained at the National Institutes of Health (NIH) which were used to create the heterogeneous N/Nih strain. However, none of these strains drank as much as the High Alcohol Drinking (HAD) line of rats that was selectively bred from the N/Nih strain (Li et al., 1993). The FH/Wjd rat was found to drink almost as much alcohol voluntarily as selectively bred lines such as the P, AA, and HAD rats in the early 1990s (Rezvani et al., 1990, 1991) and recent studies confirm the high alcohol intake of this strain (Rezvani et al., 1995; Overstreet et al., 1997b).

However, the literature has become clouded in recent years by the publication of reports on a FH strain of rat which was inbred by Gordon Harrington (1981) at Northern Iowa University (originally known as the IR or Iowa Reactive rat but referred to here as FH/Har). Investigators have failed to confirm the high immobility observed in the FH/Wjd rats (Lahmame et al., 1996). In a recent direct comparison of the FH/Wjd and FH/Har rats, the high immobility and voluntary alcohol intake of the FH/Wjd rats was confirmed; in contrast, the FH/Har rats were quite active in the swim test and drank only moderate amounts of alcohol (Overstreet and Rezvani, 1996). The present communication presents a replication of these original observations but includes, in addition, a comparison with two other inbred Fawn-Hooded strains, those selectively inbred for high blood pressure (FHH/ Eur) or low blood pressure (FHL/Eur) at Erasmus University in the Netherlands (Provoost and DeKeijzer, 1993).

According to the original report of Li and Lumeng (1984), the ACI/N rats were among the lowest voluntary consumers of alcohol, with intakes less than 0.5 g/kg. It was surprising, therefore, that the inbred ACI rats which we obtained from Harlan Sprague–Dawley for our factor analytical study drank around 2 g/kg ethanol (Overstreet et al., 1997a). Subsequently, we obtained several breeding pairs of ACI/N rats from NIH and established our own breeding colony. This communication reports that the ACI/N rats do indeed consume very little alcohol and that, relative to the FH/Wjd rats, are very active in the swim test. In addition to the data on the two parental lines, this communication provides data on the reciprocal F1 crosses between the two strains and on an initial group of F2 progeny.



After their original acquisition from the New York State Department of Health (courtesy of W. Jean Dodds) and the NIH (courtesy of Carl Hansen), respectively, the FH/Wjd and ACI/N rats were maintained in a viral-free animal facility approved by the American Association for Accreditation of Laboratory Animal Care (AAALAC). Animals used in the experiments described here were selected at approximately 70 days of age from breeding colonies established at the University of North Carolina. Members of the FH/Har strain (IR) were selected from the first set of progeny from rats obtained from the National Cancer Institute. Members of the FHH/Eur and FHL/Eur strains were kindly provided by Abraham Provoost of Erasmus University and were tested at about 90 days of age after they had completed quarantine.

Rats were maintained in standard laboratory conditions with constant temperature (22°C) and relative humidity (50%) and with free access to food and water, except as described below. The rats were maintained on a reversed light:dark cycle with lights out between 10:00 and 22:00. Initially, all rats were group-housed until the completion of the forced swim test and elevated plus maze tests. When intake of specific solutions was required, a two-bottle choice paradigm was used and rats were housed individually in metal hanging cages containing two Richter drinking tubes. In a subgroup of FH/Wjd and ACI/N rats, the sequence of testing was reversed, with the two-bottle choice test preceding the swim and plus maze tests. The University of North Carolina Institutional Animal Care and Use Committee approved these experiments.

Elevated plus maze test

The elevated plus maze consisted of a wooden maze painted brown with two enclosed arms and two open arms opposite each other. The rat was placed at the intersection of the arms at the beginning of the trial and the number of arm entries and time spent in the open arms of the maze were typically recorded over a 5-min trial. This test was developed to provide an index of anxiety-like behaviour in rodents without using aversive stimuli (Pellow et al., 1985). The test was included because there have been reports that the alcohol-preferring P rats spend less time in the open arms than do the alcohol-non-preferring NP rats (Stewart et al., 1993), although other studies have been less conclusive (see Overstreet et al., 1997a).

Forced swim test

The forced swim test involved placing the rat in a cylinder (18 cm diameter, 40 cm tall) with enough water at 25°C so that the rat could not touch the bottom with its hindpaws. Immobility (the absence of movement in three of the four paws) was recorded during a single 5-min session. This test was included, because the FH/Wjd rats are typically very immobile in the test (Rezvani et al., 1990; Overstreet and Rezvani, 1996) and this test has been characterized as a useful screening test for antidepressant drugs (Borsini and Meli, 1988; Overstreet et al., 1992). The FH/Wjd rat may represent an animal strain which has a co-morbid depression/alcoholism (Overstreet et al., 1992), but it is not known whether these are separate disorders or different aspects of the same disorder, e.g. serotonin dysfunction.

Statistical analysis

Data are presented as means ± SEM. Significant differences among the different groups were determined by one-way ANOVAs, followed by Newman–Keuls post-hoc tests.


The data on the different substrains of inbred Fawn-Hooded rats are summarized in Table 1. There were substantial and significant strain differences for almost every measure, but the pattern of inter-strain differences varied with the measure. For example, the FH/Wjd voluntarily drank significantly more alcohol than the FH/Har, as expected (Overstreet and Rezvani, 1996), but also significantly more than the FHH/Eur and FHL/Eur strains. However, the FH/Wjd rats did not exhibit significantly higher total fluid intakes than the FHH/Eur or the FH/Har rats, indicating that high fluid intake and high voluntary alcohol intake are separate characteristics of the FH/Wjd rats (Table 1).

View this table:
Table 1.

Behavioural differences among different inbred FH substrains

Intake (g/kg/day) of
SubstrainFoodTotal fluidAlcoholSaccharinImmobility (s)Time in open arms (s)
*P < 0.05, **P < 0.01, ***P < 0.001.
Data represent the mean values ± SEM for five to 10 rats. Groups with different superscript letters are significantly different, Newman–Keuls test. ND, not determined.
FH/Wjd60 ± 1a124 ± 9a5.9 ± 0.7a475 ± 32a183 ± 5a42 ± 7a
FH/Har66 ± 2a108 ± 4a2.7 ± 0.4b307 ± 23b108 ± 9b9 ± 2b
F154 ± 1b91 ± 5a,b4.2 ± 0.3c300 ± 35a,b144 ± 12a,b17 ± 7b
F267 ± 2a119 ± 4a2.9 + 0.4b314 ± 30a,b141 ± 10a,b22 ± 3b
FHH/Eur69 ± 2a124 ± 9a1.8 + 0.9b307 ± 33a,b187 ± 8aND
FHL/Eur62 ± 2a76 ± 4b1.6 + 0.4b427 ± 34a185 ± 7aND

The FH/Wjd rats were significantly more immobile than the FH/Har rats, replicating our earlier findings (Overstreet and Rezvani, 1996). However, for this variable, the FHH/Eur and FHL/Eur strains were also significantly more immobile than the FH/Har rats (Table 1). Thus, the strain distribution pattern for immobility in the forced swim test is different from that for voluntary alcohol intake.

The findings for voluntary saccharin intake, which tends to be associated with voluntary alcohol intake in many strains (e.g. Overstreet et al., 1993), revealed yet a third strain distribution pattern. The FH/Wjd strain drank the most saccharin and the FH/Har drank the least, as expected (Overstreet and Rezvani, 1996), while the FHH/Eur and FHL/ Eur strains drank intermediate amounts (Table 1).

A final important observation from Table 1 is that the FHH/Eur and FHL/Eur strains do not differ on any variable, except total fluid intake. This finding suggests that the behavioural variables assessed in this study are unlikely to be associated with differences in high blood pressure in the two strains.

Although the FH/Har rats drank significantly less alcohol than the FH/Wjd rats, their intake is generally higher than typically seen in alcohol-non-preferring lines (Table 1, see also Overstreet et al., 1997a). We, therefore, selected the ACI strain, reported in earlier studies to drink very little (Li and Lumeng, 1984). We obtained a sample from Harlan Sprague–Dawley initially, but found that this substrain also drank around 2 g/kg, a much higher amount than previously reported (Li and Lumeng, 1984; Overstreet et al., 1997a). We then obtained some ACI/N rats from the NIH (courtesy of Carl Hansen) and established our own breeding colony once the first group studied drank an average of 0.45 g/kg/day.

A complete summary of the data for the FH/Wjd and ACI/N male rats and their reciprocal crosses is presented in Table 2. The strain distribution patterns for the key variables of voluntary alcohol intake, immobility in the forced swim test, and saccharin intake are illustrated in Figs 1, 2 and 3, respectively.

View this table:
Table 2.

Behavioural differences among FH/Wjd and ACI/N rats and their reciprocal F1 crosses

Intake (g/kg/day) of
GroupTotal fluidAlcoholSaccharinPhenylalanineImmobility (s)
Data represent the mean values ± SEM for 19–25 rats. Groups with different superscript letters are significantly different, Newman–Keuls test.
FH/Wjd152 ± 7a5.4 ± 0.5a476 ± 36a174 ± 19a191 ± 6a
ACI/N84 ± 3b0.6 ± 0.2b161 ± 12b64 ± 4b85 ± 11b
AF F1104 ± 6b2.6 ± 0.3c241 ± 18b148 ± 12a200 ± 7a
FA F1130 ± 8a2.5 ± 0.3c385 ± 30a162 ± 16a191 ± 7a
Fig. 1.

Voluntary alcohol intake of FH/Wjd and ACI/N rats and their reciprocal F1 crosses.

Rats were trained to drink alcohol by a standard protocol of 1 day of water only, 3 days of alcohol only, and then 10 days of choice. The data represent the mean values ± SEM in g/kg alcohol intake for 19–25 rats. Groups with different letters are significantly different.

Fig. 2.

Immobility in the forced swim test of FH/Wjd and ACI/N rats and their reciprocal F1 crosses.

Rats were given a 5-min exposure to the swim tank either before or 2 weeks after exposure to alcohol. The data represent the mean values ± SEM of immobility in seconds for 19–25 rats. Groups with different letters are significantly different.

Fig. 3.

Saccharin intake of FH/Wjd and ACI/N and their reciprocal F1 crosses.

Rats were given a choice between 0.1% saccharin and water for a single day. The data represent the mean values (ml/kg) ± SEM saccharin intake. Groups with different letters are significantly different.

The FH/Wjd rats differed from the ACI/N rats on all of the variables summarized in Table 2. They drank more alcohol and more saccharin, had a high total fluid intake, and exhibited increased immobility in the forced swim test. Indeed, except for alcohol intake, which was much lower, the ACI/N rats closely resembled the FH/Har rats (cf Table 2 with Table 1). The FH/Wjd and ACI/N rats did not differ for food intake or time in the open arms of the plus maze, so these variables were not included in Table 2.

Alcohol intake in the reciprocal crosses (ACI × FH, FH × ACI, male parent first) was intermediate between the two parental strains (Fig. 1), suggesting that additive genetic factors were influencing alcohol intake. In addition, there was no significant difference in alcohol intake between the two crosses, suggesting the lack of maternal influences on alcohol intake.

In contrast to the pattern seen for alcohol intake, the ACI × FH and FH × ACI rats tended to exhibit exaggerated immobility in the forced swim test, resembling more the FH/Wjd strain than the ACI/N strain (Fig. 2). Again, the FH × ACI and ACI × FH rats had very similar immobility scores, suggesting no maternal influences.

The data for saccharin intake revealed yet a different strain distribution pattern (Fig. 3). As expected, the alcohol-preferring FH/Wjd rats drank significantly more saccharin than the alcohol-non-preferring ACI/N rats. However, the ACI × FH and FH × ACI rats exhibited different amounts of saccharin intake, suggesting the influence of maternal factors. Phenylalanine intake was also studied in these rats, because of reported differences in inbred mouse strains (Bachmanov et al., 1997). As shown in Table 2, the two F1 crosses and the FH/Wjd rat drank significantly more phenylalanine solution than the ACI/N rats.

Subgroups of the FH/Wjd, FA, AF, and ACI/N rats were tested in the swim test first and others were tested in the swim test 2 weeks after completion of the screening for alcohol intake. There were no significant differences among these subgroups, either for alcohol intake or for immobility, so for all further testing the swim test was measured first, approximately 3 days before the start of the screening for alcohol intake.

To date, 43 male F2 progeny have been tested in the various conditions employed on the parents and F1 progeny. Analyses of these data with Pearson product–moment correlation revealed significant positive correlations between saccharin intake, forced alcohol intake, and phenylalanine intake (r = from +0.44 to +0.60, P < 0.01), but not between these variables and voluntary alcohol intake or immobility (Table 3). Immobility in the forced swim test did not correlate with any other variable, whereas both phenylalanine intake and voluntary alcohol intake were correlated with forced alcohol intake (r = +0.44, P < 0.01). These analyses confirm the independence of saccharin intake from alcohol intake and immobility suggested in Tables 1 and 2 (see Table 3 for all correlations). Further evidence of the independence of these measures came from the distribution patterns of the F2 data for the three parameters, as illustrated in Fig. 4. The data for alcohol intake resembled a normal distribution, but those for immobility did not. The data for saccharin intake also did not resemble a normal distribution, nor did they resemble those for immobility (Fig. 4).

View this table:
Table 3.

Intercorrelations among behavioural measures in FH× ACI F2 progeny

MeasureAlcohol intake (V)Alcohol intake (F)Saccharin intakePhenylalanine intakeImmobility
V, voluntary; F, forced.
*P < 0.05; **P < 0.01.
Alcohol intake (V)0.44**0.120.10–0.16
Alcohol intake (F)0.44**0.60***0.44**–0.03
Saccharin intake0.120.60***0.59***0.19
Phenylalanine intake0.100.44**0.59***–0.04
Fig. 4.

Distribution of scores for alcohol intake, immobility, and saccharin intake in male F2 progeny from an FH × ACI cross.


The present findings have reinforced the conclusion made earlier about the FH/Wjd and FH/Har substrains of Fawn-Hooded rats, namely that different inbred strains coming from different sources do not behave similarly (Overstreet and Rezvani, 1996). On the other hand, the FHH/Eur and FHL/ Eur strain, developed from a commercial Fawn-Hooded strain in Europe by selective inbreeding for differences in high blood pressure (Provoost and DeKeijzer, 1993), were similar for most behaviours. Yet these two strains resembled the FH/Wjd for immobility, but the FH/Har for alcohol intake. This outcome provides strong evidence that high immobility and high voluntary alcohol intake, both of which have been associated with serotonin dysfunction (Overstreet et al., 1992), are probably not linked. Earlier literature purporting to use FH rats needs to be carefully scrutinized. It is likely that some of the discrepancies in previous studies can be attributed to the investigators using different substrains (Aulakh et al., 1993; Lahmame et al., 1996).

This difference among subgroups of inbred strains is not a novel finding. It is very common in inbred mice. It appears that the ACI/N and ACI/ Hsd (Harlan Sprague–Dawley) also differ on a key behavioural variable. While the ACI/N substrain avoids alcohol (Fig. 1, see also Li and Lumeng, 1984), the ACI/Hsd substrain drinks modest amounts (Overstreet et al., 1997a). If an investigator had selected the ACI/Hsd and FH/Har substrains to study, there would have been very few differences (Overstreet and Rezvani, 1996; Overstreet et al., 1997a) and little interest in follow-up studies. Clearly, if one is interested in studying the genetics of voluntary alcohol intake, the best choice for parents are the FH/Wjd and ACI/N strains, where there is an almost 10-fold difference in alcohol intake (Fig. 1). It is of particular interest that these two strains also differ quite dramatically on two additional key variables: voluntary saccharin intake and immobility in the forced swim test (Figs 2 and 3).

The fact that the strain distribution pattern for immobility in the forced swim test did not resemble that for voluntary alcohol intake (Tables 1 and 2; Fig. 4) was expected. In our previous factor analytical study, immobility was not associated with the alcohol intake factor; in fact, some alcohol-preferring rat lines are less immobile in the swim test than their alcohol-non-preferring counterparts (Overstreet et al., 1997a). Nevertheless, there are substantial and significant differences in both immobility and voluntary alcohol intake in the two parental lines. Simultaneous measurements of both variables in conjunction with genome screening in F2 progeny may reveal unique Quantitative Trait Loci (QTL) for each variable.

The present study has provided evidence that questions the closeness of the relationship between voluntary saccharin and alcohol intake. Many previous studies, including several from our own laboratory, have suggested a positive association between saccharin and alcohol intake in animals (Ninomiya et al., 1984; Kampov-Polevoy et al., 1990, 1996; Gosnell and Krahn, 1992; Sinclair et al., 1992; Bisaga and Kostowski, 1993; Overstreet et al., 1993) and between sweet liking and alcoholism (Kampov-Polevoy et al., 1997). However, there have been other instances where the association has broken down. Overstreet et al. (1996) reported that a line of rats selectively bred for high 5-HT1A receptor sensitivity drank more saccharin, but not more alcohol, than its control line. Rats which have been selectively bred for differences in saccharin intake do not drink differential amounts of alcohol (Badia-Elder et al., 1994). Finally, several genetic studies in mice have failed to report overlapping genetic control for the consumption of sweets and the consumption of alcohol (Phillips et al., 1994; Melo et al., 1996; Bachmanov et al., 1997). Note that each of the above examples has a genetic component. The data in this study, which strongly question the association between saccharin and voluntary alcohol intake, are also genetic in nature.

Despite the lack of support for a close association between voluntary saccharin and alcohol intake, there is a large difference between the two parental lines for each variable. Therefore, obtaining saccharin intake data as well as alcohol intake data in the F2 progeny, followed by genetic screening, will permit a more conclusive statement about the possible overlap of genetic factors influencing these two variables. The fact that Brown et al. (1996) have conducted a genetic study on hypertension in the FHH/Eur and ACI/Hsd rats indicates that a similar study on alcohol intake and related behavioural measures is quite feasible.

In summary, the present communication has presented findings indicating that different substrains may exhibit quite different behaviours and urges readers to verify the origin of any rat line/strain. In addition, although the FH/Wjd and ACI/N rats differ quite dramatically for voluntary alcohol intake, saccharin intake, and immobility in the forced swim test, preliminary analyses of F1 and F2 progeny suggest that these behaviours are influenced by independent genetic factors.


We wish to thank Elijah Clark Jr, Lee Gause, and Ying Yang for technical assistance. The fruitful exchange of ideas with Michael Tordoff of the Monell Chemical Senses Center in Philadelphia is also acknowledged.


  • * Author to whom correspondence should be addressed.


View Abstract